Google Web Accelerator

Lukey Boy writes "Google has released a free web accelerator product for both Firefox and Internet Explorer. According to their information page the software uses Google servers as a proxy for web content, delivering the pages to your system more rapidly and compressing them beforehand."

Comments

I keed! I keed!

[coupland]

I'm using it now and couldn't be happier! It's already saved me over 10 seconds, and there's no catch!

---
Find Google results for "catch"
Sign up for free webmail at http://gmail.google.com/
Resistance is fut... er... Try Google, we're not evil!

I see they've compressed slashdot's comments...

[Anonymous Coward]

Nothing for you to see here. Please move along.

Re:I see they've compressed slashdot's comments...

[Anonymous Coward]

The comments? Recently it seems they could replace slashdot itself with that. That or they could automatically remove dupes to reduce bandwidth...

Smart. Scary.

[lecithin]

Cute...

First, they collect your search information. Next they collected your email. Now they collect your destination. You put it all together, that is quite a bit of information.

What is next?

Very Smart..Very Scary...

Tinfoil, Post!

Well I'll Be Damned

[Anonymous Coward]

It's in beta, who'd have thunk it?

Start your engines gentlemen!

[grazzy]

And let the conspiracy theories begin!

Re:Start your engines gentlemen!

[Anonymous Coward]

Ok, listen. The complaining is not primarily about conspiracy. The problem with Google doing the things they do isn't that we think they're evil. The problem is that a) power even corrupts saints and b) Google is a publically traded company and will at some point not be directed by people who try not to be evil. The concentration of information in Google's databases is critical. People with experience in protecting data know that the only way to ensure that data isn't abused is to make sure that it is not stored. Google's success with a monolithic approach is very detrimental to the development of privacy enhancing algorithms which need to avoid high concentrations of information. Centralized algorithms are the easy way out, and we will pay the price some day.

Woohoo!

[versiondub]

Hopefully I can just get google to hack a server, give me any and all illegal information, serve me with a summons, take me to court, jail me, and then parole me in the future without any effort on my part!

Google turns Evil

[weasello]

When is google going to learn that aggregation is not the way of the future? They will eventually become so large their shareholders will be able to turn them into a giant evil machine, much lik current companies.

Re:Google turns Evil

[Anonymous Coward]

When is google going to learn that aggregation is not the way of the future?

Google's value: $4.8 billion
weasello's value: $29.93

Re:Google turns Evil

[Anonymous Coward]

that's $13.37

Re:Google turns Evil

[weasello]

Turns out that I don't value billions of dollars as an asset; it's a liability. You have 4.8 billion dollars worth of decision making at your helm, which can affect quite a few people. But if I screw up and have to correct my mistakes, I just have to squeegee windows for a day to make up the difference. :P

Re:Google turns Evil

[bprime]

Brand new account and already spreading the FUD. Google's hardly perfect, but so far they've done a great job of fulfilling their "do no evil" mission statement. I'm sure they're well aware of the potential loss of control to stockholders.

Re:Google turns Evil

[weasello]

SO FAR... That's why they're TURNING evil. Their biggest mistake was to go public - it was a method to simply get a rapid infusion of cash so that they could... what? Is not going public the epitome of evil in the first place? Google no longer has control over their mission statement; the shareholders do. It can change the day after they get 100% market penetration.

Re:Google turns Evil

[EpsCylonB]

When is google going to learn that aggregation is not the way of the future? They will eventually become so large their shareholders will be able to turn them into a giant evil machine, much lik current companies.

so whats your solution ?, not allow companies to grow beyond a certain size ?, lets give google the benefit of the doubt until they do start behaving evily.

Re:Google turns Evil

[Anonymous Coward]

And don't forget that common disclaimer "They can change their policies whenever they want without telling anybody."

"But AC, then I'll just stop using them when I find out."

By then, it is too late.

Re:Google turns Evil

[joeljkp]

What makes you think that going public equals evil? Because you consider certain public companies evil, so it must be a trait of the entire system?

Re:Google turns Evil

[hattan]

I agree with the statement that going public doesn't necessarily equate to going public. I think the big fear is the loss of power to shareholders. However that won't happen at google. I've just completed a research paper for school on their governance practices and its interesting to note that company executives have a different kind of stock than most folks out there. Their stock is worth exactly the same economically but offers substantially more voting rights. Basically that means, the majority of power will still be in the hands of the company and hot the shareholders, so taking that into account I don't think they are turning evil.

Re:Google turns Evil

[PReDiToR]

lets give google the benefit of the doubt until they do start behaving evily

Those cheeky North Koreans have funny little nucular[sic] weapons, let's give them the benefit of the doubt too, at least until they start launching them.

What happens when there is a hostile takeover of Google, Inc. by Microsoft, Claria, Yahoo or even a spam queen? Will it be too late to think of an archive of information with proportions like that in terms of risk then?

The risk is now, the risk is here. Denial is a river in Egypt, not a way to live.

Re:Google turns Evil

[EpsCylonB]

Whats does google have to gain by telling everyone that you like porn ?, if your really that worried don't use it.

Re:Google turns Evil

[Surt]

That's not a good strategy. That's like saying, lets let nations get as powerful as they want until they drop the first nuclear bomb.

Re:Google turns Evil

[Bronster]

What makes you think that going public equals evil?

a) Going public means you are required to maximise shareholder value.
b) In the short term, going evil is very profitable, trading goodwill and other unmeasurable 'intangibles' for gold mulah.
c) QED

Thanks for playing.

Re:Google turns Evil

[Anonymous Coward]

Turns out that I don't value billions of dollars as an asset; it's a liability.

Dude, let me know if you run into any significant liabilities of this kind. I would be happy to take them off your hands.

Re:Google turns Evil

[Anonymous Coward]

I agree with the statement that going public doesn't necessarily equate to going public.

How very Zen.

Re:Google turns Evil

[Kris_J]

Google already caches a huge amount of the 'Net. Why do you think this is so far away from their core business? I'd say that Gmail was more of a stretch than this.

Re:Google turns Evil

[cicho]

Because you are no longer free to take the company in the direction you want. Wait until a startup figures out a better way to search than Google has, and Google starts losing ad revenue big time. That's when they start cashing in on all the other uses for the data they have.

GWA, GMail, google toolbar, search history - these are not good deeds. They are investments.

Re:Google turns Evil

[jdgeorge]

Because you are no longer free to take the company in the direction you want. Wait until a startup figures out a better way to search than Google has, and Google starts losing ad revenue big time. That's when they start cashing in on all the other uses for the data they have.

GWA, GMail, google toolbar, search history - these are not good deeds. They are investments.

First of all, somebody better give Larry Ellison, Scott McNeally, and Bill Gates wake-up calls to tell them they can no longer take their companies in the directions they want because they are publicly traded. Oh, wait, right, they still do control their companies. Becoming publicly traded does not equate to losing control over a company, as long as you control the majority of voting shares.

Second, some of these neato keen features which are arguably a sign of Google's turn to evil, these "investments" in GWA, GMail, Google toolbar, etc. (which admittedly is a fair description from a business standpoint), were released quite a while before Google's IPO. These features arguably may be bordering on evil, but they have absolutely no relation to any loss of control Google might experience as a result of going public.

All this implies is that Google was always on its current track. If that track is evil, then they were always on the track to evil; their IPO has not changed that in any case. This is, of course, what you should expect; the company is still controlled by the founders, not by the common stockholders.

Re:Google turns Evil

[Anonymous Coward]

Depending on whenever you are for 1984 style bigbrotherism or not, google may already be evil.

Re:Google turns Evil

[lachlan76]

Those cheeky North Koreans have funny little nucular[sic] weapons, let's give them the benefit of the doubt too, at least until they start launching them.

Substitute North Koreans for Americans and see how different your opinion is.

Re:Google turns Evil

[PReDiToR]

Being a Brit, it hasn't changed one bit.

Both countries are run by tyrannical manipulative dictators who haven't any respect for law or democracy.

I apologise to any North Koreans reading this.

Hmm,

[killa62]

But how does it know how many minutes you save?

Re:Hmm,

[CyanDisaster]

... But how does it know how many minutes you save...

I assume it would calculate your current download speed as well as the size of the information you're retrieving, then do the same based on going through Google's servers, and come up with an approximate value of saved time.

Something like that anyhow I think.

Hope be with ye,
Cyan

Re:Hmm,

[SSalvatore]

(ORIGINAL_Kb - COMPRESSED_Kb) / bandwith (Kbps) ?

I know, it is phony anyways. If I have a connection to google that is much slower than the connection to the real page's server, then I may actually waste time by getting the page from google as opposed to the orig server.

Re:Hmm,

[j_hirny]

But how does it know how many minutes you save?

How many times you clicked on some AdWord * g, also known as google constant.

Re:Hmm,

[caluml]

Imagine if it sent you a bill for the money you saved at the end of the month :)

Slashdot effect?

[blackmonday]

Could this solve the slashdot effect problem, if we're all running it? Are ads associated with it?

Re:Slashdot effect?

[LewsTherinKinslayer]

Are ads associated with it?

I'd imagine anyone running this would have AdBlock as well. I know I do.

Re:Slashdot effect?

[Anonymous Coward]

Could this solve the slashdot effect problem, if we're all running it?

What, and slashdot google?

Re:Slashdot effect?

[tcpli]

It can help to damping the slashdot effect by providing extra cache (copy) at google's server

Re:Slashdot effect?

[Ninwa]

That's actually an interesting topic, and maybe it should be a feature for subscribers for a few extra dollars. I mean, have Slashdot mirror (on slashdot servers) all linked pages in the articles for subscribers only.

Re:Slashdot effect?

[winkydink]

You mean like this?

Re:Slashdot effect?

[EpsCylonB]

good question, I am not sure how this prefetching works, I not sure we can count on google knowing about some tiny site showing off the latest silly case mod before the slashdot crowd, then again.

Re:Slashdot effect?

[jangobongo]

The "What Webmasters Need To Know About Google Web Accelerator" page touches on this:

Will Google Web Accelerator affect my server load or usage statistics?

It depends on whether your pages are cacheable. You can identify page requests prefetched by Google Web Accelerator through the HTTP header X-moz: prefetch. You can learn more about this header on the Mozilla website.

Am I reading this right? If the page has been cached at Google, Google will use that cache for the preloading. And webmasters can do certain things to aid the prefetching function.

So it sounds to me that if the website being slashdotted is cache-able (and the slashdotters have this accelerator), it could ease the website's server load.

Re:Slashdot effect?

[doctong]

Yes! I just tried loading up a link from Slashdot to How Stuff Works on my Linux box and the images weren't coming in. Slashdotted!

Curious about this very question, I tried the same link on a PC box with the accelerator and it came up instantly.

-MB

Do no evil, i hope

[neoform]

Considering this looks like a way for google to simply track every site i visit.. i sure hope they really aren't.. "evil".. :-/

Re:Do no evil, i hope

[Signal_Noise]

They aren't

Someone explain?

[eth8686]

How is it faster to use a proxy than to just get the content straight from the server?

Re:Someone explain?

[SSalvatore]

The larger your proxy is, the more probabilities you have of finding that page that you are looking for in the proxy's cache.

Google has a tremendous amount of cached pages and images.

That's where the real gain comes from.

Plus, the trad proxys that I know do not compress. I don't know how relevant this is because for the most part, web pages' heavy stuff is not text but images that are already compressed. Still, this will save you some time.

Then keep in mind that the google servers are probably faster than other servers and more resistant to slashdot-and-the-likes web tsunamis.

Re:Someone explain?

[Spy der Mann]

How is it faster to use a proxy than to just get the content straight from the server?

If the pages are static, I think google could use the same copy for different IP's.

But this would also mean that google would get a massive overload when people start using it. Frankly I doubt that even google can provide the infrastructure to become a global centralized proxy.

Re:Someone explain?

[PReDiToR]

Frankly I doubt that even google can provide the infrastructure to become a global centralized proxy

Well it just so happens that Google just got a few Trillion dollars from Scott Richter for the whole Gmail accounts database.

Apparently Larry and Sergey were thinking about buying Switzerland, but someone said that this idea was more geeky and only cost $5.00 more.

free webstats

[fitsnips]

will they provide you with your web surfing trend stats?

Re:free webstats

[Anonymous Coward]

No, you dont get that information, CIA/TIA does.

The irony..

[OlivierB]

Webaccelerator's page is slashdotted...

Re:The irony..

[mattdev121]

Quick! What's the google mirror... Oh Wait....

hmmm

[willscott]

is this really able to speed stuff up if you have broadband? not sure if i really belive them.

Re:hmmm

[cornjones]

sure, there is no real difference between the functionality on broadband vs dialup. If you compress the data you are sending fewer bits down the pipe. Whether those bits are travelling at 5kps or 500kps, the total download will be quicker than w/o it.

Of course, the difference between it taking 10 seconds compressed vs 13 seconds uncompressed on dialup (and really good compression rates) is noticeable whereas 1 second compressed/1.3 seconds uncompressed may not be.

Re:hmmm

[kagelump]

we can say that this is more related to accessing google cache directly...
for example, you normally access sites quickly with broadband
but this should be able to speed up various 'slashdotted' sites
where the bottleneck is at their servers, not yours

Break out the tin foil

[djtripp]

Sounds great. But if you want your viewing habits documented, well, this is not product for you. If you don't mind, go for it.

More info

[ranson]

More information about GWA is posted here: http://blog.searchenginewatch.com/blog/050504-1453 07 Also, browsers other than Firefox and Mozilla can take advantage of GWA if you set them to proxy requests over Localhost:9100 while GWA is running in the system tray. It should also be pointed out that this is apparently geared towards broadband users.

Re:More info

[alexburke]

Does it *ONLY* listen on the localhost interface, or will machines without firewalls become open proxies? Surely they thought of this and locked it down to localhost-only, but this reminds me of the old Norton AntiVirus (2003?) POP email scanning tray-icon that listened on *:110 as a virus-scanning POP proxy, opening up unfirewalled boxes to remote DoS exploits once a buffer overflow was found in it...

Re:More info

[Kris_J]

Well, I use the proxy product "Proxomitron" and it has the option to either only listen to the local PC (default), or open it up to an IP range (I use this on a server to bounce requests through a separate network -- or I did, the Google Web Accelerator product doesn't work if I do this, and appears to have no way to get it to pass its requests through a proxy.)

isn't this basically a proxy?

[yagu]

I've RTF(ine)A and I give... what makes this different/better/faster/whatever than a proxy server?

And, while I'm at it.... I submit my vote that Google make linux/*nix versions of their stuff more quickly/readily. I find it no small irony that a company that relies on over 10,000 linux servers (actually I think the number may exceed 40,000) essentially making them one of the largest benficiaries of the OSS community they don't yet have a Google Desktop, nor are offering a beta of this accelerator for the linux community.

Don't get me wrong, I like Google, think they've done great stuff, but come on -- how about paying back a little to the hand that giveth.

Re:isn't this basically a proxy?

[Mifflesticks]

"Don't get me wrong, I like Google, think they've done great stuff, but come on -- how about paying back a little to the hand that giveth."

Well, they employ a guy to work full-time on firefox....

Re:isn't this basically a proxy?

[Anonymous Coward]

It's because they already know that Linux users wouldn't allow their web browsing to tracked, documented, and compiled for future usage.

Re:isn't this basically a proxy?

[Zocalo]

what makes this different/better/faster/whatever than a proxy server?

Nothing really that I can see other than that it will always compress which is something that some sites do not have enabled, which should offer some speed ups and help reduce over all web traffic. I'd assume that this is tied into Google's cache used on the search engine, so if you request a page through the proxy for which the cached data is stale it will update that also, then re-index the data for the search engine. If so, this could be *very* useful for alleviating things like the Slashdot effect, although it would need to pull the graphics to be of any real use here. The problem with caching the graphics though, is that it's going to make it *really* difficult for Ad-Blockers to work out which files are ads and which are not...

Re:isn't this basically a proxy?

[Anonymous Coward]

oooops... forgot about that... certainly a heck of a lot more than many others do, but still... I've used their Google Desktop on windows, and LOVE it... tried but failed to get Beagle up and running on linux. Really want one or the other for my linux platforms. I know it will arrive eventually but sometimes frustrated at the lag time. Normally I'm happy to wait. Sigh. (But, thanks for the reminder about the firefox thingy)

Re:isn't this basically a proxy?

[_undan]

Draw a circle. This is all the people using a F/OSS desktop environment.

Now, draw another circle inside that one, almost exactly the same size, but not quite. These are the F/OSS zealots who won't install anything unless it's GNU licenced.

The area between the boundaries of those two circles are the only people who would install it. And I don't know about you, but I'm pretty sure the other guy in that part of the chart understands that.

Re:isn't this basically a proxy?

[natrius]

I find it no small irony that a company that relies on over 10,000 linux servers (actually I think the number may exceed 40,000) essentially making them one of the largest benficiaries of the OSS community they don't yet have a Google Desktop, nor are offering a beta of this accelerator for the linux community.

Uh, it's not ironic at all. As you said, they use Linux servers, not desktops. Those servers don't need Google Desktop or Webaccelerator.

Don't get me wrong, I like Google, think they've done great stuff, but come on -- how about paying back a little to the hand that giveth.

You think they're trying to do Windows users a favor by releasing these products? They're doing it for themselves. They make money off of these products by solidifying their mindshare and marketshare. Releasing Linux versions (or OS X versions, for that matter) obviously isn't worth it to them.

Re:isn't this basically a proxy?

[LiquidCoooled]

Your ad blocker won't know if its come from a google cache, or from the original site?
The URL will be the same, the connection details will remain the same, how can there be any difference?

Your files already go through many servers and proxies before they end up on your screen, and you can still determine the path of them.

I think your right about the refreshing THEIR index using your selections (hey, we only need to refresh the stuff thats used, right?)
It helps keep them ontop of the game and fresh.

They could also then remove the delay between changing content on your site, and the adverts adapting (after a refresh). The adverts would then be as live as they currently are with gMail.

Re:isn't this basically a proxy?

[CvD]

Want a Linux (unix, os x, etc) version? Send them an email asking for it:

mailto:labs+webaccelerator@google.com

Perhaps if they get enough responses they will get it out faster (or decide to put it out if they had decided not to publish it).

Re:isn't this basically a proxy?

[Anonymous Coward]

Google Desktop for Linux? Perhaps Linux should start by having a desktop. Or are you implying Google should build a desktop for Linux and then port Google Desktop to it?! You're asking a bit too much.

Re:isn't this basically a proxy?

[YrWrstNtmr]

how about paying back a little to the hand that giveth.
...over 10,000 linux servers (actually I think the number may exceed 40,000)

It appears they are.

Re:isn't this basically a proxy?

[HG2]

Spamming them is not going to make it any faster...

Re:isn't this basically a proxy?

[Anonymous Coward]

The problem with caching the graphics though, is that it's going to make it *really* difficult for Ad-Blockers to work out which files are ads and which are not...

And it's really good that there aren't many web sites that rely on hit counts to be able to set ad rates...

Re:isn't this basically a proxy?

[Anonymous Coward]

"...what makes this different/better/faster/whatever than a proxy server?"

The question misunderstands the best answer. It's not JUST A proxy server. It's the MOTHER OF ALL proxy servers.

Re:isn't this basically a proxy?

[Manw3]

Doesn't the fact that they use Linux servers mean they ARE "paying back back a little to the hand that giveth"?

Re:isn't this basically a proxy?

[yagu]

I wouldn't consider it spam if it's a letter-writing campaign. A concerted plea from a community is the heart and soul of how a lot of things get done. Please don't associate or confuse letter-writing campaigns with advertising whores.

After all, the poster just lists an e-mail address to which you must compose your own request... it's not a click on an ASP button that generates some pre-canned spam. Heck, he doesn't even pre-populate the subject field!

Re:isn't this basically a proxy?

[Kris_J]

I subscribe to Opera's compression/optimistation proxy for their mobile web browsers (I have Opera on my N-gage). The concept is simple and sound. A browser and proxy pair can adopt techniques that are not general standards. The boring, original page is fetched by the proxy, then highly optimised for the (known/modified/extented) browser. The result is a smaller download.

To give an example, Google could implement Allume's JPEG compression on images such that the JPEG data your browser now downloads from Google's server is 20-25% smaller. No standard proxy or browser sould support this, but Google's proxy and Google's Firefox extension could.

Re:isn't this basically a proxy?

[kbjnash]

On the contrary, wouldn't you consider this giving back? What could be nicer of them then offering the dial up masses (I don't see much benifit for broadband users) a faster intenet connection thru proxies? As long as they don't do evil and don't use our surfing trends against us, then what is the harm? Just because a company benifits from OSS, doesn't automatically mean they have to come up the their own distro...

Re:isn't this basically a proxy?

[Anonymous Coward]

I find it no small irony that a company that relies on over 10,000 linux servers (actually I think the number may exceed 40,000) essentially making them one of the largest benficiaries of the OSS community they don't yet have a Google Desktop, nor are offering a beta of this accelerator for the linux community.

1. I thought the majority of desktop users' computers run some version of Windows.
2. YOur comment make me wonder the following. What OS does Google use in development, nevermind the actual search servers? Maybe they would be obligated to make source code available if they released binaries?

Re:isn't this basically a proxy?

[agusus]

>> What OS does Google use in development, nevermind the actual search servers?

They mainly use linux. Some windows of course, obviously including the teams that develop the windows products, but linux is preferred by most I think.

So that raises the point - google doesn't just have thousands of linux servers, they have a couple thousand linux desktops too! So why don't they make a Google file search for linux? updatedb/locate certainly isn't perfect, so the google app could be better.

Re:isn't this basically a proxy?

[Air-conditioned cowh]

Actually by releasing Windows-only products they are signing their own death warrant.

They are making it that much more difficult for anyone planning a desktop migration to Linux because there is now one more must-have application that the users will demand that isn't available in Linux.

By halting Linux desktop migration Google reduce the mindshare of Linux generally. Right now Linux and open source in general needs more mind share so that it reaches the point where large organisations push for legislation and standards that /help/ open source instead of the current situation where we have software patents and more and more closed standards trying to lock open source out of the party.

Re:isn't this basically a proxy?

[HG2]

When was the last time you posted an email address on slashdot with out the email feeling the slashdotted effect?

Re:isn't this basically a proxy?

[vldmr_krn]

how about paying back a little to the hand that giveth. They're paid back plenty--open source coders use Google. Or does only self-sacrificial help count?

Re:isn't this basically a proxy?

[Anonymous Coward]

draw a circle.
now draw another circle.
now draw a circle inside a circle inside a circle.
now draw 3 more circles.
now circle those circles.
now draw a circle inside of that circle (but not exactly) and the draw circles inside the space between the circles.

Re:isn't this basically a proxy?

[Anonamused Cow-herd]

That's such crap. There are plenty of linux users who don't mind using proprietary software; I would argue that most people do in fact use proprietary non-f/oss software on free operating environments. You just have to make it worth it for them

Re:isn't this basically a proxy?

[RhettLivingston]

Actually, what makes this "faster" than "a proxy server" is that its not just "a" proxy server. Its Google. Google's extra fast response / low latency / fat pipe to virtually anyplace on the web combined with caching of almost everything visited with any meaningful frequency makes this like having all of the web a few milliseconds from your ISP. On average, their servers will almost always serve you vastly faster than the originals.

Great for dial up users

[binaryspiral]

This is great news for dialup users that are being charged for this service through their own ISP.

Re:Great for dial up users

[Anonymous Coward]

If you actually read the article, you would note that it's designed to speed up computers with broadband internet.

While it may "work" on dialup, who's to say it's:
1. Going to remain free
2. Actually make browsing faster on dialup

?

Re:Great for dial up users

[LiquidCoooled]

No it isn't, at least not directly.

They specifically say this is designed for broadband, and the ISP level speedups are by a cache reducing the complexity and detail in images.

Now, what do you think would happen if YOUR dialup ISP decided to install and use this technology on their servers?

As in, all your requests are fed through google to speed up the html side of things, then their proxy trims the images as well.

Re:Great for dial up users

[Rikus]

While this would probably seldom provide a speed boost due to better response times or throughput, zlib compression (supported by many "modern" browsers) should help just about anyone by allowing more than one packetworth of data to be crammed into a single one. As long as the computer itself can decompress fast enough, dialup users should benefit from that part.

Re:Great for dial up users

[binaryspiral]

Lucky you posted as a coward... I actually RTFA you dolt. Here's a clip:

Google Web Accelerator uses various strategies to make your web pages load faster, including:

Sending your page requests through Google machines dedicated to handling Google Web Accelerator traffic.
Storing copies of frequently looked at pages to make them quickly accessible.
Downloading only the updates if a web page has changed slightly since you last viewed it.
Prefetching certain pages onto your computer in advance.
Managing your Internet connection to reduce delays.
Compressing data before sending it to your computer.

This sure as heck sounds like accelerating a website (no matter what speed internet connection you have) - including dialup.

A local ISP is selling a service like this for $5/month... Google offers it free and offers a much larger caching engine.

Re:Great for dial up users

[binaryspiral]

Some days I just get too excitied about new stuff... Snarf!

3. Can I use Google Web Accelerator with a dial-up connection?

Dial-up users may not see much improvement, as Google Web Accelerator is currently optimized to speed up web page loading for broadband connections.

Re:Great for dial up users

[Lord Haha]

Considering this tech is mostly used for highspeed internet... I doubt if your dialup ISP (which are quickly going the way of the dodo) would even bother the small improvement you might gain would outweigh the cost of them using it in a for profit situation (as in google is fine with personal use probably, but for a business... or an ISP ... that's probably a different story.)

Is this like...

[cs02rm0]

...a proxy which just compressed stuff on the server and then decompresses it on the client?

Oh... yes.

Google Web Accelerator uses various strategies to make your web pages load faster, including:

* Sending your page requests through Google machines dedicated to handling Google Web Accelerator traffic.
* Storing copies of frequently looked at pages to make them quickly accessible.
* Downloading only the updates if a web page has changed slightly since you last viewed it.
* Prefetching certain pages onto your computer in advance.
* Managing your Internet connection to reduce delays.
* Compressing data before sending it to your computer.

Re:Is this like...

[Anonymous Coward]

and adding the pages to thier index if they dont know about them, so they dont have to crawl for them.

This is a great way to make sure popular pages are fresh in the search engine index.

Re:Is this like...

[NightLamp]

add to that

* ...Whenever your computer sends cookies with browsing or prefetching page requests for unencrypted sites, we temporarily cache these cookies in order to improve performance

Re:Is this like...

[Anonymous Coward]

* Prefetching certain pages onto your computer in advance.

This can get interesting. Who is now responsible for content downloaded to your computer? Can you now blame on Google porn on your work computer and/or child porn on your home computer?

Re:Is this like...

[Lord Ender]

That's all crap! Other than compressing and prefetching, all those things you listen actually make your browsing slower.

Re:I keed! I keed!

[MrAnnoyanceToYou]

Holy Cow! Google got slashdotted!

Hm...

[Ninwa]

I'm dissapointed that it doesn't work with Windows 98, but that aside... how is this different than all of the other "Web Accelerator" products out there?

Re:Hm...

[Vann_v2]

It's free, functional, and doesn't pester you with ads?

Re:Hm...

[Anonymous Coward]

You forgot the word "yet"

Re:Hm...

[airjrdn]

Theirs is "Designed for Broadband", while Propel, and Proxyconn are designed mostly for dial-up users.

What does Google gain from this?

[Spy Hunter]

How does this benefit them? The only thing I can think of is data-mining people's surfing habits to improve their web search results. This could easily drastically improve their ranking of results, but couldn't it also be easily spammed? Is there another reason they might want to do this that I'm not seeing? I imagine this will result in an awful lot of extra bandwidth consumption they'll have to pay for, they must have a good reason...

Re:What does Google gain from this?

[jesseraf]

sounds like spyware in disguise.

Re:What does Google gain from this?

[Spy Hunter]

More ideas for Google's interest in this:

• More aggressive preloading of top search results (made possible because Google is providing the bandwidth, so they're not wasting other people's bandwidth), makes Google search results more responsive => people rely on Google more
• In the future, improved google ad relevancy by serving you ads related to your browsing habits (Sign me up! maybe I'd actually get ads that are useful to me instead of the normal crap ones. You can always turn it off when you want privacy, so stop frothing at the mouth already.)

Of course, people are going to be crying "spyware"! But this is different from most spyware. Firstly, it doesn't clog up your Windows installation and slow down or crash your computer; in fact it speeds up your browsing. Secondly, you can turn it off, or uninstall it if you want. Thirdly, you only get it if you explicitly download it. Fourthly, it might actually improve Google's relevancy for search results and ads, which would benefit me directly. And finally, so many people are watching Google right now that the instant they do something evil, everybody on the globe will know about it. If that happens, it's trivial to switch to a competitor. And that's exactly why they *won't* be doing anything evil.

Re:What does Google gain from this?

[LiquidCoooled]

The adverts should be as fresh as they currently are on gMail.
Because the page you requested is fresh in its cache, the adword lookup will be fully live. There will be no delay between google scanning the content and displaying ads which then become slightly outdated by changing content.

It should work already for people using the service.

A good test would be choose a quiet backwater site with google adverts on, post to the forum with big adword phrases and do a simple refresh.
The original unchanged ads will remain.

Switch on your filter and reload the page. If the ads are different and use your keywords, then we have cracked it ;)

Re:What does Google gain from this?

[MrNonchalant]

Firstly, it doesn't clog up your Windows installation and slow down or crash your computer; in fact it speeds up your browsing. Secondly, you can turn it off, or uninstall it if you want. Thirdly, you only get it if you explicitly download it. Fourthly, it might actually improve Google's relevancy for search results and ads, which would benefit me directly.

Sixthly, it tells you very clearly in the installation process the dangers. On a separate screen from the legalese, in plain and bulleted english. You have to check a box to agree, and even the label for the box gives you some hint.

Re:What does Google gain from this?

[Anonymous Coward]

You forgot:
- Improving search results by giving preference to popular web pages (as in most visited).
- Improving search results by ignoring spam-pages identified by simply looking for well-linked pages that nobody visits.

Re:What does Google gain from this?

[PReDiToR]

If that happens, it's trivial to switch to a competitor

JohnBoy, can you grab that stable door please? I'm going to go look for that darned horse =/

If they started using the information evilly we would all have to change our names, our SS numbers, our vehicles, our email addresses, phone numbers, CC details and who we bank with, our online chat services, friends' URLs, and in fact everything that we use the internet for.

How much of your life can be gleaned from the information in and out of your modem in one week? Emails can be included in this because most of us have Gmail addresses, don't we?

In one week how much would GOOG know about you?

I think anyone having this much information is too much of a risk. They may be evil already but biding their time.

Re:What does Google gain from this?

[Spy Hunter]

I didn't forget, that was in my original post, which seems to have been ignored. Anyway, I think that using this data in search rankings would allow a huge improvement in search result quality, at least at first. We could see a return of the Google of old, with great search results on every topic, where popular keywords haven't been spammed to death. But what happens when people start spamming *this* system? Hired armies of surfers in India or China could skew results. Popular sites could employ hidden frames to inflate the hit counts of other sites. Armies of infected zombie PCs could send Google bogus surfing data, rendering this data source as nearly useless as Google's current sources (page content and links). The war against spam will march on.

Re:What does Google gain from this?

[Mac Degger]

Yeah, yeah...and they have a satelite too now. The real question /is/ 'what are google going to do with this information?'.

My guess: make money. Duh. Nothing else: just make relevant adwords and rake it in. This is a company which made scholar.google.com. I do beleive they are in this for good (and profit!!!) at the moment. But anything else nefferious they do will come out in their papers...remember, they're a public company. We'll know when they start buying an army or searching through webresults/mail for info they shouldn't have just by what they purchase.

Re:What does Google gain from this?

[Bryan_W]

Server-side Spyware prevention

Re:What does Google gain from this?

[Anonymous Coward]

But:
-This drives up the cost of googlebombing. Instead of thousands of automatically generated pages, they have to hire armies of surfers in India or China, who aren't all that cheap.
-Hidden frames are included in the page, which google is caching and can therefore look for relatively easily, and then reducing the pagerank of the source page.
-"Armies of hired surfers" would have very different browsing habits than normal users. If google figures out a way to set them apart, they can ignore all the data from those users.
-Related to the above, if they can classify users according to their browsing habits, they can tailor search results to each group. E.g. Automatically guessing which languages you can read and prefer, whether you read mailing list posts (Note this doesn't mean google has to figure out which pages are from mailing lists, but rather that people with your habits do read those pages). This would have the added benefit of making the armies of hired surfers get different search results than the rest of us.

The biggest threat I see are zombie PCs, as you mention. But that's a bigger problem, one step at a time.

Re:What does Google gain from this?

[QuantumG]

It's not spyware, it's voyeurware. If that's what you're into then all the more power to you, but I think the majority of us would prefer not to share our private browsing habits with Google.

Re:What does Google gain from this?

[Spy Hunter]

True, they should definitely do it. Anything to help slow the tide of spam, even if it only works for a while.

Re:What does Google gain from this?

[volkris]

I think the majority of us would prefer to get something we want and let Google make money off of it especially when it costs us nothing.

Which is the situation here.

This is useless...

[Anonymous Coward]

This product is pretty useless for broadband connections.

The only useful thing that could come from this is being able to view the cache of a website that has been /.'ed to death.

I don't understand why companies continue coming out with products for dialup. Dialup has long been on its way out! Get over it!

It's on par with trying to sell MC Hammer CD's

Re:This is useless...

[Blakflag]

Well the FAQ says the exact opposite:

3. Can I use Google Web Accelerator with a dial-up connection?

Dial-up users may not see much improvement, as Google Web Accelerator is currently optimized to speed up web page loading for broadband connections.

Re:This is useless...

[Anonymous Coward]

Speed up webpage loading for broadband?!

How much more speed do you need? My connection is almost on par with changing the channel on a TV...

I don't think they're gonna be happy until the data is being directly fed into your brain via gold-plated wires. You know, to get rid of that pesky eyesite lag and whatnot.

Re:This is useless...

[Anonymous Coward]

I don't understand why companies continue coming out with products for dialup. Dialup has long been on its way out! Get over it!

Netcraft confirms it.

It works with Firefox!

[EvilStein]

From the FAQ:

6. What are Google Web Accelerator's system requirements?

To use Google Web Accelerator, your computer must have a Windows XP or Windows 2000 SP 3+ operating system. Google Web Accelerator works for the Internet Explorer 5.5+ or Firefox 1.0+ browsers.

7. Can I run Google Web Accelerator on a browser other than Internet Explorer or Firefox?

For other browsers running on Windows, you'll need to manually configure your proxy settings to 127.0.0.1:9100 for HTTP connections.

Re:It works with Firefox!

[WillAffleckUW]

But not on Linux or BSD or Apple (Mac) ...

From the FAQ:

6. What are Google Web Accelerator's system requirements?

To use Google Web Accelerator, your computer must have a Windows XP or Windows 2000 SP 3+ operating system. Google Web Accelerator works for the Internet Explorer 5.5+ or Firefox 1.0+ browsers.

7. Can I run Google Web Accelerator on a browser other than Internet Explorer or Firefox?

For other browsers running on Windows, you'll need to manually configure your proxy settings to 127.0.0.1:9100 for HTTP connections.

Re:It works with Firefox!

[EvilStein]

Wow. Go fuck yourself. IDIOT.

Half lf Slashdot doesn't bother to read the articles, much less the headlines.

Asshole.

Re:Smart. Scary.

[Anonymous Coward]

Next they modify the data you receive to influence your opinion.

Nothing for you to see here. Please move along.

[roman_mir]

Now I can get this message: " Nothing for you to see here. Please move along." - way faster! Thank you Google!

Major privacy concerns

[chrispyman]

The major problem I see with this is the privacy implications. Sure, each individual site you visit nowadays has some information about you in their logs, but Google's proxy would store information on every site you visit. Sure, their privacy policy might be good now, but who's to say it won't change without notice at any point in time, or perhaps the government could get a subpoena for them.

I'm a little too paranoid for this one...

[allism]

At least they were decent enough to point out that you need to READ their Privacy Policy:

• Google Web Accelerator sends requests for web pages, except for secure web pages (HTTPS), to Google, which logs these requests. Some web pages may embed personal information in these page requests.
  
• Google receives and temporarily caches cookie data that your computer sends with webpage requests in order to improve performance.
  
• In order to speed up delivery of content, Google Web Accelerator may retrieve webpage content that you did not request, and store it in your Google Web Accelerator cache.
  

To learn more, read our Google Web Accelerator Privacy Policy (http://webaccelerator.google.com/privacy).

Does anyone know if the accelerator gives you the option to omit certain webpages from your accelerating experience, or is this going to turn into a huge information mine? (Not that the two are exclusive, there are going to be users who just blindly send anything through the accelerator regardless).

Re:I'm a little too paranoid for this one...

[niteice]

Yeah, you can block on a domain level.

Re:I'm a little too paranoid for this one...

[ranson]

Does anyone know if the accelerator gives you the option to omit certain webpages from your accelerating experience, or is this going to turn into a huge information mine? (Not that the two are exclusive, there are going to be users who just blindly send anything through the accelerator regardless). Yes, you can set up exception domains in the GWA settings, along with some other cool settings like prefetch tweaking.

Re:I'm a little too paranoid for this one...

[mikeswi]

Well, domains yes, specific pages, no. And they even let you turn off the autoupdater if you want. First time I've ever seen that from Google.

What I'd to know is how this helps a broadband connection but not dial-up. My connection already loads most pages nearly instantly.

Re:I'm a little too paranoid for this one...

[allism]

Maybe it's more a case of us broadband people helping out all the little dial-up folk...

Re:I'm a little too paranoid for this one...

[MSTCrow5429]

I've installed the thing, and under Preferences, you can select domains to be unaccelerated.

Re:I'm a little too paranoid for this one...

[Crashmaster007]

If they dont allow you to disable some pages they are gonna be getting a lot of porn.

Re:I'm a little too paranoid for this one...

[arkham6]

Yes, it is able to specify sites you don't want it to cache/accelerate.

Right click on the clock icon in your toolbar, select preferences, and at the bottom of the page displayed there is a section that reads "Specify sites that should not go through Google Web Accelerator."

I plan on putting all my banking info in there. Don't want them knowing stuff like that.

Re:I'm a little too paranoid for this one...

[Shrubbman]

Does anyone know if the accelerator gives you the option to omit certain webpages from your accelerating experience

Yes, it lets you specify what pages/domains you don't want it to cache. I just installed it and they made it quite clear this was as option. What I found annoying was it's insistance on placing yet another icon in my system tray. I've already got like half a dozen icons down there, and that's after weeding out the stuff I don't need/want running down there. I couldn't see how to disable the tray icon while keeping the accellerator enabled, so it didn't last 5 minutes on my machine before I was hitting the uninstall.

Re:I'm a little too paranoid for this one...

[yaddayaddaslashdot]

What I found annoying was it's insistance on placing yet another icon in my system tray.... I couldn't see how to disable the tray icon while keeping the accellerator enabled....

At least in XP Professional, if you right-click in the taskbar and then choose "Customize" in the "Notification area" section, you can set any tray icon to "Always hide" to make it disappear.

Re:I'm a little too paranoid for this one...

[Anonymous Coward]

Please do not accidentally follow spam links to kiddie porn. You will prefetch all the images and be in jail for life!

Re:I'm a little too paranoid for this one...

[RevAaron]

If they dont allow you to disable some pages they are gonna be getting a lot of news.

...and? Do you think that Google, being in the position that they are, doesn't know that *gasp* people look at porn?

Re:I'm a little too paranoid for this one...

[aprilsound]

Does anyone know if the accelerator gives you the option to omit certain webpages from your accelerating experience...

At the bottom of the preferences page:

Don't Accelerate These Sites
Specify sites that should not go through Google Web Accelerator. Put each entry in a separate line. Examples:
www.domain.com
.domainsuffix.com

So yes.

Re:I'm a little too paranoid for this one...

[Gopal.V]

> Google receives and temporarily caches cookie data that your computer sends with webpage requests in order to improve performance.

Cookies with session ids are the most common auth systems in practice. So what happens when google gets my cookie ?. (think about your webmail for example).

Re:I'm a little too paranoid for this one...

[Anonymous Coward]

It doesn't accelerate https protocol.

So let me get this straight ...

[Draoi]

... Google will log every URL you visit via their proxy logs. They'll ultimately forward on the requested page with their own AdWords and possibly mask other sites' adverts. Not sure if I like that ..

Re:So let me get this straight ...

[Anonymous Coward]

So let me get this straight: you just made half that up without informing yourself of the facts first. Not sure if I like that.

Re:So let me get this straight ...

[Kris_J]

Well, I've installed it on this and my own personal ad filter is still working. Requests appear to be directed at localhost:9100, but still through the proxy configured in Firefox's settings. I have had to bypass a remote proxy I used to use. This has a few negative implications and I'm not sure if I'll see a net reduction in traffic through our ADSL connections. But I'm a compression junkie, so I certainly want to give it a go.

internet filters

[jaxle]

I wonder if you could get around internet filters with this since it tunnels through google servers. This could be of use to college students who have to deal with unintelligent internet filters at school (like me). I would test it now but I'm on a public terminal.

Re:internet filters

[germanStefan]

in the mean time try http://redirect.stefangeorg.net/

Re:internet filters

[Anonymous Coward]

It totally does bypass the Internet filters. I'm on a campus connection right now where they use agressive filter (Bible belt Christian school) and it works wonderfully.

In theory, they could respond one of two ways - block Googles servers (high probability) or inspect the packets, since ultimately your requests are included in plain-text.

Cool!

[exp(pi*sqrt(163))]

Not only will you be able to waste time more efficiently, it comes with a little clock that shows how much more efficient your time wasting was.

Re:Cool!

[Sox2]

ergo: you are obviously able to be deceived into thinking a deity created an entire universe

Re:Cool!

[exp(pi*sqrt(163))]

I disagree. In fact, there are very few things I think you can prove about a deity, but one of them is the opposite of what you say. In order to be able to deceive you it must exist. But if the deity exists, it can't deceive you into thinking it exists because believing the deity exists wouldn't be deceit, it would be correct. Therefore a deity cannot possibly deceive you into thinking it exists.

Unless you mean by "able to be deceived" you mean "deceived by a person not a deity" or simply "mistaken".

Re:Cool!

[Anonymous Coward]

i, for one, like the sig. :)

Squeezable Software

[Doc Ruby]

I always wondered why HTTP didn't allow a compression option in its C/S protocol. SSL encrypts HTTPS at the socket layer; why not just compression, without the scrambling? Especially for low-bandwidth mobile devices, compression of today's hundreds of KB of HTML (apart from multimedia objects) would really smooth the Web surfing.

Re:Squeezable Software

[eddy]

We already have compression of HTTP content, it's just that some idiots don't enable it on their servers.

The magic words are: apt-get install libapache-mod-gzip

Re:Squeezable Software

[SirSlud]

You can pretty much do it anyways, and most browsers support it, right?

Re:Squeezable Software

[TelJanin]

HTTP 1.1 allows gzip (bzip2 also?) compression.

Re:Squeezable Software

[SirSlud]

Incidentally, seperating code from view from data is probably the most effective way of cutting bandwidth. A report you look at 20 times a day, with different data, will download the file 20 times. The 'view' layer doesn't need to be downloaded again .. just the data!

This is why CSS is a good thing. You're not downloading the look & feel of the site every time you make a non-cached request. Getting the data out of their too would go a long way towards cutting down on the amount of useless bits browsers have to download over and over again.

Re:Squeezable Software

[Enahs]

apt-get install libapache-mod-gzip

which works beautifully on FreeBSD...

Re:Squeezable Software

[Anonymous Coward]

HTTP does have support for compressing responses. Firefox sends a "Accept-Encoding: gzip,deflate" header to the server to suggest that the server compress the content and the server will decide whether to respond with gzipped content and a "Content-Encoding: gzip" header.

Re:Squeezable Software

[Andrewkov]

Compression does exist as the others have mentioned, but still, most of the bulky files are already compesssed (jpeg, avi, etc). Compressing a few KB of HTML text won't make things *that* much faster (except for dialup users, maybe). Plus there is CPU overhead on the server and client.

Re:Squeezable Software

[Anonymous Coward]

I'm sure FreeBSD users are smart enough to know how to compile a module, should their package repository be sucky.

Re:Squeezable Software

[Anonymous Coward]

It's called Transfer-Encoding, and there's no less than three compression algorithms (gzip, compress, deflate) that are IANA-standard.

Compressing the headers is really beyond the scope of HTTP, which doesn't really aim at being a low-bandwidth protocol there. IPv6 has header and payload compression. There's probably an RFC to repurpose some of the weird bits in IPv4 for compression, but if so, it never really caught on, probably because the applications that needed it were already doing their own end-to-end compression.

Re:Squeezable Software

[hankwang]

Incidentally, seperating code from view from data is probably the most effective way of cutting bandwidth. A report you look at 20 times a day, with different data, will download the file 20 times. The 'view' layer doesn't need to be downloaded again .. just the data!

Unless you use MSIE, whose handling of the cache is terrible. If you have an URL www.example.com/style.css MSIE will cache it. However, if the URL is something like foo.example.com/style.css, it will not cache it, thus actually increasing the bandwidth. The same happens for .js, .html, .png, .jpg, .pdf and any other file extension.

To prevent this, the webmaster will find out only after quite a lot of web search efforts how to tweak the headers that Apache sends such that MSIE actually caches the data.

Re:Squeezable Software

[nsayer]

which works beautifully on FreeBSD

Fine.

cd /usr/ports/www/mod_gzip && make install

Re:Squeezable Software

[jrockway]

Yes; especially if you FreeBSD is OS X and you install Fink! apt-get _does_ work on non-Debian systems! Amazing!

Re:Squeezable Software

[SirSlud]

that and keep-alive over ssl and .. man, the list goes on with ie.

Anonymizer?

[alewar]

Do they provide also an anonymizer service with this accelerator/proxy??

Re:Anonymizer?

[Anonymous Coward]

Even if they did, it still greatly increases the amount of information Google has about your surfing.

Re:Anonymizer?

[Coppit]

Heck no. In fact, I'm sure they collect information about your browsing habits.

Re:Smart. Scary.

[Anonymous Coward]

They already know where you go. Adsense is everywhere. Remember DoubleClick?

Just what I need...

[CyberZCat]

...now when Google goes down I can't even use the web. What am I waiting for.

So once google owns the entire internet...

[Liquidrage]

what are they going to do with it?

Not that I'm anti-google. But it's amazing all the things they've gotten themselves into. Now they're apparently going to cache (pieces of) the internet for us.

Though this might finally be a usefull tool to get around the /. effect.

Re:So once google owns the entire internet...

[MrAnnoyanceToYou]

The thing is, they can't really 'own' the internet until this web accelerator can stop you from accessing certain sites. While that's possible, it's not something that really fits into their charter. They're not really skirting anything here. And my guess is that their top twenty searches will become MUCH more accurate in the future as well as faster due to this.

Re:So once google owns the entire internet...

[PReDiToR]

they can't really 'own' the internet until this web accelerator can stop you from accessing certain sites

What happens when a site changes their content to something GOOG (or their sponsors) don't like and they conveniently forget to update their cached version?

It would be a little like the MiniTruth*, wouldn't it?

I fear for the freedom of information in the digital age, bits and bytes are a lot easier than print to manipulate.

* 1984, George Orwell: The Ministry of Truth, the government department responsible for adjusting historical documents and books to conform to today's version of history.

Re:So once google owns the entire internet...

[ronsta]

bits and bytes are a lot easier than print to manipulate.

perhaps, but on the client side, bits and bytes are much more difficult to deny access to simply because they are easier to copy, and hence, more accessible, right?

i too can quote a novel from high school english class here. In Fahrenheit 451, they burnt hordes of books successfully, did they not?

Tracking?

[Anonymous Coward]

So now, not only will google have the "hooks" in place to know what I search for, but also everywhere I go? How do I know they don't have tracking software on the server side?

"Browser 334928394 spends alot of time looking at anti-depresent web sites,... "

Seems like a lot of information to give one company. How soon till they offer a service to save my CC #'s for me a la hailstorm?

Privacy issue

[TheKubrix]

2. How does Google Web Accelerator work?
Sending your page requests through Google machines dedicated to handling Google Web Accelerator traffic.

Don't get me wrong, I trust google, but even this makes me lift an eyebrow.

Re:Privacy issue

[Anonymous Coward]

If you trust Google, you are a moron. Now that they are public their sole goal is to maximize profits for the investors. Selling their soul should bring a good price.

no automated queries

[mcguyver]

From Google, You may not send automated queries of any sort to Google's system without express permission in advance from Google.

It's good to see Google abiding by their own rules.

People on WMW have pointed out that you can stop Google pre-fetching by banning IPs 72.14.192.0/20.

Re:no automated queries

[Mifflesticks]

But they are google, and I imagine they have permission from themselves... I don't see the contradiction.

Re:no automated queries

[Anonymous Coward]

Google says you can't do automatic queries to Google, but Google will be automatically fetching your web page whenever it wants. You can't do it to them, but they are doing it to you.

Re:no automated queries

[Anonymous Coward]

Google is a fucking indexer! How do you think they index? Magic fucking faeries typing in the webpages through remote sensing? Jesus!

So now google runs a proxy, every proxy will have to fetch the page that the user requests. Is it suddenly different because it's a google proxy?

Asshat.

Re:no automated queries

[Anonymous Coward]

I hear they have this thing called a "spider" that makes lots of automated fetches. Their prefetch respects robots.txt as well.

Re:no automated queries

[LiquidCoooled]

They GIVE you permission to use this software.

They let you download their software and explicitly say it will do this.

Re:no automated queries

[Anonymous Coward]

Paying for bandwith for spiders - ok. Paying for bandwith for automated tools that hit your servers - not ok.

Re:no automated queries

[Anonymous Coward]

Google proxy means *less* bandwidth, dipshit. It's a proxy. It's a proxy. (one more time) It's a proxy. The first user to request a page gets a hit on that site. *just like the user requested it* Then the rest of the people don't hit the site because it's already cached. Idiot.

Re:no automated queries

[Anonymous Coward]

Someone forgot to take their meds today.

Works in other browsers too

[Anonymous Coward]

7. Can I run Google Web Accelerator on a browser other than Internet Explorer or Firefox?

For other browsers running on Windows, you'll need to manually configure your proxy settings to 127.0.0.1:9100 for HTTP connections.

Still windows only... has anyone tested this with WINE?

What's in it for them?

[Anonymous Coward]

Great feature, even if it is for windows only, but I have to wonder why they even bothered to offer such a feature. Since it is via proxy they can see what pages are being browsed and add the information into their page ranking system.

Sure I won't, but I am still annoyed I can't!

[Anonymous Coward]

How far does Jon von Tetzchner have to frick'n swim before Google starts supporting Opera?

Re:Sure I won't, but I am still annoyed I can't!

[fbjon]

Well, gmail now works... at least.

Re:Sure I won't, but I am still annoyed I can't!

[Anonymous Coward]

Is there a valid reason why google doesn't support Opera?

Just curious. I love Opera, but absolutely hate the fact that Gmail/GoogleMaps (Satellite images in particular)- whatever else - does not want to work with Opera.

I understand that Opera is in the minority, but how are they so non-WC3 compliant that things suck to Google in particular?

Re:Sure I won't, but I am still annoyed I can't!

[Guspaz]

GWA works fine with Opera. If you notice on the info page, you can maually set Opera's proxy settings to use GWA, you just don't get the "x seconds saved" figure in the browser.

Re:Sure I won't, but I am still annoyed I can't!

[Yakman]

Google starts supporting Opera

Considering Google works fine in the other 95% of browsers out there I would say it's a case of when will Opera start supporting Google.

Re:Sure I won't, but I am still annoyed I can't!

[xQx]

7. Can I run Google Web Accelerator on a browser other than Internet Explorer or Firefox?

For other browsers running on Windows, you'll need to manually configure your proxy settings to 127.0.0.1:9100 for HTTP connections.

Re:Sure I won't, but I am still annoyed I can't!

[Anonymous Coward]

Opera's already fast enough. GWA simlpy makes IE and Firefox just as fast as plain Opera.

Re:Sure I won't, but I am still annoyed I can't!

[Guspaz]

I disagree. Opera's advantages seem to be in rendering speed, not transfer times. GWA accelerates transfer times, which would benefit Opera.

Re:Sure I won't, but I am still annoyed I can't!

[uhlume]

Yes, there is: because Opera's Javascript/DHTML support is abysmal.

Re:Sure I won't, but I am still annoyed I can't!

[Onan]

Considering Google works fine in the other 95% of browsers out there I would say it's a case of when will Opera start supporting Google.

Because we all know that marketshare is more important than standards.

Re:Sure I won't, but I am still annoyed I can't!

[masklinn]

It does, you just have to setup it manually and you don't have the bar.
Does Opera support extensive third party plugins anyway?

Re:Sure I won't, but I am still annoyed I can't!

[baadger]

Opera uses the netscape navigator plugin API, as does firefox. If you mean EXTENSIONS then no. Opera is very limited in open source or third party extensions..infact, I can't think of one.

Re:Sure I won't, but I am still annoyed I can't!

[uhlume]

Seriously, I'm not trolling: I've literally lost count of how many times I've written simple DHTML scripts that executed consistently in IE, Firefox, Netscape, Safari, even Omniweb -- only to discover that they caused Opera to absolutely shit itself executing the same code. I don't think it's exaggerating greatly to state that Opera is to DHTML roughly what Netscape 4 was to CSS: so bad that I've simply stopped even trying to support Opera with any but the most basic scripts.

Re:Sure I won't, but I am still annoyed I can't!

[baadger]

I'm using Opera 8 and it seems to support all of Google web services flawlessly.

Maybe it's time you upgraded

Re:Sure I won't, but I am still annoyed I can't!

[Anonymous Coward]

Google will start supporting Opera when Opera stops sucking.

Re:Sure I won't, but I am still annoyed I can't!

[kundor]

Something that works on 9 out of 10 browsers but not the 10th IS the standard.

Re:Sure I won't, but I am still annoyed I can't!

[Onan]

Something that fails 10% of the time is a pretty crappy standard.

Do you feel that it's a good plan to allow entities with overwhelming market share to always make their own proprietary design choices without any external accountability or documentation? Without any thought given to interoperability, or indeed specifically so as to preclude interoperability?

Microsoft and Cisco could probably choose to collaborate on their own IP-like protocol, and move very aggressively toward deprecating IP in all their existing and future products. They could also easily arrange for that protocol to be unusual and dynamic enough that no one else's implementations of it would ever work very reliably. (Or just use the DMCA to the same effect. The point of the question is not about the technical details of how they would do such a thing, but rather about whether they _should_.)

Do you feel that this would be the best thing for technology? For the market? For users? Even, in the long run, for Microsoft and Cisco?

and, okay... a technical question

[yagu]

Since I want it on linux.... could I have the proxy part of it on a Windows machine on my network, point to that port 9100 as a proxy for my firefox browser from my linux machine?

Re:and, okay... a technical question

[Anonymous Coward]

Nope, that doesn't work, I just tried it. The Google proxy rejects all connections that aren't from the local computer.

Also, the Google proxy is stopping me from posting on Slashdot...not good. ; )

Re:and, okay... a technical question

[s2r]

You could redirect for eg. port 12345 to port 9100 in the windows machine and use 12345 as the port to connect to the internet from behind the proxy pointing to the windows pc.

Since all gSoftware just listens to connections that come from 127.0.0.1 the above should work.

But theres a problem, I dont know any prog. on windows that redirects ports. :(

Hope it helps.

Re:and, okay... a technical question

[Kris_J]

Yes, but you'll need another proxy that accepts non-localhost connections. I recommend Proxomitron. Then set Proxomitron to accept connections from your network, and forward them to the webaccelerator.

This is a good idea on any home network with multiple PCs, Linux, Windows, Mac or whatever. This puts a single high-tech cache your side of your Internet connection talking to a single, massive, high-tech, compressing, optimising cache on the other side of your network. I think it's time for my home network to gain an always-on, fanless, Windows PC.

Re:Smart. Scary.

[dmf415]

I think they should be able to implement ads in the pages that their proxy cache's...

Ever notice the advertisements in gmail? For some reason i get these Brazillian Jiu-jitsu ads, how did they know? Scanning my email... hmmmm

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Smart. Scary.

[soupdevil]

What's next? Hopefully a calender. I'd love a free online replacement for Outlook.

No difference here

[grahamsz]

I tried loading some big pages and some pages off slowish sites and certainly didn't notice any difference.

It's already been uninstalled here. Right now i'm on a fairly slow dsl connection and expected to see a reasonable difference - but nothing remarkable.

Is this part of the Big Brother (TM) Google pack?

[WillAffleckUW]

Just wondering, because the privacy implications of prefetching your queries make my eyebrows rise ...

Only for windows

[a3217055]

This only works on windows, that is what the webpage says, but there must be one for others using firefox on other platforms

Re:Only for windows

[rjw57]

Not necessarily. It works by setting itself up as a webproxy on your local machine and then pointing the browser to that proxy. The browser is, in fact, pretty irrelevant.

Google FAQ's don't answer the questions

[killa62]

1. What does Google Web Accelerator mean for my site?

It means that you don't need to modify your website in order for your users to enjoy a faster experience.

Seems they're trying to hide something..

Oh, now this sounds juuust perfect

[LNO]

Google Web Accelerator uses various strategies to make your web pages load faster, including:

* Sending your page requests through Google machines dedicated to handling Google Web Accelerator traffic.
* Storing copies of frequently looked at pages to make them quickly accessible.
* Downloading only the updates if a web page has changed slightly since you last viewed it.
* Prefetching certain pages onto your computer in advance.
* Managing your Internet connection to reduce delays.
* Compressing data before sending it to your computer.

No, boss, honest! I have no idea how those pictures got onto my computer! I don't even LIKE Bob Saget goat porn!

Re:Smart. Scary.

[DogDude]

First off, it's called a "calendar". Secondly, there already is a kick-ass free online calendar.

Where is the Mac version for Firefox/Safari?

[r0d3nt]

Since Firefox for Mac is similar to the Windows version, where is this accelerator for Mac? Safari might be a different beast though...

Re:Smart. Scary.

[Scruffeh]

Yeah, I like the part where they say they wont give your information to thrid parties. No mention of what they might do with the data themselves though. I do like the idea of an accelerator for fast internet connections. Would definately get points from me for comedy value if I could get the conspiracy theories out of my head

No thanks!

[sanermind]

Google reserves the right to modify these Terms and Conditions from time to time in its sole discretion, without notice or liability to you. You agree to be bound by these Terms and Conditions, as modified.

Re:No thanks!

[Anonymous Coward]

I understand what you mean. Google was supposed to represent a new way of doing things (not evil they said) but they use the same underhanded techno-babble shit as everybody else. Common now, get real google.

Re:No thanks!

[WhatAmIDoingHere]

How is that techno-babble shit? Everyone has that in their EULA. From Blizzard to Apple. They don't have to tell you when they change the EULA. Blizzard actually does but they don't have to.

Re:No thanks!

[Koiu Lpoi]

Um... what licence agreement doesn't include a phrase very similar to that?

Re:No thanks!

[vhogemann]

AFAIK this can't be enforceable... at least not here at Brasil. When two parts have an agreement, and one want to change the terms, both have to agree and be noticed of the change.

Is the law in the US so flawed that this kind of term is actually valid, or Google lawyers just put there as a "safety measure", even knowing that's invalid?

Re:No thanks!

[autophile]

...?

Then you better not sign up for anything, anywhere, because all Terms and Conditions have that phrase.

--Rob

ha!

[roadrunnerro]

the accelerator is down - anybody got a mirror?

Oh the irony

[exp(pi*sqrt(163))]

One great thing about the Internet is how distributed it is. It was designed to be this way from the beginning so as to make it robust against (physical) attack. And it seemed pretty obvious, that by its very nature, the Internet had to be distributed. But now we find that the whole damn Internet actually fits inside one company and ultimately we're all going to be surfing just one web site - Google's. Amazing!

And scary too. There will come a point when we are dependent on Google and then the terrorists will know where to strike. As I mentioned, the distributedness was all about defense.

Re:Oh the irony

[LiquidCoooled]

Google is not a single whitebox server anymore.

Google is everywhere.

Re:Oh the irony

[exp(pi*sqrt(163))]

Pretty centralized compared to this.

They Want Usage Information. Good Business Step

[osewa77]

By offering the use their servers as a proxy they are able to get collect more accurate information about what sites you visit, outside the google network. If a search result results in visiting a site and you viewmore pages on that site it is probably more useful! By "speeding up" your experience they are giving us yet another reason to ignore privacy issues and trust them Overall, an interesting move by Google. They seem to be taking all the right steps now that they are public! (GMail, etc)

Google Ads show up faster!

[binaryspiral]

Wow, this caching technology has really improved the load times of google ads. Rawk on!

Re:Google Ads show up faster!

[Anonymous Coward]

You're joking, but soon their ads will be flashy!! So the caching will help them.

http://www.buffalonews.com/editorial/20050502/1067 216.asp

Soon, google will be an AD MACHINE!@ EULA

[Eric(b0mb)Dennis]

Google Web Accelerator sends requests for web pages, except for secure web pages (HTTPS), to Google, which logs these requests. Some web pages may embed personal information in these page requests.

Google receives and temporarily caches cookie data that your computer sends with webpage requests in order to improve performance.

In order to speed up delivery of content, Google Web Accelerator may retrieve webpage content that you did not request, and store it in your Google Web Accelerator cache.

To learn more, read our Google Web Accelerator Privacy Policy ().

Re:Soon, google will be an AD MACHINE!@ EULA

[SirSlud]

> In order to speed up delivery of content, Google Web Accelerator may retrieve webpage content that you did not request

Kinda scary if you're one or two links away from illegal content. Is this an issue, say, if you end up with bad material cached on your machine, or is it just inherently obvious that if you're concerned about these sorts of things, you dont use it?

Re:I keed! I keed!

[Anonymous Coward]

I had the best story ever and they rejected it.

It was my first attempt. Is there life after /. rejection?

Re:Smart. Scary.

[soupdevil]

Bleh. I use the Yahoo calendar at work. It's OK. But I want a GoOOogle calendar. Because I want to keep my contacts, search results, etc., in one place. And I really dislike Yahoo mail, at least the free version.

Re:Smart. Scary.

[airjrdn]

They don't have to. They just let the other parties search for it.

great idea

[Anonymous Coward]

anyone else see a trend in google's blatant moves to exploit our privacy?

Well my results are in

[MerlynEmrys67]

I have saved 0.0 seconds since installing the cache accelerator and surfing back to /.

I guess in the grand scheme of things - this whole time should count as negative

Re:Well my results are in

[Trejkaz]

Same here, still on 0.0 after a whole hour of ordinary usage. Seems pretty pointless, doesn't it? I wonder how many years it takes to get up to 7.5 minutes, like their screenshot shows.

Exactly.

[CarpetShark]

Google offering to proxy the web for everyone cannot make sense unless they're planning to make a lot of money from your personal browsing records. In all honesty, and without wanting to sound like a troll, I think "Don't be evil" just died.

Re:Exactly.

[Johnboi Waltune]

Google has spent years maintaining the highest ethical standards... I don't think they would piss away their credibility for profit, especially since they aren't hurting for cash in the first place.

I'm prepared to give them the benefit of the doubt. There are lots of cool things they could do with the information, used in aggregate. They could recommend websites to you by correlating your browsing history with others, kind of the same way Amazon.com recommends products. I for one think that would be cool.

Re:Exactly.

[Sancho]

Can you honestly not think of one single non-evil reason for offering a free web proxy? How about filling in the missing gaps? Those webpages that are not linked and thus generally unsearchable?

Re:Exactly.

[bluelip]

They can insert ads into the pages. increases their revenue. Appears reasonable enough for me.

Re:Exactly.

[enosys]

Their ads into others' pages, like Gator^H^H^H^H^HClaria? That would definitely mean "Don't be evil" just died.

Re:Exactly.

[erebor]

Look for google to begin inserting ads into proxied pages. They probably don't care about your personal browsing history.

Re:Exactly.

[natrius]

Let's look at the information Google currently has to generate search results from.

1. The content of each webpage (text, images, video, anything really)
2. The number of pages that link to a page in question
3. The words that people use to link to a page
4. The sites that people click on after searching for a term

These by themselves generate pretty good results, but sometimes this information can be deceptive. The more metrics you have to measure relevance by, the better.

So now, Google offers to cache the Internet from everyone. What can they get out of this? Well, everyone here is speculating about the evil things, so I'll leave those as a given. What I haven't seen so far is a very valuable piece of information they get from this: web traffic. They get to see how many people go to web sites, what time, where they got referred from, and anything else that can be deciphered from someone's web traffic. Not only can they rank pages by how many people link to a page, they get to see how often each link is actually used to get to the page. That's extremely valuable, because it's hard to fake convincingly. Web sites won't be able to plant links around the Internet to increase their ranking, because if no one actually clicks the link, then it's not important in the first place. That is awesome.

Why didn't I think of that?

Re:Exactly.

[EnronHaliburton2004]

money from your personal browsing records.

They want to know what everyone is searching for in a given moment, and model their advertising business around that information. This is the purpose behind Gmail and Google Groups.

This is their business model. They are an Ad business first, and a search engine second.

They will gain information from your personal browsing records. Their advertising business can use this information with direct-market advertisements, future trend prediction, etc.

Re:Exactly.

[Jugalator]

Google offering to proxy the web for everyone cannot make sense unless they're planning to make a lot of money from your personal browsing records.

Hmm, money? Yes, in the end of course they need to profit from it. Google is not charity organization, and have a ton of expenses. However,money how? is a more interesting question.

I can't believe Google will simply sell the results to some third party -- that would look pretty bad PR-wise, and Google has so far tried to avoid these things as well as possible. Something more commonly seen with Google is beating the competition by providing good and accurate search services. If they do that, they gain a larger market share since they're simply better, and that will make companies willing to pay more for AdWords. Tadaa, Google in a nutshell, and how they've always worked.

So I basically think it may have something to do with this. What better foundation for a TrustRank system can you get, than one where you know how visited sites are? Scam sites would only get sporadic visitors from fooled Internet users and have their PageRank drop like a rock, while news sites, popular gaming sites, and so on, would get large numbers of returning users. Cross-linking scam sites would find out that their exploits wouldn't work very well anymore, and Google could possible tune their rank system to let both PageRank and TrustRank have an influence on the final rank. Sounds like the regular Google philosophy of conquering by improving. And they'd need our browsing habits to pull it off.

Re:Exactly.

[Jugalator]

Nah, that would get them into a lawsuit crossfire and would be PR suicide by using their marketshare to advertise on a lot of non-Google sites.

Re:Exactly.

[shadowsurfr1]

Epic 2014 is starting to happen with them aggregating information about browsing. It's almost scary. Check the website linked for a view into the future of what Google (and Amazon) may be doing do next. It's all speculation but sounds very interesting.

Re:Exactly.

[ottothecow]

Until they start trying to combine browsing habits into one.

"Based on your recent browsing habits, Google would like to suggest MidgetsHavingSexWithFerretsInSpace.com"

All I wanted was a smaller computer, a pet toy, and some homework help....GOOOOOOOGLEEEEEEEEEe

Re:Exactly.

[vertinox]

Scam sites would only get sporadic visitors from fooled Internet users and have their PageRank drop like a rock, while news sites, popular gaming sites, and so on, would get large numbers of returning users.

Great! Now, I can actually use google to find Mame ROMs again!

Re:Exactly.

[Zak3056]

Google has spent years maintaining the highest ethical standards...

Such as censoring their search results at the behest of the chinese government? Licking the boot of a dictatorship just to avoid getting kicked out of the market is not "maintaining the highest ethical standards."

Re:Exactly.

[Anonymous Coward]

Why didn't I think of that?

Because someone else already did. This is exactly what the google toolbar is designed to do (or the alexa or yahoo). Except the tool bar doesn't require them having asmuch bandwidth as the rest of the interweb combined.

Re:Exactly.

[Ninwa]

Very interesting video, well edited, and definitely thought provoking. Thanks for making me aware of it. :]

Re:Exactly.

[shadowsurfr1]

No problem. I added it to my bookmarks as soon as I found it. I knew it would come in handy someday!

Re:Exactly.

[natrius]

The Google toolbar only sends the URL of the page you're visiting back to Google, with a harmless google.com URL as the referrer. People would scream about privacy complaints otherwise. With this, the only information sent to Google is stuff that's necessary for the service to work. Fortunately for Google, that information happens to be important.

Yahoo and presumably Alexa do send referrer information if WebRank is enabled. I don't know what percentage of people turn off WebRank, but with Web Accelerator, there's no opting out if you use it.

Re:Exactly.

[Koiu Lpoi]

Because everyone knows that there is no grey area. Something is either good or evil. You're either with us or against us.

Re:Exactly.

[ThePromenader]

I agree with you on the Trustrank possibilities of this, but don't forget that only a certain "chunk" of the population will be using this feature at first - those less worried about privacy issues, or the "home user" department - so I would tend to doubt that any info gathered there would be a "real" reflection of web use. At least in its first years.

Just a personal thought, but if they ever would care to publish their proxy statistics I would be VERY interested - here's an occasion to see first-hand statistics of what people are really looking at. Hell, they could even start a sort of virtual "top of the pops" page...

Re:Exactly.

[Urusai]

Well, the IPO kind of tipped me off. "Because it profits the shareholders" is the universal corporate excuse for all manner of chicanery.

Re:Exactly.

[Anonymous Coward]

Google has spent years maintaining the highest ethical standards... I don't think they would piss away their credibility for profit

This is the normal way some people do business. I can't say if Google is one of these people or not, but you can see it clearest on eBay. The most successful crooks play fair for a long time and then go crooked and flee with huge sums of money.

To capitalism, people's trust has value. The next step is to figure out how to make as much money for your shareholders out of that trust. Go free market!

BTW, Google admits they will log every URL.

Re:Exactly.

[Rutulian]

But...how good of a sample will it be?

How many people will install the web accelerator? How many computers will come with if preinstalled?

If the only people using the service are Joe Blow on a dial-up connection, then the sites they visit will be better represented. Say they check Fox News everyday. Well, a typical slashdotter might prefer any of the Google News sites (NY Times, BBC, CNN, etc...), but Fox News is going to get the PageRank. I don't know, it seems like they would have to weight the results carefully to keep an even distribution. Besides, you don't really want "most popular site takes all." I mean, say Fox News really was hit the most often. Should it really have a better PageRank than NYTimes just because it is more popular? It is the unpopular sites that need a search engine. Everybody already knows about the big ones.

Re:Exactly.

[maddugan]

They could easily use the data to augment there searching so pages people actually visit get higher rankings.

Re:Exactly.

[Anonymous Coward]

Good God! They use cookies?! I'm never going back to that site again.

Re:Exactly.

[vslashg]

Man, I was so disappointed when that link didn't work.

Re:Exactly.

[yppiz]

I am guessing that Google's bandwidth use is mostly from crawling. If this is correct, then they may have a fair amount of 'free' upload bandwidth that they aren't using. This application would use that bandwidth.

In return, they get real-time data on sites, indicating how relevant links are and also how frequently they should recrawl certain pages.

--Pat / zippy@cs.brandeis.edu

Re:Exactly.

[pigwin32]

This issue isn't necessarily just about Google staff accessing your personal data, it is also about exposing your personal data via Google to villains. This is the same problem with the likes of Marketscore who also proxy web pages. Recently most banks in NZ denied access to their respective internet banking sites from all known Marketscore addresses because Marketscore was proxying SSL connections. I don't see how this new Google "service" is any different. Avoid it.

Re:Exactly.

[NetSettler]

Google has spent years maintaining the highest ethical standards... I don't think they would piss away their credibility for profit, especially since they aren't hurting for cash in the first place.

Of course, it's not always "now" that these problems occur. One reason that one maintains strict ethical breaks between various organizations is not to protect them when they're strong, but on the assumption that one is not always strong every day.

I heard a few years back that Reader's Digest was not doing economically well and that their biggest asset turned out to be a repository of the reading habits of a huge part of the US population. Even if they were not inclined to sell out, they were still candidate for takeover by another company buying them just for this data and not for their editorial work or revenue stream. I didn't end up following the news, so I don't know how it turned out, or even that this account I'd heard was correct. (Maybe someone else knows better can offer more info here.) But even if you take it only as a hypothetical, it seems pretty plausible that such things could happen.

Big companies have sometimes fallen. And one would like to believe we haven't entered a political climate where that will never happen again, even if one doesn't have a deathwish for any particular big company. So what if Google gets all this stuff and then gets either nervous or outright cheap... If their size and economic power is what protects us now, what protects us then?

Re:Exactly.

[fm6]

If you judge somebody by the amount of evil they could do, Google went beyond the pale some time ago, since they already have a lot of capability to spy on people. But it's usual to judge people the amount of evil they actually do. If Google implements a reasonable privacy policy, and sticks to it, they're off the hook.

Re:Exactly.

[Andrewkov]

They already have a striped down cache of the whole internet .. It's not that far a stretch to turn that cache into a caching proxy.

Re:Exactly.

[Mad_Rain]

Google has spent years maintaining the highest ethical standards...

Which makes me wonder:

...My company's firewall filters some objectionable content

...My company's firewall does not filter Google

...would I (or others) be able to surf for objectionable content through Google and bypass the company firewall this way?

Re:Exactly.

[MobileC]

Licking the boot of a dictatorship just to avoid getting kicked out of the market is not "maintaining the highest ethical standards

Or the US government?

Re:Exactly.

[Anonymous Coward]

"I don't think they would piss away their credibility for profit"

Hahahahahahahahahahahahahahahahahahahahahahahaha ha hahahahahahahahahahahahahahahahahahahahahahahahaha hahahahahahahahahaha

The naivety of slashdotters really is too funny.

Re:Exactly.

[rbarreira]

Ethical standards aren't above the law. Not for corporations' wide-known measures at least...

Re:Exactly.

[jericho4.0]

Cookie paranoia is soooooo 20th century.

Re:Exactly.

[Petronius]

They want to know what everyone is searching for in a given moment

that's the easy part:

- lesbian sex
- natalie portman
- desperate housewives
- desparate housewives having lesbian sex with natalie portman

Re:Exactly.

[d474]

+5 Funny pleeez - Excellent punch line man! er, or woman.

Re:Exactly.

[SacredNaCl]

Banning advertising of those they politically disagree with (such as gun, ammo, gun rights ...etc).

Re:Exactly.

[ColMustard]

The sites that people click on after searching for a term...

Actually, I don't think this is one. The links an result pages look like they link directly to the site instead of going through some redirection script on Google's own server. The 'Similar pages' feature could be used for this purpose, though.

Re:Exactly.

[Jah-Wren Ryel]

Until they start trying to combine browsing habits into one.

"Based on your recent browsing habits, Google would like to suggest MidgetsHavingSexWithFerretsInSpace.com"

Plus, we can look forward to a bunch of "Google thinks I'm gay!" jokes since the tivo ones have grown passe.

Re:Exactly.

[natrius]

It's a proxy. All your traffic at least goes through Google's program, and most of it will be Google's cached copy. When you request the copy, they get all the information they need.

Re:Exactly.

[Onan]

Only people above a surprisingly high threshold of technical savvy would even think of installing a tool of this type. That same set of people would find the addition of ads to be a vastly bigger deal than any modest speed gains. Approximately zero people would ever use such a tool.

So without even needing to raise the question of whether Google is good or ethical, we can just look to whether they're the very small amount of smart necessary to see that such a ploy would not work. Which, yes, I personally am willing to grant them.

Re:Exactly.

[__aailob1448]

Because who would have imagined that google would have enough bandwidth to act as a proxy for everybody ? I still have a hard time believing it...

Re:Exactly.

[mikkom]

Recently most banks in NZ denied access to their respective internet banking sites from all known Marketscore addresses because Marketscore was proxying SSL connections. I don't see how this new Google "service" is any different.

It's different - Google doesn't proxy https:/// connections :-)

Anyway, I don't quite get these "web accelerator" services anyway, if some site is too slow, you just should stay away from it.

Re:Exactly.

[Corrado]

Yes, you can blow right past your company's firewall/proxy! I usually cannot get to Azureus, but with Google Web Accelerator I can download to my hearts content.

Hmmmm...I wonder how long it will take before my company recognizes that I am no longer opening connections to multiple sites... :)

Re:Exactly.

[volkris]

And seriously, what's wrong with making a lot of money from your personal browsing records?

You talk like you're somehow hurt by this.

If they can make money off of me without harming me in the slightest, and in fact while giving me something I want, then by all means!

Re:Exactly.

[andywebz]

Which is why you should create 2 firefox profiles. One that uses the proxy and one that doesn't. Or if you can't do that with this, then use mozilla for regular browsing, and the google accelerated firefox for searching all your naughty stuff.

Re:Exactly.

[Anonymous Coward]

Is there any way to point directly to the Google proxy and force all traffic to go through that for a particular profile?

Re:Exactly.

[gmplague]

Or you just turn of GWA when you're browsing normally, and turn it on when browsing to naughty sites.

Re:Exactly.

[Relgar]

If your company runs an internal DNS server/relay, won't they be able to log the sites you're going to? They may not be able to see the content, but they might wonder why you're trying to hit a banned site so often.

Re:Exactly.

[Anonymous Coward]

Following your logic, google should raise the ranking of sites that no one visits because they need the exposure. If more people are going to a site it is probably because that site has the information that they are looking for. In your world, google should link first to sites that no one wants to use. Those people should just stay with MSN search. I prefer google giving the site most likely to give me what I am looking for.

Should a site that is more popular should have a higher page rank? MOST DEF!!

Re:Exactly.

[Anonymous Coward]

I think "Don't be evil" just died.

If I were going to be evil, my motto would totally be "Don't be evil"

Re:Exactly.

[tekunokurato]

Ethical standards aren't above the law

That very much depends on one's political philosophies. I'd hazard that Nazi soldiers and German citizens who felt that their ethical standards were above the law were correct, while those who did not were probably incorrect.

Re:Exactly.

[mikehilly]

The best way to blow around the firewall for content depends on what you are looking for. If it is data on a site that is blocked then search for the site and bring up the google cache of the site. Then when you click on a link on that page and comes up blocked, pop that web address back into google and open the google cache of that link address. Kind of cumbersome, but it gets around. For pictures, just go to googles image search and search there. While you won't get the full sized original, you will get thumbnail sized pics... Don't get in trouble through because usually someone will eventually hit the logs for content and see what you are looking at.

Re:Exactly.

[Anonymous Coward]

Grandparent was refering to just visiting google.com, not using GWA.

Re:Exactly.

[lloydtesterman]

midgetshavingsexwithferretsinspace.com is available.

AVAILABLE EXTENSIONS
The domain name you searched for is available with the following extensions:
midgetshavingsexwithferretsinspace.ne t
midgetshavingsexwithferretsinspace.org
midgets havingsexwithferretsinspace.info
midgetshavingsex withferretsinspace.biz
midgetshavingsexwithferret sinspace.tv

SIMILAR AVAILABLE DOMAIN NAMES
The following names are available:
mymidgetshavingsexwithferretsinspace.c om
midgetshavingsexwithferretsinspacecentral.com
themidgetshavingsexwithferretsinspace.com
midget shavingsexwithferretsinspaceweb.com
intermidgetsh avingsexwithferretsinspace.com

I sort of feel like I should have Posted Anonymously......

Re:Exactly.

[rbarreira]

Yeah, but my point was that the fact that they were correct wouldn't help them much during a trial...

By the way, are you invoking Godwin's law? :D

Re:Exactly.

[tekunokurato]

Not expressly, as it wasn't so much a comparison as it was an acid test, but... ah, who am I kidding, I lose...

Re:Exactly.

[pigwin32]

It's only one letter different. Marketscore too started out only proxying non-secure connections. Personally I think it sucks to give all your traffic to a marketing firm. There's not a whole lot of privacy on the web anyway but this is whittling away a little more of it. Plus it is even more suckful to advertise it as a means of speeding up your browsing. This is the same bullshit story Marketscore used.

Re:Exactly.

[Anonymous Coward]

I've been using Google's cache to bypass a company proxy for years. All you have to do is...

http://216.239.39.104/search?q=cache:http://www.na ughtyurl.com

If you're sneakier, you'll create an IE search alias so you can type "cache www.naughtyurl.com" in your address bar.

Of course you can't click on links, you'll have to copy then and do the process all over again.

Re:Exactly.

[Sir Pallas]

I think they're probably just trying to use a better ranking system, namely the one where users actually click on things. It's one thing to know which pages link to which, but if you know which links are actually used you can better optimize a search engine. I talked to Rob Pike about this after a talk he gave for my graduate group and this is what he seemed to indicate. (Not that he said anything explicitly.) The goal of Google, of course, is to make the world's information readily accessable; "don't be evil" are just guide lines. This is the same sort of FUD that popped up when they started Gmail.

Re:Exactly.

[CarpetShark]

Well, to be fair, many people would say that those initial concerns about gmail have turned out to be true. I'm partly basing my concerns about this on their previous track record with gmail. But yes, I see that there are other possible driving factors here, so I'll wait and see.

Re:Exactly.

[AJC123]

Have you never heard of Comscore? They have (until Google beats them) the same offer -- web acceleration in exchange for total usage visibility. They can predict all sorts of stuff based on analyzing their data. Their internal statistics accurately beat national labor results by about three days. They can predict elections and American Idol results. This is the next business for Google. Feeding leading interest indicators to the companies, investors, and politicians that can afford it. It's not you that Google is interested in. It's a statistically valid representation of you.

Re:Exactly.

[jasonjacks0n]

- desparate housewives having lesbian sex with natalie portman

Link?

;-)

Re:Exactly.

[Anonymous Coward]

Google is allredy doing this ( the colecting of browser histroy with the google histoy service

https://www.google.com/searchhistory/login

So i see GWA as an exsention of this with added advantages of 'faster' browsing.

Re:Exactly.

[william.gunn]

That would be great, because Amazon does such a good job with recommendations currently.

If website recommendations were to work like Amazon's, I'd get recommendations to visit www.slashdot.org, while I'm reading a thread.

Re:Exactly.

[william.gunn]

You mean like this?

Re:Exactly.

[ThePromenader]

Um... yeah. Exactly that. Thanks for the heads-up : )

Re:Exactly.

[mdfst13]

"they get to see how often each link is actually used to get to the page. That's extremely valuable, because it's hard to fake convincingly."

Err, no it isn't. It's dead easy to fake. Just send requests through the Google Web Accelerator. The page rank links are harder to fake, since you can cancel the benefits of certain sites manually (i.e. Google has the capability to ban people from their index). It's hard to fake a link from nbc.com. It is easy to fake clicking on a link on fake-site.com.

The best response that I could see to that would be to only use long established accounts for the page rank. Of course, that just makes virus zombies valuable again. If you corrupt a PC, you can spam their GWA proxy with your links. As a result, they will probably only use a small amount of this info in Page Rank.

Let me guess...

[Anonymous Coward]

...it's "4x Faster than DSL**!!"

Re:Smart. Scary.

[hansonc]

there already is a kick-ass free online calendar.

who cares if Yahoo! has one? It doesn't interact with my gmail account so it's useless (to me)

Restrictions

[Bifurcati]

Note that at the moment it's limited to users in North America and Europe, which limits us poor Aussies...

Also note that despite being available for Firefox (yay!) it's only available on Windows (booo...) - no support for Macs or Linux. That would have to be my only (minor) gripe with Google - they don't develop any applications for the minority. Of course, I understand why, but these are people that tend to be quite loyal, and since no-one else is really developing these things...

Re:Restrictions

[Bifurcati]

That said, a closer examination of their FAQ reveals that as long as you configure your proxy right, any Windows browser will work.

Does anyone know if this works if you already have to go through a proxy server, as I do at my University?

SLOWER!

[SirDrinksAlot]

Theres not much you can do to speed up 4MB DSL. This is really hurting my browsing performance.

I'd also like to know how it figures it save you time. Consitering i've "saved" half a second going to several websites now that load 5x faster with out it.

Re:SLOWER!

[WillAffleckUW]

Theres not much you can do to speed up 4MB DSL. This is really hurting my browsing performance.

Good point. Here at the UW we've got gigabit internet, so my guess is it would crawl if we used that, and probably similar results for people using Seattle's Speakeasy gigabit WiFi network.

Wake me when it works for Linux or the Mac.

Re:Smart. Scary.

[shakezula]

Yes, Google's scanning your email, but they aren't reading it...

1. Is Google reading my email?
No. Google scans the text of Gmail messages in order to filter spam and detect viruses, just as all major webmail services do. Google also uses this scanning technology to deliver targeted text ads and other related information. This is completely automated and involves no humans.

Re:I keed! I keed!

[mtz206]

Hell, I submitted this same story 3 hours earlier and it got rejected. Must be my breath or something...

Calm Down

[Anonymous Coward]

This probably won't even be visible since I've lost the user/pass for the umpteenth time, but here goes:

There's no reason to panic. Google is not forcing you to use their proxy. If it is a worthwhile tradeoff for people, then they'll use it.

For example, my parents live out in a small town with only fixed wireless broadband. They (and most of the people in town) are sticking with dial-up. Quite a few dialup companies are trying to capture that market using their own proxy/caching servers as a hook. "Surf up to 5x to 10x faster with CrapNet"

If google want's to make that kind of ability freely available, I won't have any problem suggesting it to my parents (the "speed boosting software" that the local dial-ups use costs an extra $5 a month)

Sorry for rambling, I've got my last final in 2.5hrs so I'm kinda frazzled...

Re:Once again Google forgets us.

[TelJanin]

RTFA. It works with Opera also, you just have to tweak your browser a bit.

Re:Smart. Scary.

[saforrest]

First, they collect your search information. Next they collected your email. Now they collect your destination. You put it all together, that is quite a bit of information.

Add to that your Usenet posts, where you're going or where you live, what you're buying, what kind of news you're interested in, and maybe even who your friends are.

But all that's only true if you give them the information. Even so, the quantity that Google could know about me just given all the Google stuff I've used from one single IP address is rather alarming.

But I don't mind. This is partly because I don't think they're jerks (as far as public corporations go, anyway), but mostly still because I don't think they really care.

If we had a lot of evidence they did care, then I suspect that there would immediately exist a movement for 'free', anonymous versions of whatever services Google currently provides.

Its an odd thing...

[riprjak]

Ive blogged on the disturbing accuracy of the targetted adds in gmail (like realising we are writing haiku where we havent actually mentioned the word "haiku"), I am fully cognicent of their MASSIVE database and penetration of all things search. They know more about me than I would normally be comfortable for people to know as I conduct ALL my searches there. Now this... a product which I feel drawn to using, in much the same way I reflexively watch star wars movies no matter how bad they are...

And a small voice at the back of my mind muses that perhaps this is how the German people felt as they voted back in 1933. It SEEMS like a good idea but you cant help but wonder where it is all going to lead.

For now I just take solace in the fact that Microsoft's ridiculously large staff probably sit around in horror, wondering how the fuck Google keeps doing this :)

err!
jak.

prefetch denial of service

[clinko]

I'm giving myself a self inflicted prefetch denial of service attack by including:

Then including my url which you should see double underlined here: Clinko!

Why?

[Sairret]

Is there really any demand for a web accelerator anymore? Dialup is slowly dying, and I don't really feel the need to speed up my DSL connection.

I'm sorry, Google

[Eric(b0mb)Dennis]

Dear Google,

I've always loved you, sometimes a bit too much. When all the others told me you were bad, I didn't listen. I followed you blindly. Google, everything you have done until now has been the upmost joy of my life.

This "Web Accelerator" is a abomination! It's akin to Paul Atreidies forgoing his addiction to the spice melange.

A "Web Accelerator" designed for broadband connections, what kind of crack are your PhDs smoking? Maps, Suggest, Gmail, Web Accelerator (ACCELERATED FOR BROADBAND)

Wtf AOL?!?

Re:I'm sorry, Google

[lachlan76]

A "Web Accelerator" designed for broadband connections, what kind of crack are your PhDs smoking?

It doesn't just accellerate your connection, it probably uses a cached page so that you can negate things like a Slashdotting, Googleblatting, etc.

Sounds interesting....

[FS1]

I think that a word of caution is needed here. Now Google, in it's current state, seems to defy the "laws of business." I for one hope that it remains an honest company that continues producing software that is innovative and desired. People trust Google way more than any other company in recent memory. Google has access now, through their software, to every file, search, website you visit, password, personal detail, and photo you have (assuming you use all their software).

Am I the only one a little shocked at this? What's to stop another company from swooping in and buying Google with all your assorted information? Or, to stop Google itself from using this information in a way that most people wouldn't want them to?

Obligatory Murphy's Law derivative quote: "If everything seems to be going well, you have obviously overlooked something."

Re:Sounds interesting....

[citking]

You're absolutely right...what is there to keep Google seperate from other businesses who will collect information and later sell out to the highest paying bidder, information and all? The truth is, nothing. But I still trust Google because they haven't lost my trust yet. I usually put my trust in an institution that has a reputable track record for success and customer service. How do I know my credit union will be there tomorrow, other than the FDIC insurance? How will I know if I am actually going to get my 401(k)? How will I know that my car will start tomorrow? I know nothing of the future, but I know from track records that my credit union will be there tomorrow because they have a wise and active Board of Directors. I am a little less certain that my 401(k) won't be nabbed from me by a renegade governor (I work for the state, so I am a bit more secure than others may be about it) but that is still unlikely. My car, a 94 Cavalier Z24 which has "Service Station" listed under my insurance's primary residence, could quite possibly not start tomorrow. I'm not gonna worry about it all though because I have better things to do

Google though is a firm institution in many people's eyes. They support F/OSS in so much as I've seen (even if just by keeping certain sections Beta forever and granting the Mozilla FireFox default start page), they play fair, they push out what people want, and they give, give, give. Picasa 2, GMail, Image Search, Satellite maps and directions, and, of course, the most comprehensive search box on the net. Their ads are unobtrusive yet effective, their founders are ones we can relate to, and most importantly they have a sense of humor! Personally I'd be very upset, like many of you, if Google were to ever betray that trust. But since they haven't yet, I'm not going to hold a grudge against them "just because they might".

A legitimate Internet Accelerator

[_RidG_]

Good God, I never thought I'd live to see the day a legitimate Internet Accelerator was released. After all, these are precisely the programs that I see advertised via pop-ups on spyware- and malware-infested PCs: "BOOST YOUR DOWNLOAD SPEED 100%!!!! DOWNLOAD OUR INTERENT ACCELERATORX NOW FREE AND WIN A FREE IPOD AFTER PUNCHING THE PINK MONKEY BOXING PRESIDENT BUSH!!!!"

What an age we live in!

Re:A legitimate Internet Accelerator

[Kris_J]

True Internet accelleration is easily achievable with a browser plug-in/server pair. Other than that, your best bet is a good ad filter. What's more, they stack.

Re:Smart. Scary.

[Skal Tura]

Yeah, and marketing firms would just LOVE and pay ANYTHING to get their hands on all of that information.

It's a little bit specific, but oh wow, would they learn TONS of new tricks to sell us stuff we don't really need...

/me smacks himself in the head remembering that he bought last night to his car a brand new spanking DVD player car stereos & TFT screen...

Web stats for hosts will not be accurate

[texas neuron]

Since they will host frequently used pages on their web site, it is not clear to me that my web stats will be accurate. I can specify if pages are prefetched but if they have chosed to cache a page and then the user clicks on another cached page, it is not clear to me that my web stats will know this.

do no evil!

[SuperBanana]

Very Smart..Very Scary...

"Don't worry. Their motto is 'do no evil', so we can trust them!", say the geek masses.

Dow Chemical's motto is "Living. Improved Daily". Unless you're one of 15,000-30,000 people in Bhopal, India, of course.

Ford's motto is "Ford: Quality is #1". Well, except for the Ford Pinto (or its modern equivalent, the Ford Crown Victoria, which is burning police to death left+right). Or Ford Explorers, where management ignored engineering reports saying the roof pillars were substantially weaker. Or ignition switches in millions of Ford vehicles which would catch fire- even if you weren't using the car? Then there's the Ford Focus, which I think is close to setting the world record on factory recalls...

Then there's GE- "we bring good things to life". Well, I don't think the people who have been harmed by dioxin poisoning would agree with you there. But hey, GE will sell you a nice water filtration system (seriously- go into Home Depot, GE is the featured brand. Note how it brags about removing industrial toxins?)

Microsoft says "enabling people and businesses to realize their full potential", something I think we can all give a good chortle about, considering how grossly unreliable virtually every Windows release has been, how incompatible their software is one year to the next, piss-poor interoperability, anticompetitive practices, licensing costs, spyware, viruses, etc.

Need I go on to prove that corporate PR lines are just that- nothing more than PR lines? Or should I mention that Google AdSense terms prohibited AdSense customers from discussing, in public or private, their experience/satisfaction with AdSense? Hmm. Now, why would a "do no evil" corporation do something like that?

Re:do no evil!

[oldwolf13]

I wish I had mod points for you... you spoke my mind for me.

I think if Google actually wanted to adhere to "Do No Evil", they wouldn't be gone public with their IPO.

Public companies do whatever they can to maximize profits, I've even read (although I believe this was on /.) that they HAVE to do this.

There is also a long ways between "Do No Evil", and "Do Good Things".

On a side note, I use both gmail and google, I remember the pre-google days when searching was just painful. I hope that google will not start to do things the way most of the corporate world does.

Re:do no evil!

[Anonymous Coward]

Dow Chemical's motto is "Living. Improved Daily". Unless you're one of 15,000-30,000 people in Bhopal, India [wikipedia.org], of course.

That was Union Carbide. And besides it was the local inept UC people too coupled with the India government.

Re:do no evil!

[Anonymous Coward]

Why do you hate America?

Re:do no evil!

[joeljkp]

Public companies are required to maximize the value of the company, not profits. Doing good things while making a decent profit is certainly more valuable than making a massive profit off of lies and extortion (see Enron).

Re:do no evil!

[oldwolf13]

While I'll take your word on the subject of value versus profit, companies use profits to increase the companies value.

One of the http://www.theregister.co.uk/2005/05/04/ibm_fire_1 3000/*Slashdot goodguys* is reportedly getting rid of THOUSANDS of employees BECAUSE they didn't have a good enough quarter. While some might consider this kind of behaviour NOT EVIL, it certainly isn't good.

Now I don't think IBM had to do this to save their business, and maybe it's a long term strategy that I don't follow, what I do see is over ten thousand people losing their jobs in order to "maximize" the companies "value"

By the way, the original poster's URL in his sig is great.

Re:do no evil!

[oldwolf13]

damnit. I meant the url in SuperBanana (662181)'s sig.

"The Apple Product Cycle":
http://slashdot.org/comments.pl?sid=148392&cid=124 36782

Apparently also, I suck with posting URL's.

Re:do no evil!

[iammaxus]

-1 Troll, for sure. The amount of good that all the companies you have mentioned have done (except for possibly Microsoft) is immesurable. Sure there have been many setbacks along the way, but these companies have been innovating and improving our lives for literally more than a century each. As for Google, I can't sem to find any refereence to what you described in AdSense's terms and conditions, have a look for yourself https://www.google.com/adsense/localized-terms/

Re:do no evil!

[natrius]

"Don't worry. Their motto is 'do no evil', so we can trust them!", say the geek masses.

The reason I personally trust Google is because they've earned it. Their business ethics are far better than most companies out there, so I trust them with the messages people send me, the locations I drive to, my writing, and my interests in general.

I'll wait to see how they use the data they get from this before I consider using it (not that I can now anyway, since it's not available on Linux), but I'm pretty sure what they're going to use it for anyway.

Re:do no evil!

[Anonymous Coward]

So, what's your favorite web search? :-)

Some people just bitch but don't give any constructive thoughts on the matter...

Re:do no evil!

[rfunches]

Or should I mention that Google AdSense terms prohibited AdSense customers from discussing, in public or private, their experience/satisfaction with AdSense?

As an AdSense customer I never understood the purpose of this in the terms of service.

...what, you were expecting me to say something else? Because I'm going to violate the ToS of Everyone's Favourite Company...

Re:do no evil!

[Anonymous Coward]

You're british, you bastard! And I know it, because you spelled "favorite" wrong. Limey pricks.

Re:do no evil!

[Dun Malg]

Dow Chemical's motto is "Living. Improved Daily". Unless you're one of 15,000-30,000 people in Bhopal, India, of course.

Nice troll. Inflamatory, and correct only by a tenuous strand of tortured logic. It was Union Carbide who gassed Bhopal, which didn't merge with Dow until 1999, a full fifteen years after the incident, and five years after Union Carbide sold its 51% interest in the Bhopal facility.

Re:do no evil!

[coolsva]

Dow Chemical's motto is "Living. Improved Daily". Unless you're one of 15,000-30,000 people in Bhopal, India, of course
FYI, it was Union Carbide who owned the Bhopal plant. After the accident, UC (smartly) sold themselves to Dow, who took over the assets and part liabilities (the multi million dollar settlement to the Indian Govt.).

Re:do no evil!

[Anonymous Coward]

Why do you hate America?

Because you only constitute 4.5% of the world's population but you consume 25% of its resources.

No one hates America, not even terrorists, despite what you're being told. We feel sorry for you. You're a wonderful, friendly, open and honest people who also seem to be hideously naive. You're being exploited by a corrupt government and you're losing a priceless constitution that other populations in the world would sacrifice their lives for (as did your forefathers, remember?). We just wish you'd be a tiny bit more open minded - take a few more holidays abroad and see how others live. Consume a little less relentlessly, campaign to stop the hegemony and ideally, kick that warmongering lunatic out of the White House.

Re:do no evil!

[Anonymous Coward]

Sure buddy, they've done us all a world of good. That's why the gap between the rich and poor has never been wider. Are you completely naive? Do you really believe these people are caring, loving or selfless? Personally, from what I've seen, they're clearly greedy, selfish and unprincipled. These people have been exploiting the ecology and the economy for years. They've perverted and twisted every principle this country was founded upon. It's so totally amazing that so many people still can't tell when they've been swindled. BTW, did you know you sure come off as a fanboy for the rich and famous?

Re:do no evil!

[Anonymous Coward]

There is one technical difference here. Most of the slogans you cite are advertising creations. They try to persuade you that the company is doing good, but they have no real significance beyond ads. "Crazy Lenny" doesn't have to be crazy. "Nuclear Hot Wings" don't have to be radioactive.

Google, on the other hand, put "do no evil" in their SEC filing papers. IANAL, and I haven't been paying much attention, either, but presumeably any stockholder could sue Google for doing evil, because Google stated that they would "do no evil" in their filings. That has real legal weight, where advertising slogans don't.

Re:do no evil!

[cicho]

"Sure there have been many setbacks along the way"

Say that again when you happen to be part of the collateral damage in one of those 'setbacks".

Re:do no evil!

[cicho]

And the difference is... what? A slogain coined by a different PR firm?

Re:do no evil!

[Zphbeeblbrox]

Great examples on all those companies failing to live up to their motto's. Now where was your google example? hrmmm don't seem to see it...

one moment......

nope you didn't list it. Ok I guess you have no proof that google isn't living up to it's motto. Until such time that they are demonstrated to have gone against their motto I will trust them. Trust that has been earned. I didn't hop on the google bandwagon at first. They earned my trust. Now all they have to do is keep it.

Re:do no evil!

[Anonymous Coward]

I think if Ford is burning pigs to death, more power to 'em.

Fuck the po-lice. Fuck 'em.

Re:do no evil!

[radish]

So? When you do something like Union Carbide did, you have a responsibility. If you get bought out by some other company (Dow in this case) they just bought that responsibility. They should not be allowed to wash their hands of the whole mess just by selling the plant and then selling the company.

Re:do no evil!

[Anonymous Coward]

Ok... so let's say that Google in some wierd twist of fate one day ends up buying Microsoft. So then Google would have their hands dirty with all of Microsoft's evil deeds? That is incorrect and you know it.

Re:do no evil!

[incognitopoet]

Ahhh. Ok. Since the guy didn't explicitly list the merger/aquisition history exactly correctly, his entire opinion is wrong. If you think the guy is wrong about big corporations not always acting in ethical ways, say so. Attacking some trvial error in an effort to discredit the author does not strengthen your position. Calling him a 'troll' does not give me any information about your opinion, which may be valid. (I don't know if it is valid or not, because you failed to state your position.) Labeling his post 'inflamitory' does not provide information on your position. Now, if your thesis is that corporations are always ethical, or that they do harm only without intention, let us know why we should agree with you. Maybe you can enlighten those of us that wrongfully believe that some corporations have in the past ignored the common good with the intention of controling costs or increasing revenue. You can start by correcting my understanding of the tobacco companies. I'd mention Phillip Morris, but aside from a hazy memory of something about "Altera" I don't understand the merger history well enough to discuss the subject.

Re:do no evil!

[d474]

Slashdot: "News for Nerds. Stuff that matters."

If that isn't the biggest oxymoron too! I mean, if Nerds that like stuff isn't...no wait,...Nerds matter and that's stuffy news...mmmm no...damn it! These PR lines are friggin' confusing me...

Re:do no evil!

[BlueLightning]

Ok... so let's say that Google in some wierd twist of fate one day ends up buying Microsoft. So then Google would have their hands dirty with all of Microsoft's evil deeds? That is incorrect and you know it.

I'm sorry, but there is no comparison. Microsoft hasn't killed anyone - the Bhopal disaster killed 20,000 people and continues to affect the lives of survivors and the people still living in the area which has still not been fully decontaminated because buck-passing corporations like the one you are apparently defending hold up their hands and say "it's not our problem".

Re:do no evil!

[Anonymous Coward]

They are also good (for us gov) becuase they know quite well how likely it is that you become a terrorist. You are already ranked in a demographic database based on all data harvested from you. "THEY" want you to beleve this is used primarily to provide tailored ads, thats not the primary objective.

Re:do no evil!

[Dun Malg]

And the difference is... what? A slogain coined by a different PR firm?

The difference is, the slogan was coined by Dow Chemical PR monkeys, and the Bhopal incident involved Union Carbide. The implication of the original poster was that the slogan was disingenious because of the the Bhopal incident. This implication is absurd because Bhopal and the slogan had fuck-all to do with one another until fifteen years later. It's like complaining that the US of A claims to guarantee religious freedom when the Spanish missionaries in the 1600's were forcibly converting indians in California to catholicism-- the two have fuck-all to do with one another. Cripes, I can't believe I have to spell this out for people. in 1984, Dow didn't gas Bhopal: Union Carbide did! The fact that Dow and Union Carbide merged a decade and a half later is fucking irrelevant!

Re:do no evil!

[Dun Malg]

So? When you do something like Union Carbide did, you have a responsibility. If you get bought out by some other company (Dow in this case) they just bought that responsibility. They should not be allowed to wash their hands of the whole mess just by selling the plant and then selling the company.

News flash, genius. First, Union Carbide sold their interest in the plant five years before the merger, so Dow was never involved in any of this purported "hand-washing". Second, the now-combined company hasn't attempted to abdicate its responsibility for the after-effects of the incident-- they continue to work towards mitigating the damage to this day.

Re:do no evil!

[Anonymous Coward]

Yeah!

When you do something like run down a pedestrian in your car, and then pay a large sum of money to compensate the damages, you still have a nebulous responsibility. If I later buy your car, OF COURSE IT'S MY FAULT that you hit the pedestrian.

By the way, Union Carbide didn't wash their hands of the mess. The fact that you're unhappy with the result doesn't change the fact that a democratic government representing the victims declared the matter settled, and that's as final a determination of what's just as is practical in this world.

"In May 1989, Union Carbide and Union Carbide India Limited (UCIL) entered into a $470 million legal settlement with the Government of India, which represented all claimants in the case. The settlement was affirmed by the Supreme Court of India, which described it as "just, equitable and reasonable," and settled all claims arising out of the incident. Ten days after the decision, Union Carbide and UCIL made full payment of the $470 million to the Indian government."

out of date cache?

[weighn]

From TFA: Caching of pages on Google's own servers + Differential fetching (instead of downloading the entire page, GWA will try send only what might have changed on the page)

must be some chance of inaccuracy if you're watching a page with frequent updates eg. live sport scores or stock quotes?

Re:out of date cache?

[Kris_J]

I imagine there's a hash exchange between your local cache and the server to determine what version you've currently got.

Re:Smart. Scary.

[lecithin]

"But I don't mind. This is partly because I don't think they're jerks (as far as public corporations go, anyway), but mostly still because I don't think they really care.
"

I apologize, but I think that you are being naive.

Perhaps they are not 'jerks' but they do care. Every thing that they log is information. Knowledge is Power.

Just my thoughts.

bandwidth?

[silicongodcom]

I've seen a bunch of info about their network, but are there any stats available regarding their actual bandwidth usage?

Spyware ? Banking, and whats wrong with gzip ?

[Anonymous Coward]

90% of web servers already serve content using gzip (even IIS4 has been for years and browsers all support it) so how is this going to speed up already compressed pages ? hmm snake oil

also what will the banks think about it ? banks/schools are already banning marketscore (who do a similar thing) i think that anti-spyware products will have to target this as its a threat to your privacy (like marketscore) and is technically spying on you (even if you know it is)

it would be a shame to see google dumped into the spyware pool but you reap what you sow and google being a public company profit is all their owners (the stock markets) are interested in

Re:Smart. Scary.

[Ninwa]

Don't forget they even know your oppinions! But I agree with the parent, I don't think they care. They have no reason to aggregate all of this data together, what would they do with it? Sell it? I think they have enough money as it is. I believe they're just offering a service for people which will enhance their browsing experience. I'm not sure how they will make money on the WebAccellerator but they do a lot of things that do not give them any proffit, so this isn't very alarming. Take your tinfoil hats off...

Re:Smart. Scary.

[okvol]

Why should Google mirror and cache everything? This seems inefficient. Why don't they just host the whole Internet? Maybe they can't because AOL owns the patent for that...

Re:Once again Google forgets us.

[grub]

Read the info page. You just have to point your proxy settings to localhost:9100.

Imagine the repercussion

[ghingy]

Google's innovation may seem to be making life easier for a few of us right now. We use their products even though we know that certain level of our surfing habits are monitored by Google and this is because they have so far done no apparent evil and we want to trust them. However, just imagine in the future, given that Google is public and if there's a shareholder coup or certain group of large shareholders would like to cash out, making Google susceptible to a takeover by another search engine, say Yahoo or worse, Microsoft. What then? There'll be no turning back for any of us then. There should be a policy that if a takeover happens, all past data that Google collected should be destroyed and the whole process should be made transparent.

Thw world's smallest violen plays....

[/dev/trash]

For the .0001% that have no speedup performance.

Re:Thw world's smallest violen plays....

[SirDrinksAlot]

Well my point is its slowin stuff up but staying its faster. Its complete BS. I'm up to 12 seconds now after half an hour of browsing.

I'd have to say i've lost 45 seconds to a minute now.

I'm calling shenanigans on this.

If maybe all the websites on the internet were being slashdotted all the time it would be an improvement.

I say we just start building Coral P2P Proxy into all web browsers and call it quits at trying to speed it up. Lets just average everything out for an acceptable speed while maintaining availability.

FreeNET would also be a good thing to start moving to - Bittorrent clients built wiht freenet supprot for trackers and such would be exellent.

Re:Smart. Scary.

[benow]

Tried Horde?

I like it...am I a freaking idiot?

[PenguinBoyDave]

Let's see here...can anyone say "Big Brother"? Sure seems like it, but damn...it's fast!

Your breath

[Anonymous Coward]

I wasn't going to say anything, but a mint every now and then couldn't hurt.

I am going to start a /. rejection support group.

Please meet me tonight in downtown Seattle at McMenamins Pub, Queen Anne.

http compression

[j1m 5n0w]

I wondered the same thing, and a few google searches came up with this.

Interesting project. It looks like apache supports compression, but not automatically (the user has to manually gzip the pages to be sent compressed).

Re:Smart. Scary.

[Neurotoxic666]

First, they collect your search information. Next they collected your email. Now they collect your destination. You put it all together, that is quite a bit of information.

I know this doesn't apply to the majority of the slashdot crowd, but I still find it strange that people will panic about a company that collects some personal information yet they'll cope with the fact that there's a god, somewhere, knowing all...

no linux

[protomala]

Well, nice thing, exept that I can't use because I'm a linux user.

Already one privacy problem

[Anonymous Coward]

Also, browsers other than Firefox and Mozilla can take advantage of GWA if you set them to proxy requests over Localhost:9100 while GWA is running in the system tray.

So basically, if anybody else is logged into the same system as you at the same time, they can figure out whether or not you have visited any given page by connecting to your GWA installation and seeing whether or not the page downloads faster than your Internet connection speed.

Re:Already one privacy problem

[Mojojojo Monkey Inc.]

Paranoid much? It's just as much of a "privacy problem" as IE or Firefox's local cache. Which means if you don't want your mom to know you were looking at porn, you'd clear your browser history and cache, and also clear your GWA cache which is easily accessible through the program options.

Re:Already one privacy problem

[Anonymous Coward]

It's just as much of a "privacy problem" as IE or Firefox's local cache.

Huh? Users have separate caches and can't access other users' caches. How, exactly is user 'a' going to retrieve resources from the Internet Explorer cache of user 'b'?

Re:I keed! I keed!

[oozer]

Slashdot needs an button for stories like this (any story involving Google really).

dial-up or metered gprs/umts alternative

[ironhide]

Google Web Accelerator is meant to speedup broadband browsing by prefetching.

There is another accelerator that does http gzip compression with jpeg/gif resizing and recompression; perfectly suited for metered gprs/umts or dialup modems.
http://rabbit-proxy.sourceforge.net/

WOAH

[gsibble]

Damn. This thing works well. My internet connection is moving faster than a crack addict from the NYPD.

Re:Smart. Scary.

[mikael]

First, they collect your search information. Next they collected your email. Now they collect your destination. You put it all together, that is quite a bit of information.

What is next?

They buy out www.archive.org and collect the evolution of your web site. It's going to be rather ironic to have google caching archive.org, and archive.org archiving google.

Then they buy out your ISP and collect the IP's that visit your web page (Of course, if your ISP has an online statistics page, they can already do this).

bummer! Re:isn't this basically a proxy?

[Anonymous Coward]

Funny response! Sorry you got modded troll. I thought it was funny and your response was to my post. Thanks for the yucks anyway....

Re:Smart. Scary.

[EpsCylonB]

I still find it strange that people will panic about a company that collects some personal information yet they'll cope with the fact that there's a god, somewhere, knowing all...

I don't know if there is a god (I prefer to believe in the provable) but the fact that I can cope with a possible god knowing everything about me doesn't mean I like it. Theres not a hell of a lot we can do about a possible god, google on the other hand...

This'll mean...

[Anonymous Coward]

that all information flows through Google, as they already practice censorship along the Chinese Firewall...

The internet will atrophy and from the ashes the Googleplex...under a single ownership.

Wow just like the last Doctor Who episode.

The Infancy of the Infocracy.

The AOL and NetZero web accelerators

[alphakappa]

Now who will pay $9.99 per month for those web accelerators :-)

A more pertinent question - how much RAM's it use?

[mbourgon]

On my new XP machine right now, doing nothing, I have...

• Mozilla using 33mb of ram
• Explorer using 28.5 mb of ram
• svchost using 19mb of ram
• CCAPP using 10mb of ram
• Corecenter using 8.5

etc, etc. My "commit charge" is 225mb, and my system "only" has 512. It sounds interesting, but so does not buying more ram.

My hypothesis is...

[Tamerlan]

... that this yet another revenue stream project for Google waiting to emerge. No, Google is not evil. Just another corporation.

This would be scarier if ...

[Skapare]

... it were offered by the RIAA and MPAA.

Privoxy is the best web accelerator

[xwin]

Privoxy (http://www.privoxy.org/ is the best web accelerator. It accelerates by removing ads which probably comprise about 50% of data on the page. It improves any connection and has added benefit of improving privacy. It is also cross platform.

Re:Smart. Scary.

[Anonymous Coward]

"Sell it? I think they have enough money as it is."

If they were still a private company, it would be possible to agree with you. But they are public now, their only goal is to make as much money as possible for the investors. Your data will be used to ensure profits. There is no doubt of this happening.

Re:Smart. Scary.

[soupdevil]

Am I missing something? I don't want a collection of applications or development tools (several of which do look promising, by the way). I want a web-based Outlook. So how do I use Horde to access my calendar, email and contacts from various computers. Is anyone hosting a Horde portal yet?

Re:Smart. Scary.

[Anonymous Coward]

"I prefer to believe in the provable"

Ever been in love?

I for one Welcome my new Google overlords

[MagikSlinger]

I'd like to remind them that as a trusted TV personality, I can be helpful in rounding up others to toil in their underground PigeonRank facilities.

(What? How did you think TrustRank was going to work?>

Ugh no Linux version?

[PatrickJ_M]

Dangit there's no Linux version. I emailed them asking for one. They'd better make one... I mean, all of their servers run on Linux, so...

Re:Ugh no Linux version?

[Anonymous Coward]

Did you actually plan on buying a Linux version and actually using it?

Or is this a case of platform penis envy?

Does it anonymize requests?

[Mustang Matt]

Effectively, if it's a proxy, couldn't it be used to anonymously access the web?

Not that google doesn't keep logs to let law enforcement see who you are, but in theory, the logs of the sites visited would see google unless they explicitly told them you're ip correct?

Re:Does it anonymize requests?

[Anonymous Coward]

Dear Sir or Madam: Your post has been rejected by our new email/post monitoring system.

Improper usage of "you're" please replace the offending word with "your".

Good day, Your Corporate Email/Post Monitor

This Has Little To Do With Web Acceleration...

[smug_lisp_weenie]

...and everything to do with decreasing loads/speeding up Google sites. After using it for several minutes, I noticed that any froogle/googlegroups/google search I do has marked time savings- more than any other sites I found (except CNN front page, which is also much faster and well suited for this kind of thing...)

Basically, running the web accelerator allows google to have compressed copies of all their pre-generated search pages and use the proprietary webaccelerator internals to give them a strategic advantage over web publishers/services/searches- Imagine the benefits this could have on their internal server load if adopted by 90% of web suers...

In typical Google fashion, a very clever move!

Re:This Has Little To Do With Web Acceleration...

[Kingofearth]

I don't really see this as a bad thing. I mean, people using this program will also use the google search engine so its helping their users. Now if they used it to slow down the time it took to load MSN search or Yahoo, then it would be a bad thing.

No Linux support either

[Anonymous Coward]

Not only that, you can't run it on Linux. I can't see why the Firefox version couldn't be a normal XPI instead of a Windows executable.

Re:No Linux support either

[mrchaotica]

XPIs are transparent, since they use Javascript and whatnot. Maybe Google didn't want people to be able to see their source code?

Re:No Linux support either

[SA Stevens]

Maybe Google didn't want people to be able to see their source code?

Hmmm.

Re:No Linux support either

[Myen]

(Note: I'm going by the info page, since I havn't installed this yet)

It's (partially, at least) a local HTTP proxy. Which means some sort of a binary (much like the desktop search stuff), and thus platform specific.

This also means it should work at least somewhat with other browsers (like Opera), you probably just won't be access the config screen easily. If you find the right page to access though you probably can.

Re:No Linux support either

[Anonymous Coward]

XPIs can be written in C++ as well.

http://www.iosart.com/firefox/xpcom/

However, I wouldn't be suprised if Google was hacking into the winsock or something and not just doing normal plug-in stuff.

Whoa, welcome to 1998 -- Slashdot automatically makes links!

Re:No Linux support either

[willCode4Beer.com]

I think there were probably other reasons.
Binaries are trivially easy to disassemble and pretty easily decompiled.
Its more like that the team working on it were more experienced at C++ or Delphi and chose that route. It memory serves me correctly the Google toolbar was a Delphi project.

If someone doesnt want anyone to know what their code is doing their best option is to not distribute their app.

Re:No Linux support either

[noamsml]

ummm... it installs a local proxy server and then sets the local proxy to contact google in a special way. this means that it is a binary.

Re:No Linux support either

[sifusam]

Yes, user of other browsers can setup a proxy to 127.0.0.1:9100 to take advantage of the web accelerator.

Re:No Linux support either

[sigaar]

Surfing on dial up is faster on linux in any way, so what are you complaining about?

Re:No Linux support either

[Panaflex]

No Such thing.

I looked over the install, it's a "simple" transparent proxy.

The XPI watches for browser interaction events and communicates these to the Client. For instance, a mouse over preloads the anticipated URL.

There's two primary C++ programs, GoogleWebAccClient.exe, GoogleWebAccWarden.exe.

The firefox files are:
GoogleWebAccFirefox.dll (GUI controls, event forwarder)
GoogleWebAccFirefox.jar (look and feel elements)
GoogleWebAccFirefox.xpt (Event monitor)

The IE files are:
GoogleWebAccToolbar.dll

The controllers are:
GoogleWebAccClient.exe (Receives events, controls proxy to google)
GoogleWebAccWarden.exe (Manages cache and search)

Just a complete guess though...

Re:No Linux support either

[mjtg]

I wonder how long until someone reverse-engineers this and releases an open-source version ?

Hmmm...

[whoami-ky]

I installed it. Seemed OK until I tried to check WindowsUpdate. Browser stopped cold until I turned it off.

Re:Smart. Scary.

[jp10558]

I can't really see what google (or anyone for that matter) can really do to accelerate web content on broadband connections. As I've said elsewhere, it sounds like they are running a caching server for you. I would guess you could get very near the same effect without the privacy problems with something like squid or Allegrosurf.

The only thing they might do is some compression, but this assumes a number of things to make your connection faster:

1) the content isn't already compressed. Lots of sites already gzip html etc...

2)The google servers have a faster connection to the server than you do (they might have a faster general net connection, but the effective speed changes by the minute due to server load, net congestion etc...)

3) Your connection to the google servers is faster than your connection to the destination server. This is likely true right now, will it be with massively more load - IDK.

4) Your computer can decompress content and run the google background process faster than it can recieve textual information over a pretty fast line. This might be true, it depends on your PC, and it's load.

The only thing I haven't talked about is whether google is going to compress images. Personally I think it'd be kind of stupid becasue:

1) they'd have to do it lossy, and so now pics look like crap.

2) broadband is rather fast for most web pics now adays. Broadband is just tending to get faster. It doesn't really have the limitations dial-up does that make this attractive in any way.

There is no good reason to sign up for this. Unless you like feeding more info about yourself to a big company, or can't manage to get effective caching through cheap or OSS software on your PC.

No, because it's Windows only

[Anonymous Coward]

I remember when Slashdot was a Linux site.

Re:No, because it's Windows only

[Anonymous Coward]

Duh! You're supposed to run it through Wine, with your Windows version of Firefox or IE. : p

Alright! I saved 0.1 seconds!

[ianbnet]

Which I then squandered writing this post. Still, I'm going to put this on my parents' machine with their ghetto-ass cable connection, and see if it helps.

Re:Alright! I saved 0.1 seconds!

[Trejkaz]

Mine is still on 0.0 seconds after a whole hour of visiting the sites I normally visit. It seems like this thing isn't what it's cracked up to be.

Re:Smart. Scary.

[EpsCylonB]

No.

woot, I am the first one~

[dance2die]

W00t, I am the first poster of the thread~

Ah darn it, i turned on "Never Check for New Versions" and for 10 minutes i have been seeing the cached pages only... argh..

Re:Smart. Scary.

[77Punker]

Google news is news, not usenet. Google groups is usenet.

So what?

[Anonymous Coward]

System Requirements
Operating System: Win XP or Win 2000 SP3+

Fat lot of fucking good it does me...

Google (proxy)hosting??

[siliconeyes]

Here is another angle that /.ers might not have thought of.

What if, with a minor addition, Google decided to actually host every site on its servers?? It already does something like with Google cache, but that is limited to text only. All they need to do is cache all the other content as well, and they will overnight become the world's largest hosting company. Whenever a user opens a website, Google diverts the request to its nearest server through the Google Web Accelerator and satisfies it from there.

This would also mean the end of us /.ing a server, and I suppose would put quite a few hosting companies out of business. I can't decide if this would be a good thing though. On one hand, it would mean having the whole bloody Internet cached on a server close to you, but on the other would mean that hosting companies would be out of business. The only one hitting their servers would be Google itself, which would then take over and serve the whole public!

Re:Google (proxy)hosting??

[Anonymous Coward]

Thats how proxy caching works..

What would be the point of this service if it wasn't caching? It would be slower!

Improved Page Rank

[skraps]

This could be used to provide a better Page Rank. Instead of determining worth based on links that exist, they will determine it based on links that are used.

correct

[adpowers]

Exactly. Now they can find pages that are rarely linked, yet may be valuable. I wonder if this also might allow them to search the 'deep web'. Imagine a user with this browsing an online chemistry database where the only way to find info is by filling out some text fields on a website. Now Google will be able to find this deep websites by having users due all the grunt work.

Also, they might use info about popular pages and browsing habits to improve search results (like I'm sure they are doing now with the Search History feature).

Andrew

PS: As soon as I saw this on GoogleBlog I realized the 'privacy' freaks were going to flip. If you don't like it, don't use it.

Re:correct

[Anonymous Coward]

Not only that, but it is also a beautiful solution to all the googlebombing, keyword-linking pages.

You know what I mean. Thousands of pages with nothing but keywords, some random readable text, and links to pages whose ranking they want to pump. These have become sofisticated enough that you can't tell them apart from real web pages just by looking at their linking patterns.

So what's the difference? Real pages are actually visited by people while spam pages aren't. You can use aggregated browsing data to set apart useful from non-useful pages.

Add this to Trust Rank and you got a winner. All you need is a very large amount of bandwidth.

Re:correct

[ergo98]

Privacy freaks??? You can be a privacy pacifist and still be a little concerned about this.

I realize that some people have the Google blinders on, so just sit back in your chair, close your eyes, and imagine a world where Microsoft did this. They would be absolutely EVISCERATED.

Re:correct

[DustMagnet]

Imagine a world where Microsoft did this. They would be absolutely EVISCERATED.

I wish. They've done far worst and are doing just fine. Slashdot doesn't really have the power to eviscerate anyone (or anything, Microsoft isn't truly human).

Re:correct

[Motherfucking Shit]

Exactly. Now they can find pages that are rarely linked, yet may be valuable. I wonder if this also might allow them to search the 'deep web'. Imagine a user with this browsing an online chemistry database where the only way to find info is by filling out some text fields on a website. Now Google will be able to find this deep websites by having users due all the grunt work.

This brings up an interesting point. Some content on the web is never linked because it isn't supposed to be found. I realize that robots.txt and .htpasswd are there for a reason, but if I place a file on a website and I don't link to it from anywhere, I should be able to reasonably expect that it will never be found by a search engine.

As for your chemistry database example, that piques my interest even more. Let's say that instead of a chemistry database, you're browsing LexisNexis. Is Google going to cache the LexisNexis data, which is only supposed to be available if you're a paying member?

What about other for-pay sites, where only registered users are supposed to be able to see the content? Let's say you login to site.com and call up pagex.html through the Google proxy. A few minutes later, I call up http://site.com/pagex.html through the Google proxy. Even though I didn't log into the site, will Google show me the content they cached during your visit? Will that content show up as a Google result, or in the Google cache?

There seem to be a lot of potential problems when someone as large as Google decides to set up a public proxy.

Re:correct

[Anonymous Coward]

What about other for-pay sites, where only registered users are supposed to be able to see the content? Let's say you login to site.com and call up pagex.html through the Google proxy. A few minutes later, I call up http://site.com/pagex.html [site.com] through the Google proxy. Even though I didn't log into the site, will Google show me the content they cached during your visit? Will that content show up as a Google result, or in the Google cache?

Actually, it's worst than that. If it works the way you describe, then I can deliberately poison google's cache. Set up a transparent proxy to redirect me to midgetporn.com whenever I visit microsoft.com. I type microsoft.com in my address bar, google thinks I'm visiting microsoft.com, but I'm actually pulling data from midgetporn.com thanks to a transparent proxy (with a filter to rewrite urls) running on my linux router. So, this google tool transmit all this midget porn back to google central thinking it came from microsoft. Hilarity ensues.

Re:correct

[adpowers]

Good point, may I add to it?

This may also help them determine which links are also useful. Remember that blog software company that was caught hosting spam pages (it was on /. a few months back)? They were a perfectly valid website and had lots of visitors, but hidden on the front page were a bunch of links to the spams. Google would hit these links, but almost no actual humans would. That is sort of along the lines of what you said, but it would allow them to be even more fine grained (find the links that users are likely to hit). Hmm, hopefully that makes sense, it did in my brain.

Andrew

Re:correct

[vitaflo]

I'm a little scared by this given its a Windows only thing ATM. Pages viewed pertaining to computers will tend to be very Windows centric (even more than they really are) given the fact that you can't run this on a Mac or Linux box.

It reminds me of an old script you could add to your website way back when that would do stats for you (and the company aggrigated all the data), but the application to do this was Windows only. Usually a person making thier own site hits it quite a bit, and this artificially inflated the rankings of how many people online use IE.

If they're going to use it to increase relavance, it'd be nice if they included the entire computer universe.

Re:correct

[SEE]

"if I place a file on a website and I don't link to it from anywhere, I should be able to reasonably expect that it will never be found by a search engine."

Does your site's referer log require a password to access? Have you avoided visiting any sites whatsoever from your unlinked page?

If the answer to either question is no, then your "unlinked" page is quite likely linked in somebody's referer log, and is therefore vulnerable to being spidered. Literally millions of supposedly unlinked pages have been automatically indexed by search engines via referer logs. Not (intentionally) linking a page is not a sufficient precaution against getting the page spidered right now, whatever Google does with this service.

Re:correct

[baadger]

What about other for-pay sites, where only registered users are supposed to be able to see the content? Let's say you login to site.com and call up pagex.html through the Google proxy. A few minutes later, I call up http://site.com/pagex.html through the Google proxy. Even though I didn't log into the site, will Google show me the content they cached during your visit? Will that content show up as a Google result, or in the Google cache?

I'm far from an expert but generally I would think access restricted (htaccess) mandatory login pages and directories will return "must-revalidate" in their Cache-Control header or "private" as the cache type type. This essentially means these pages are only cached by browser caches and other 'safe' caches.

Remember GWA is still just a normal compliant proxy.

Re:correct

[imaginate]

Yep - and how long before spammers start using the proxy programs to endlessly click their own links?

I'm tempted to trust Google for the moment, because every article I've read on search technology says that it will get better with more information. I'm sick of dealing with the crap spam links when I search, so I'm happy to help downrank them. Of course, if I catch wind of the data being sold to "market research" companies, then I'll bail.

But I wonder how useful the information will be to Google when it can ultimately also be manipulated by spammers. Maybe they have an algorithm to weed out spammers using their proxy (too many clicks, too often), but I also don't doubt that the spammers will do everything they can to appear to offer "useful" information to google in order to up their rankings.

Really, it just escalates the information war...

Re:correct

[Snaller]

As soon as I saw this on GoogleBlog I realized the 'privacy' freaks were going to flip. If you don't like it, don't use it.

Yeah, and if you think its wrong to keep slaves, then don't keep slaves!

Re:correct

[Anonymous Coward]

The company's reputation isn't enough for me to give an application the benefit of the doubt. Browsing habits be damned, when I pull up cPanel on my domain and it asks for my password, I want to know what happens to that password--if Google sees what I'm doing and decides to spider my control panel, then we've got a real security issue, and a possible liability for Google.

Perhaps it's safe, but it is also worth looking into.

Re:correct

[Anonymous Coward]

There's nothing wrong with improving search results. Microsoft didn't do it because they don't know how.

Re:correct

[adpowers]

Then make sure you access your cPanel using HTTPS. The Web Accelerator won't send HTTPS requests through Google's servers, so you won't have any problem.

Re:Smart. Scary.

[spagetti_code]

Google say they will do no evil. Great, and I trust that.

But what I also trust is that they will open their doors and computers very wide to the first FBI agent with a supboena, especially with the full weight of The Patriot Act.

Judges are handing wiretapping orders out like confetti, so you need to consider that any information held by any company belongs to the government at any time. All your base belong to us. And what's even scarier is that no-one is allowed to talk about it - all requests for info come with gag orders.

I'd be willing to bet that Google have already been approached for information.

What i'd like to know is what sort of data mining expertise the FBI is gathering in preparation for getting their hands on all googles files.

why no talking about adsense

[KalvinB]

most likely because they want to maximize the value of adsense. If everyone were all talking about how much money they were making on AdSense people would start propping up pages to target the most lucrative ads (they do already). The value of those ads would then go down. As it is it's all a big mystery and so people for the most part don't consider AdSense when deciding what content to put on-line.

The other problem with talking about AdSense performance is that your success or failure a) can't be proven and b) could influence other's decisions to or not to market using Adsense. How well or not someone else's site is doing with AdSense has exactly zero to do with how well it will do on your site but people think it does anyway.

If Google took away the gag you'd have thousands of people bitching about how little their site is making and it would make Google look bad even though it has nothing to do with them. Sorry but your crappy little Geocities site isn't going to generate enough traffic to allow you to quit your day job. You'd also have people going on and on about how much they're making which would cause people to have unrealistic expectations.

Google wants entire control of the PR side of AdSense which is reasonable. It's how they pay the bills and make investors happy.

Re:why no talking about adsense

[Anonymous Coward]

"Google wants entire control of the PR side of AdSense which is reasonable. It's how they pay the bills and make investors happy."

What..? You say it's reasonable that Google would want to gag (your word) the users of AdSense, because by gagging them Google can make the service more profitable.

What a justification! I wonder if the people modding you up actually read your post...

I'm not going to install this thing

[microbee]

Not just because of technical reasons (it might reduce the latency but it incurs more traffic and load on the machine and the Internet), but I am starting to feel uncomfortable of how aggressive Google has been trying to be. "Do no evil"? I hear the similar thing when Larry started to give away Bitkeeper to Open Source developers. Not that I say Larry is evil, but a company is a company. I cannot trust them without limits.

Re:Smart. Scary.

[Anonymous Coward]

Don't worry, pretty soon those unobtrusive text ads will be animated!!!

http://www.buffalonews.com/editorial/20050502/1067 216.asp

HTTP_X_FORWARDED_FOR

[Anonymous Coward]

most proxies anon requests dont so i doubt google will either

Re:HTTP_X_FORWARDED_FOR

[Anonymous Coward]

Check it here if you want

Re:Smart. Scary.

[blueadept1]

My god, can you not wait a just a few decades? Soon we will have the ultimate calendar. Prone to BSOD, though.

Builds Database better

[kgruscho]

Actually I think this is just a method of optimizing their cache and search capabilities.

You look for page, it loads page to you but also into its cache. This way google's cache is most up to date for the most used websites.

(This makes their webcelerydskfaslhf the best most likely besides having their huge caching capacity, it will also be fastest, now if we could just run sort of torrent system and make the whole net into a cluster)

It also lets google get all sorts of information of where people go AFTER a give google search, which is data about what the people really wanted in the first place.

Re:Smart. Scary.

[Anonymous Coward]

If you are plotting some scheme the government may not like, don't use Gmail, dont't search using google, don't use groups and don't look for bomb making materials on froogle. Especially don't do this from your own IP, or the IP of your *local* library.

If you are looking up restaurants, the newest MP3 player, cool music, places to live, jobs to work at and so on, by all means use Google.

Re:Smart. Scary.

[EnronHaliburton2004]

I'm sure Horde's great, if I have the time and resources to install, configure, host and maintain it on a spare computer in my house.

I would love to do that, but I don't have that sort of spare time.

Are there any service providers who run a bunch of the Horde apps?

I'm oficially freaked.

[fleppir]

Just fired off an email to their FAQ.

• How do you decide what constitutes an ad?
• How do I measure pages viewed when you deliver my content?
• How do I preclude you from caching for the accelerator while keeping my pages indexed and in the regular Google cache?
• Would you stop that please? (ps This constitutes Evil (tm) behaviour by definition, remember you can be bought out and then that thing is in the hands of the NEXT stockholder, not you)

Irrespectfully my own,
*blankitynameblank*
webcitizen

Give the benefit of the doubt

[exp(pi*sqrt(163))]

And after they do turn evil I guess you'll just ask them nicely to wipe the data on you they've already collected.

SPEED UP YOUR INTENET BY 500%

[Anonymous Coward]

Firstly, it doesn't clog up your Windows installation and slow down or crash your computer;
how do you know ? iam sure those other vendors will say the same too, or do you think Google has some sort of magical coders that other million dollar software/accelerators companies in the same market cant afford ?

in fact it speeds up your browsing.
right , they all say that

Re:SPEED UP YOUR INTENET BY 500%

[Mac Degger]

"right , they all say that"

Yup...and they all got found out. So you can bether ass that if google did, we'll find out too. And then google and it's adwords/sense is history.

But why would google commit suicide like that? I'm betting they won't, you're saying they are and have. I say show me proof, or pipe down. And if nothing fishy has been detected within, oh, say two weeks?, I'm gonna keep on assuming Google's OK and that you're wrong.

Re:SPEED UP YOUR INTENET BY 500%

[Spy Hunter]

I do think that the quality of Google's coders (and therefore of their code) is much, much higher than average. Also, Google has more resources than any of those other companies. Also, Google has always been upfront about their products, unlike spyware companies. Also, every single one of Google's products to-date have been high-quality and focused on helping the user, whereas spyware companies have a philosophy of "hurt the users to benefit us". Is it really that unreasonable to believe that Google's product will be better?

How is this better...

[bpuli]

than privoxy chained with squid with prefetch? Looks to me like there are a number of privacy issues with this. Not an even trade for the "boost". Google gains a lot more than you do.

Re:How is this better...

[Kris_J]

The partnership between Google's server and Google's local software can produce a better optimised stream than anything you can do purely locally. Basically, Google copes with that initial sloppy junk from the source website, after that it's all streamlined using whatever modern compression techniques Google feels like using. No waiting around for new standards to be formed, they can implement more efficient compression systems as soon as they're available.

why does Mr. "do not evil"..

[Anonymous Coward]

releases no software for *nix? Guess they want to be the next M$. Who cares abt the open-souce community anyways ;-)

Re:why does Mr. "do not evil"..

[MCraigW]

We shouldn't assume that the product is designed for bandwidth-rich Americans and Europeans.

From http://webaccelerator.google.com/support.html
3. Can I use Google Web Accelerator with a dial-up connection?

Dial-up users may not see much improvement, as Google Web Accelerator is currently optimized to speed up web page loading for broadband connections.

PageRank Implications

[pseinstein]

If they are smart they can use the information acquired from knowing what websites people visit to help with there PageRank system. They would now be able to record how much traffic a site gets and throw that into their PageRank equation to yield more accurate results.

TrustRank

[Dialithis]

Seems to me this is the way they are going to be able to seed their "TrustRank" metrics that they want to deploy. Why bother hiring "experts" to rank webpages when they can deploy a tool that will let all their users do it for them?

While their spider has no choice but to assume that all 10 links on a page are relatively equal, you and I rarely click on links that are clear spam or advertising, and certainly would not continue to browse a site that was simply there to spam the search engines.

They can measure both where people go, what they click on, and how long they stay there with this tool, so they can determine with high accuracy how important various sites are to various internet users.

Another anonymizer question..

[SmokeyMirror]

I know this has been asked before already, but does anyone have any idea if this could eventually be used as an anonymizer?

See, I live in one of those wonderful countries where our (one and only) ISP happily blocks our internet access, so this could be great for us.

Or of course, it could get google blocked out here...

Interesting. But wise?

[Quixote]

Let's assume that Joe Schmoe installs the "web accelerator". Next he downloads child porn. Who's responsible for this? Can he sue Google, claiming they "put it there" ?

Msr. Francois in France browses a Nazi site and Google happily provides the content to him via the handy web accelerator. Can the French go after Google now? (as if they're not already).

Chinese government demands that Google strip out offensive content and replace any references to Li Hongzhi with "<insert insult here>". Will Google comply? Has such a demand been made before ?

Plus, what about copyrights and such? Will Google be held liable for pushing out outdated pages? How will the servers (from where Google is grabbing pages) get their statistics? And since Google will be sort-of screen-scraping, why does Google object to it themselves?

Just some questions that come to mind.

Re:Interesting. But wise?

[Motherfucking Shit]

Let's assume that Joe Schmoe installs the "web accelerator". Next he downloads child porn. Who's responsible for this? Can he sue Google, claiming they "put it there" ?

I wouldn't imagine so. If Joe Schmoe downloads child porn on his Comcast cable modem connection, he can't sue Comcast, they didn't "put it there." (Well, he can sue anyone he wants, but he's not going to win.) Google didn't "put it there" either - he's the one who requested the URL, after all.

Plus, what about copyrights and such? Will Google be held liable for pushing out outdated pages? How will the servers (from where Google is grabbing pages) get their statistics? And since Google will be sort-of screen-scraping, why does Google object to it themselves?

If Google hasn't been found liable for keeping local caches of billions of copyrighted web pages, then I don't think adding a proxy service is going to change things any. A lot of ISPs (think AOL) already do some level of transparent proxy caching, in fact AOL by default even compresses all .JPGs into some lossy .ART files; it's not like this is really a new idea.

I am sort of curious about the web stats issue. If all of a sudden, 20% of my traffic appears to be coming from Google, that throws off my ability to judge the geographic makeup of my traffic. I at least hope that Google will pass through the correct (well, at least the user-reported) User-Agent and Referer.

Come to think of it, there could be more serious implications with the "geo IP" issue. A lot of online credit card processors will reject transactions if the purchaser's IP address doesn't match the general region they've used for the billing address. In my case, even if the processor lets one of these through, if I'm suspicious that the IP doesn't match the billing info, I cancel the order and refund the txn.

How is Google's proxy going to factor in? Giving them carte blanche is too great a risk, but prevent all of those users from making purchases and you're going to lose money.

Re:Interesting. But wise?

[h00dLuM]

Tilting at windmills again are we Don?

Re:Interesting. But wise?

[Quixote]

It is my job. :-)

But seriously: why do you think these concerns are dismissed so lightly? I'd be happy to hear decent rebuttals and answers.

For example: Google does cache pages currently, I know; but when it shows you the cached version, it tells you in plain words that what you are seeing is a cache, grabbed on X/Y/Z, etc. If they're acting as a proxy, they will not be doing that.

And as far as child porn goes: assume, for the sake of the argument, that the porn site deliberately feeds the Google cache-grabber child porn; and to others it feeds normal porn (whatever that may be). Now if Joe User gets child porn via Google, who's responsible?

This scenario isn't as far-fetched as one might think. For example: the people at FuckedCompany.com have models modelling their t-shirts. If you link to the model images from some other page, you get bestiality porn. Similarly, the FC chaps could, for example, feed you bestiality porn via this Google proxy (if they thought it was illegal linking).

Re:Interesting. But wise?

[Anonymous Coward]

really? Click here for the animal sex

Re:Interesting. But wise?

[wiill]

credit transactions / etc are *generally* encrypted using SSL ... and if they are you shouldn't be using them! the google accelerator says that it doesn't touch SSL connections ...

a new measure against search spam?

[drew]

i see a lot of people asking what good this does google other than as a way to harvest your personal information. while that may be true, if i had to bet, i'd guess that their purpose here is as a way to fight search spammers.

by indexing the results of real requests, they eliminate the problem of cheaters attempting to serve different responses to googlebot than they do to real visitors. it actually makes it much harder to even tell google has anything to do with it. on top of that, it gives them a new weighting metric to add to page rank. besides just knowing how many pages link to a page, they also get a useful measurement of how many people visit a page and all the pages linking to it. google has figured out how to use real thinking people to accomplish what has turned out to be one of the harder tasks for a machine to perform- distinguishing real data from spam.

no more slashdotted sites?

[demmer]

the first user to click through links in a /. story puts the data into googles cache by doing so... and users afterwards only get data from google... result... no more slashdot effect?

HOSTS file is now useless...

[CypherXero]

I installed it, and now (I assume due to the way the program works), my HOSTS file no longer blocks the domains I listed. I have a LOT of domains for ads, and I almost never see any advertisements. But now, the HOSTS file is useless. Just a heads-up for anyone that has an active HOSTS file on their computer.

Orkut accelerator?

[Warlock48]

That's all good... but will it accelerate Orkut, the slowest site ever? :-P

seems to work

[mtupker]

Privacy issues aside, pages do seem to be loading faster since I installed it. I think I'll keep it.

Installed: Odd behavior.

[guidryp]

I figured what the heck.

So I installed it and watched a bandwidth graph with it on off. Here are my observations:

1: Speed difference seem infintesimal. On the order of .1 sec reported by thier own tool on a page hit.

2: Actual amount transferred actually seems larger with the cache on! Every page load actually seems to grab way more bandwidth peak, now this might be great if you are really getting the page that much quicker but this doesn't seem to be the case. Looking at the visual bandwidth track it seems the same width, just higher on the Google cache transfer.

Very odd. I look forward to what others measure.

Re:Installed: Odd behavior.

[Anonymous Coward]

I would gather that the noticeable benefit will increase over time, depending on your usage. However, this is just a guess. I myself have noticed little 'saved time'. However I intend to keep the program installed, out of geek curiosity to see what happens over time

Re:Installed: Odd behavior.

[Kris_J]

This is a caching proxy with some pre-fetch stuff bundled with the plug-in. It will consume more bandwidth at the beginning, but as it caches your regular pages you should notice the traffic drop. Hopefully there's some nice compression between the local proxy and Google's server that will make the initial impact a little lighter, and the long term gains more siginificant.

Re:Installed: Odd behavior.

[guidryp]

The odd thing was I was repeatedly clicking on the same two pages. It should be done with the pre-fetching. The interesting thing for me will be how it works with slashdotting effect.

Re:Smart. Scary.

[Southpaw018]

Tinfoil, Post!

What tinfoil? I, for one, welcome our new Google overlords.

Free Advertising for Google

[JonStewart]

When you type in a page that has an error, you get something like this: Error: Host Not Accessible The web host www.penny-arcade.com is not accessible. Possible sources of this error: * The host name is invalid * There was a DNS error * The web site may be unavailable * You may not be connected to the internet Please edit the URL, or search for it using Google. Google Web Accelerator version 0.2.52.65-pintail.a Windows XP 5.1

The big picture

[Anonymous Coward]

I can see immediately two big impacts:

1) People in China will be able to circumvent censorship, unless the government starts to block access to Google itself.

2) PageRank will be more accurate because Google now has a clear idea which pages are more popular therefore more likely becomes what other people are searching for. Instead of "page links to page", it's "visitors link to page".

Re:The big picture

[Anonymous Coward]

Assuming that Google makes it available in China. Right now it says North America/Europe only.

Re:Once again Google forgets us.

[oldwolf13]

I didn't do anything to get Gmail working with Opera. I got a warning page about it and just clicked "Yeah yeah continue anyways". From what I've heard google maps don't work with Opera.

Like the original poster, I feel like google is just ignoring us Opera users as well.

Re:Smart. Scary.

[Tyler Eaves]

You're forgetting one biggie:

A fairly (but not totally) dynamic site that generates pages on request, but is generally fairly static minute to minute. I just installed the accelerator, and I can tell a noticable improvement over my cable connection when loading the front page of a site like CNN or Fark.

Re:Smart. Scary.

[Anonymous Coward]

That which is provable is what you know, not what you believe.

Don't you believe in anything? If not, that's rather cynical. Kind of sad...

No complexty wanted for sake of security

[Sithgunner]

Just sounds like it's getting more vague where the original information is coming from.

These days, all is talked about security, like man-in-middle stuff etc, now this is getting more hard to get the proper security, if you use it.

Sounds good for average users, but it's not like system admin wants to use some public proxy for their entire web surfing, and think something may be changed or spoofed along the complicated route from the real original source to the end computer.

Re:No complexty wanted for sake of security

[Sithgunner]

Forgot another point to say, but then again, not to mention that Google is going to collect what web page people view, intead of current 'passive' way of picking the quality page, they now trying the more aggressive way of fetching the number of hits on web pages.

Now, if they don't take IP that sounds like cool, but they will and while it certainly is difficult to identify a person out of IP, they can surely tell which company is looking at what web page in particular, thus sort of telling which company is interested in what at the moment.

At this far point, it just made me think, Google is trying to control and watch the entire web, just because they fetched a lot of webpages floating around.

Re:A more pertinent question - how much RAM's it u

[pseinstein]

It is running two proccesses: GoogleWebAccClient 7.1mb GoogleWebAccWarden 9.2mb

Well, thank goodness.

[DaveJay]

Thank goodness. One of my biggest computer problems is the slowness of web pages loading. I remember back when I had a modem, and pages loaded like lightning because the Internets were not very crowded yet. Now that everyone and their brother has the broadbands on their machines, it's too crowded. I hate waiting in line. Hooray for google!

Re:Smart. Scary.

[Anonymous Coward]

Does it allow sharing between users? If I can't publish a "public" calendar for friends, it's useless to me.

Re:Smart. Scary.

[jp10558]

Yes, but have you tried the alternative where you run the caching proxy (like allegrosurf or squid) on your own PC or a local server?

Big benefits there IMHO. One, I don't care how fast google is, it's not going to be faster than hitting your memory cache or hard drive cache. It won't even be faster than hitting a LAN cache.

Second, buy controlling that cache yourself, you avoid the privacy invasion. Plus, you can tweak it yourself, perhaps telling it not to cache certain pages so you're not reloading forums all the time, or even tweaking prefetching if you want that.

Big brother is watching you

[Anonymous Coward]

The majority of Google stock is owned by the CIA.

Re:Smart. Scary.

[Anonymous Coward]

Won't be long until google narks on you for the illegal content you viewed using its own image search cache.

Browsing habits

[kbahey]

The main thing they gain from this is people's browsing habits.

Imagine how much insight they could gain by having a large datawarehouse application that sifts through the billions of daily hits they get.

They can get patterns of how people access certain sites, by region, by time of day, by day of week, by site, breaking news effect, ...etc.

This is a gold mine for Google ...

Re:Smart. Scary.

[Anonymous Coward]

60 years down the line. When the original founders have retired, and they are hemoraging money. They will hire a medieval studies scholar with no experience in engineering as CEO. She will fight with Sergey Brin's surviving son for the rights to buy Yahoo. Google will buy Yahoo and then in 2 years lose 60% of their stock value. Then the CEO will have a thought. Why not black mail people into paying them money for all shady websites that people have visited. Then slowly after the passage of the Anti Web Board act of 2064 where subscribing to blogs will be a capital offense. They track down each and every slashdotter.

Okay, maybe not like that. But they are do no evil now. But later when the founders get killed in a mysterious plane crash, you will thank me.

The new Adwords?

[grolschie]

Won't be long until they insert their own adwords into every page you download.

Proxy

[DyslexicLegume]

I downloaded it for the fact that it goes through a proxy server.

It allows me to get around the block that has been placed on my router for Blogspot addresses without having to go straight through google translator.

Quite nice...and I've saved 9.8 seconds.

I really needed to those...my shoe was untied.

Re:Smart. Scary.

[SpaceLifeForm]

OK, Then I'll take the extended warranty.

Re:Smart. Scary.

[Anonymous Coward]

Regarding 2:

Remember that they are already retrieving a huge amount of the web on a regular basis for their search engine. They may well have a "faster connection to the web server", since they might have already retrieved a copy of the resource beforehand. Also remember that, assuming it's not dynamic content, they can serve copies of the resource to hundreds of thousands of people without having to download it themselves.

Basically, it boils down to whether they can serve the content faster than the origin server, and do so often enough to outweigh the cache misses. Given that plenty of origin servers are underpowered/underconnected and that they already have a decent infrastructure around the globe, this isn't hard to believe.

Also, they can use heuristics in the client software to decide whether or not to hit the Google servers or go direct to the origin. In the case of going direct to the origin, there is no slowdown at all.

As far decompression being slower than the network connection, that won't be a problem unless you have something like a 286 on a broadband connection.

Why would they do this?

[Anonymous Coward]

This is a simple and plausible way for Google to bypass robots.txt exclusions. "We were'd spidering you site, just providing a service to our users."

Works under linux!

[Rayban]

Works with Wine:

1) Install on a Windows box
2) Copy Program Files\Google\Web Accelerator files to linux box
3) "wine GoogleWebAccWarder.exe &"
4) Set your browser proxy to "localhost" port 9100
5) Surf with speed

If it fails, check your windows\temp directory for the google logs...

Note - this comment posted with Google Web Accelerator. :)

Re:Works under linux!

[Rayban]

Whoops - typo...

s/GoogleWebAccWarder/GoogleWebAccClient/g

It seems to stop running after a few minutes, but at least it works for a little while. :)

Re:Smart. Scary.

[cakoose]

They do predictive prefeching of pages. That's probably why they said it's primarily for broadband users. Though they might use their servers for this as well, predictive prefetching can be done entirely locally.

Apparently Mozilla/Firefox already does some prefetching, but the web page author needs to be explicit about what URLs to prefetch. Since the core functionality is already in there, it might be easy to write an extension to prefetch links as well.

Lets Me Get Around Stupid DRM Restrictions

[wynand1004]

I'm living in Japan and on occasion will find that some sites will not allow me access due to my location. For example, movielink.com states that you must be present in the United States to use the site.

I just installed the web accelerator and had no trouble accessing the movielink site. Awesome!

Of course, they'll probably set up their own servers here in Japan at some point, but until then, I can surf with impunity.

I really wish I had had this when Napster offered that free iRiver H10. I couldn't sign up because of my location. I was mighty mighty upset that I couldn't get a free mp3 player.

Re:Lets Me Get Around Stupid DRM Restrictions

[Kris_J]

Funny, Showtime still blocks me;

We at Showtime Online express our apologies; however, these pages are intended for access only from within the United States.

Re:Smart. Scary.

[Anonymous Coward]

Are there any service providers who run a bunch of the Horde apps?

Lots of Cpanel-based webhosts do.

Re:Smart. Scary.

[Anonymous Coward]

True, but these are first-generation people running a darling startup. They can be honest, kind, and generous while still raking in piles of cash. Everyone gets to be happy, and nobody has to be evil.

Now, when the founders leave the company to a business guy, he will start trying to squeeze for more money. When the brilliant shine fades and people don't throw money at Google, they might have to find new revenue streams. Eventually I expect that they will likely become at least semi-evil.

But for the moment, I think they are geeks with a blank check. They want to make the web a better place, providing better services and more convenience. They haven't started to plan all the evil things they can do, because they still have so many good things to do and plenty of money to do it.

Just my guess as a total outsider.

Re:Smart. Scary.

[Class Act Dynamo]

There's one thing we must consider. Let's give them full benefit of the doubt now. They are aggregating this information for the most non-evil purposes that exist. The problem is, what if the Google culture changes five or ten years from now. What if somehow the founders are forced out and the Google is run by people with nefarious intentions. Worse, what if Google corp. falls on hard times, gets desperate, and sees selling information as a quick fix when they are in a pickle. That would be my big worry.

Re:Smart. Scary.

[nonicenamesleft]

Try hotmail. I was pleasantly surprised to find it was quite fast.

Re:Smart. Scary.

[Koiu Lpoi]

google on the other hand..

What, boycott Google? And use MSN? You're joking, right?

Re:Smart. Scary.

[jp10558]

Great, but there's already software out there that will do all this without reporting back to google. Like allegrosurf.

I've also seen many people claim that prefetching doesn't help anyway, and hurts servers by hammering them for info that many users will never see.

Re:Smart. Scary.

[SunSaw]

The Google Web Accelerator blocks Yahoo! mail.

Summary

[zogger]

This is a Windows Microsoft article, it should have been noted in the summary. One additional word would have sufficed to make that clearer. Look at it from this angle, this is just an article about some big company releasing some windows software. Well isn't that special! It's not as much about google as it is that MS continues to dominate mind and marketshare, people just assume you *must* be running windows. "Oh, but it runs on firefox". Big deal, who cares, it's still windows as the primary and only platform this new whizzbang application runs on.

No biggee, it's a picky point, just a suggestion. I bet any number of folks are annoyed with it.

With every new google article and "feature" they are coming out with, I am liking them less and less, this is just another example of why. Screw their datamining.

Censorship

[Anonymous Coward]

I've heard that Google is making agreements with old money MSM entities like CNN and NYT to replace their PageRank system for common "news" searches like "Iraq" or "UNScam" with a system that gives less weight to page links and more weight to a Google defined score of "reputability," which will heavily favor MSM sources. If Google tries to shut out independent news sources like blogs, and, in fact, Slashdot, I think they're going to shoot their own search engine in the foot. Other companies with similar technology who have a committment to free flow of information will arise. There can easily be an Internet without Google, and they'd do well to remember that.

abuse ?

[thanew]

ok.. dedicated servers to handle this stuff.. its not going to accelerate if someone starts abusing it.. not to be cynical, but thats just people do to nice things.

3rd-World ISPs

[valdean]

I could see this being very useful in 3rd-world countries where ISPs have severely limited bandwidth.

Internet cafes in China, for example, are painfully slow... I have been in cafes where it feels like the entire room is sharing a single dial-up connection. At times, overseas webmail sites like Hotmail won't load at all. I've been forced to use a ssh connection to my server rather than load a webmail program.

Perhaps the Google Web Accelerator will help in places like this, where a small amount of bandwidth is shared by a large number of people. We shouldn't assume that the product is designed for bandwidth-rich Americans and Europeans.

Re:3rd-World ISPs

[Andrewkov]

Well, it's still the same amount of data going into the Cafe (how much compression can they squeeze on jpeg's?).. But instead of coming from random servers around the world it will come from Google's servers. I don't think the Cafe users will see much benefit.

Re:3rd-World ISPs

[Baricom]

Internet cafes in China, for example, are painfully slow... I have been in cafes where it feels like the entire room is sharing a single dial-up connection.

Are you sure they weren't?

Re:3rd-World ISPs

[anti-trojan]

I don't think GWA is doing it but there are other products like GWA that re-compress JPGs at a lower quality, and decreasing the number of colors in GIFs to achieve higher compression. I used one for a while and my average surfing bandwidth reduced to one third of original.

It Hosed my Firefox 1.0.3 install

[bogie]

Right now I'm posting from IE and trying to figure out what it screwed up.

Firefox wouldn't launch after install. After rebooting I see this http://img115.echo.cx/img115/6282/firefoxhosed5wg. jpg bookmarks, the address bar, and my personal toolbar links are gone.

Not exactly what I expect from Google. Although I'm sure its working fine for others I have a plain jane install that gives me no grief. It did work on IE btw, but it totally screwed up Firefox. Uninstalling did NOT fix the problem.

Re:It Hosed my Firefox 1.0.3 install

[spood]

I can't believe I just clicked on an image link to the .cx TLD. Not even a moment's hesitation. goatse has been away for too long!

Re:It Hosed my Firefox 1.0.3 install

[overseerbrian]

I think that it's just a firefox bug. The same thing happened to my 1.0.3 install. But it was about a week ago, before I had even heard about this new google program. I had to uninstall and then go and delete my profile before it would work again.

Re:It Hosed my Firefox 1.0.3 install

[ksemlerK]

It works fine on my machine. I am running WinXP SP2 using FireFox 1.0.3 set as my default browser. I can't vouch for IE, as I do not use it anymore.

Re:It Hosed my Firefox 1.0.3 install

[Blnky]

FYI, I had the exact same problem when I upgraded to 1.0.3. However, I wasn't using this new google proxy. I finally tracked it down to a misbehaving extension. Once it was disabled, I was able to get firefox up just fine.

Re:It Hosed my Firefox 1.0.3 install

[jaromanda]

Same happened to me, even after uninstall. Simply delete googlewebaccfirefox.trash directory where firefox is installed. The directory probably only contains googlewebaccfirefox.dll

Re:It Hosed my Firefox 1.0.3 install

[Thaelon]

That happened to me.

Something in your profile is hosed.

Try deleting your prefs.js in your firefox profile dir
(..Application Data\Mozilla\Firefox\Profiles\XXXXX.somethingproba bly"default").

You'll have to redo a lot of settings, but it's better than borked, eh?

If that fails try creating a new profile and copying stuff into it from the one you use now until it breaks, then undo your last copy.

If all else fails just create a new profile (Firefox.exe -profilemanager) copy over your bookmarks.html and redo all your extensions/settings.

I've had to do this twice because I can't quit monkeying with about:config.

Re:It Hosed my Firefox 1.0.3 install

[jaromanda]

the compreg.dat file in the profile directory was the only file with any reference to google web accelerator. simply delete it (or rename it if you like) AND delete any "googlewebacc*.*" files/directories in the "C:\Program Files\Mozilla Firefox" directory (as I mentioned above).

Re:I keed! I keed!

[alatesystems]

You mean Googledot?

Re:I keed! I keed!

[plutonium83]

".. and there's no catch!"

Unfortunately, the catch is google now knows your surfing habits, and their's no privacy policy.

Geo IP users now think I moved to Mountain View

[Anonymous Coward]

I visited thehun.org while running dumeter as I wanted to see what a link heavy site would do to the Google Web Accelerator.

I noticed a geo ip advertisement at thehun that normally recognizes I live in Phoenix and offering to introduce me to hookers in Phoenix. Now though, it wants to introduce me to hookers in Mt. View.

So that sucks.

yeah

[deafff]

And this years award for teh most stupid idea goes to...mic...err...goo^H^H^H...

Of course it's for gathering data.

[Bongo Bill]

Don't automatically assume that this is a bad thing. Google's gathing information about browsing habits with this, no doubt, but what can they do with it without causing their stock value to plummet as a result of negative PR? They can improve the quality of their searches (which is good), they can deliver ads that are even more relevant (which they already do anyway), and they can take market share away from companies who don't make products that are as effective (which is good).

Request on mouse over?

[Kris_J]

I'm watching my local proxy's logs and I'm seeing a request every time I mouse over a link. Can anyone confirm this behaviour? Sounds like it would slow things down, not speed them up.

Re:Request on mouse over?

[Mardak]

Yeah, I can confirm this. I followed access logs for apache on my computer as I hovered over links, and google servers made requests to my computer.

It wouldn't necessarily slow things down. Google is the one using its bandwidth (as well as the host), but the user will get a possibly compressed file from a nearby speedy google server. And it's not all bad for the host either because many people getting copies from the google cache reduces bandwidth for the original host.

Adblock?

[Fish Heads]

Anybody notice any issues with the Adblock plugin no longer working in Firefox with this thing installed?

Neuter GWA in FireFox and IE and make it optional

[melios]

To neuter GWA:

Internet Explorer 6:
Tools -> Manage Add-ons
Disable:

Google Web Accelerator - Toolbar
Google Web Accelerator Helper - BHO

Tools -> Internet Options -> Connections
Remove the proxy entry, or use your own, like The Proxomitron or Proximodo.

FireFox 1.0.3:
Mozilla Firefox\chrome\chrome.rdf
Mozilla Firefox\chrome\overlayinfo\browser\content\overlay s.rdf
Mozilla Firefox\chrome\overlayinfo\global\skinstylesheets. rdf

Search for "googlewebacc" and remove any <RDF:*></RDF> sections that contain it.

Tools -> Options -> General -> Connection Settings...
Remove the proxy, or use your own such as the two linked above.

No more toolbars!

Now, if you want to use GWA selectively, you can use the aforementioned Proxomitron or Proximodo and configure them to use an external proxy, with the 127.0.0.1:9001 address as you would for other browsers. Then setup FireFox, IE, or whatever to use 127.0.0.1:8080 as the proxy address.

Re:Smart. Scary.

[Anonymous Coward]

There is no good reason to sign up for this...

It's very usefull if you work in a company that blocks pr0n sites...

Does this mean...

[Anonymous Coward]

FASTER PR0N OMG !!

What if I already have a proxy setting?

[hoyty]

The software just automatically changes IE (and I assume Firefox) to point to a .pac on port 9100 of localhost. This is fine and dandy for most home users. What about those on a corporate or educational network that either use explicit or transparent proxies already. How does this .pac interact with those existing settings? Any way to see the .pac file?

Re:What if I already have a proxy setting?

[Kris_J]

I had the following proxy chain;

Browser -> Localhost:8080 (Proximitron with ad filtering) -> 192.168.3.3:8080 (Proxomitron acting as a relay to jump networks) -> 192.168.1.1:3128 (Squid) -> The Internet.

When I did a default install on my local PC, the following was formed;

Browser -> Localhost:8080 -> 192.168.3.3:8080 -> 192.168.1.1:3128 -> Localhost:9100 ! Fail !

So I turned off the jump to 3.3;

Browser -> Localhost:8080 -> Localhost:9100 -> Google's server -> The Internet

Which works, but now has traffic going through the "wrong" ADSL connection. Then I worked out that I could setup Google's webaccelerator further down the chain;

Browser -> Localhost:8080 -> 192.168.3.3:8080 -> 192.168.3.3:9100 -> Google's server -> The Internet

I'm not positive that 3.3 is routing through the 192.168.1.1 gateway, but if I really cared I'm sure I could setup a default route for Google's IP range. What I'm more likely to do is suggest that staff start using the 192.168.3.3:8080 proxy, which will automatically route them through Google's accelerator. After I explain the privacy implications first, of course.

Re:I keed! I keed!

[IconBasedIdea]

They were waiting to see if Roland Piquepaille was going to submit this story with a link in his weblog or whatever...

Re:Smart. Scary.

[MerlinTheWizard]

What's next? Hopefully a calender.

(A calendar.) Great, so now they'll know what you search for, what you visit *and* what you are going to do in the future. You're right, not much freedom left after that one...

Darn, installed it on the wrong PC

[Kris_J]

I just realised that rather than installing it on the local PC, you can install it on any Windows PC with an existing proxy, and redirect all your requests from anything through it.

/heads off to install it on one of the servers.

Re:Smart. Scary.

[ryusen]

i'm still waiting for Google Porn... i mean, i'm sure they didn't sue Booble for nothign right?

Re:Smart. Scary.

[Anonymous Coward]

What i'd like to know is what sort of data mining expertise the FBI is gathering in preparation for getting their hands on all googles files.

What I would like to know is how they are planning to collect the evidence. I mean this much information about this many people would take ALOT of space to store and a LONG time to search for a specific person, which is pretty tricky to do especially if you want to avoid a lost update problem without turning off the service for the duration of the search.

Im using it now, no issues yet

[mybase.load(Chaz)]

Im using it, dont seem to notice any issues, although i dont notice too much of a speed increase either.

Wikipedia

[Anonymous Coward]

I have noticed that after installing the Web accelerator, Wikipedia seems to be much more responsive. Has anyone else noticed this trend?

Re:Smart. Scary.

[slizz]

The agreement you have to sign says exactly how your information is used. If they used your information in an evil way, it would be illegal. (and then you could sue them for some of their billions of dollars)

will the technology work?

[walterbyrd]

Putting aside the issue of whether google is evil or not. Will this really accelerate the web?

It's just more compression, right? Correct me if I'm wronge, but isn't practically everything on the web already compressed? And, won't you actually slow things down by trying to compress the already compressed? Since compressing adds it's own overhead?

JPEGs and MPEGs and MP3s are already compressed, so are .zip, and .bz2 files. Don't modems have some sort of compression built in?

Isn't this going to be as big a joke as those web-accellerator services? For $5 extra you will get your data 5X as fast?

Re:will the technology work?

[Kris_J]

This is three things;

• Compression
• Caching
• Differential downloading

I can't imagine they've improved much with the caching, but if it's true differential downloading, your web surfing will get much faster. As for compression, there are far more efficient things in the world than mod_gzip, or for that matter JPEGs. The people that now own Stuffit (Allume) have got a nice program that can compress JPEGs by roughly 20-25% -- one hopes they're scrambling to offer something to Google.

You ignorant little monkeys

[lohphat]

I don't think they would piss away their credibility for profit...

What part of "they're now a public company" do you not understand. Those in control now will not be those in control tomorrow.

Public corporations have an obligation to the shareholders, not the customers, to make every attempt to increase value by any legal means possible. That's why there's a shareholder lawsuit mill waiting to pounce on any "missed opportunity" to use their IP to generate more profit.

Just wait until the founders percentage value drops below 50%.

It's how the game was designed to work. You want altruism? Stay private and/or not-for-profit.

Re:Smart. Scary.

[rthille]

Well, there are java tools that let you use POP3 to access your Yahoo mail, if that would help.

IT STORES A HUGE HISTORY CACHE ON YOUR HD!!!!

[mybase.load(Chaz)]

After only about 5 minutes of using GWA, it had stored my history on my local HD and used over 9 mb to do it. You can view this in GWA's preferences. You can't turn this off either, just delete the history. I knew this was too good to be true.

what if google = echelon?

[VoiceOfRaisin]

every single thing google makes seems to have privacy concerns. and everything they do people question how the hell they can make money off it, yet love it cause its free. they employ a ton of people, they have bandwidth like you wouldnt believe, they have a building filled with servers to the roof. all from free to use services? doesnt this seem to anyone as too good to be true? everyones wondered how the hell echelon would work.. how could they tap everything? how would they store everything? it cant be done! well folks, it can be done, and google is doing it! look at how insane the government is about security and information, would you really put it past them? as others have said, the only thing pointing away from this is PR quotes. you dont think people could possibly lie do you? nah.

Re:Smart. Scary.

[Anonymous Coward]

I know MS is evil and all that, but: http://www.mailstreet.com/ (no, it's not free!)

Not to mention passwords!

[Anonymous Coward]

Google will have usernames and passwords for any site anyone goes to with their proxy set to google!

GWA Easter Egg(s)!

[mrklin]

Located in the menu item GWA Help -> About GWA

tempus rerum imperator = Time Commands All Things

Contact e-mail address also has a plus sign: labs+webaccelerator@google.com

ME TOO! MOD PARENT UP

[robfoo]

After installing, I got an XUL error on trying to load firefox. So I removed the google accelerator (from add/remove programs), and then after a couple of goes, firefox started, but I've got no adblock and no google toolbar.

It removed the two extensions I use the most!

Bad, bad, BAD google! I am now rather pissed off.

Re:ME TOO! MOD PARENT UP

[Anonymous Coward]

Bad, bad, BAD google! I am now rather pissed off.

Stop blaming Google for Firefox's bugs.

Re:Smart. Scary.

[d_jedi]

I, for one, welcome our new computer Overlords.

works

[kcbernfeld]

I'm using it now, and it works quite well. 3.2 minutes saved!

Google == Linux, yet no Linux Client?!?

[TheCeltic]

Shame on Google! They do great things and add alot to the community.. They run the servers on Linux, why is it that they don't write clients for ALL of their tools (googlebar, web accelerator, etc) in Linux as well as Windows? Just a rant.. but an accurate one!

Re:Google == Linux, yet no Linux Client?!?

[giacomo-b]

they understood the real role of linux: TCP/IP web server...
there's no space in the immediate future for linux as a desktop system...

What if?

[feronti]

It's really an attack on akamai? Think about it... everyone starts proxying through google, so now popular web sites no longer need akamai to balance their loads. Once akamai's dead (or even before), google starts offering preferential or fine-tuned caching for websites for a fee. Brilliant.

Continuous trafic

[se7en11]

I just installed the web accelerator and noticed that as I was doing nothing my router was going crazy. So I simply shut down the accelerator and the trafic stopped.

If I'm not active, why is the web accelerator send information?

Re:Smart. Scary.

[skaeight]

I don't think searching is a problem for google.

Re:Smart. Scary.

[timmy the large]

Google Cache is caching content on your computer. They are prefetching likely webpages, such as the second page of the article you are reading, and placing them on a cache on your machine. What will make this an improvement over regular proxies and caches is that google has a huge amount of data to use in predicting what pages you may visit from the page you are currently at and caching all of them on your computer.

The down side is that google now owns your ass. As many others have pointed out between this and other products of theirs that a large majority of surfers use google has everything about you. They may actually beat MS at becoming the Borg and swallowing humanity.

Re:Smart. Scary.

[Huogo]

Horde comes with CPanel. Many times professional web hosting companies provice CPanel, which includes Horde, Squirrelmail and Neomail as webmail options by default.

Re:Smart. Scary.

[Anonymous Coward]

Finally we know what happens when Slashdot's love for Google collides with it's hate for the large corporation.

Prefetch can be dangerous

[Presto_slashdot]

I'm surprised no one has mentioned this (or at least I didn't see it in highly-rated comments.)

Most weblinks are harmless content URLs, but prefetching certain links can actually cause a side effect - besides the obvious one (messed up weblog stats.) For instance there are plenty of web UIs that use links to delete items, log out, etc. Simply hovering over these links will cause them to be invoked.

Most links are probably well-formed, such that actions use a dynamic URL link http://whatever?action=..., and the prefetcher won't follow those. But I've seen several sites that use apparently-static links to perform actions.

Wide adoption of prefetching is going to hurt these sites much like the XP SP2 "download protection" did... Should be interesting to see how it plays out.

Re:Prefetch can be dangerous

[pbreit]

Except that GWA only pre-fetches links that explicitly request to be pre-featched using {link rel="prefetch" href="/"}.

Re:Smart. Scary.

[cicho]

"Google say they will do no evil. Great, and I trust that."

Google's *founders* said that, and you or I may trust them, because they're geeks and they're doing cool stuff. But did google shareholders say that too?

Whatever information Google now has that it is choosing not to use or is using in a benign manner *will* eventually be used to detriment of Google users' privacy if the shareholders decide it's gonna raise their "value".

This will surely display

[Anonymous Coward]

the popularity of tubgirl among the INTARWEB users >_>

Re:Smart. Scary.

[Anonymous Coward]

In Texas, it's calender...ask the president. Oh, and by the way, in Texas the calender picks you!

My short try of it..

[nolife]

Just for reference...

Kerio personal firewall popped these up during the install of the MSI (expected I guess, cert maybe?):

Remote Point: 12.158.80.10, port http [80]
Remote Point: crl.verisign.com [12.158.80.10], port http [80]

Then:
Direction: outgoing
Local Point: localhost [127.0.0.1], port 1980
Adapter: N/A
Remote Point: 127.0.0.1, port 1980
Protocol: UDP
Application path: C:\WINDOWS\system32\msiexec.exe
Description: Windows® installer

Then Spybot teatimer warned about changes to the registy (expected):

Shortly after, Microsoft AntiSpyware warned about three times about an IE toolbar being added and would I like to allow it (expected as well).

After restarting my browser and clicking on the icon, another warning from Kerio:

Direction: outgoing
Local Point: 0.0.0.0, port 2037
Adapter: N/A
Remote Point: localhost [127.0.0.1], port 9100
Protocol: TCP

Followed again by MS AntiSpyware and then Kerio alerting about the Google binary trying to contact the internet.

With all these damn protection mechanisms loaded, I cant get anything done ;)

Shortly after, I uninstalled it. I'll just stick to local network squid proxy for now..
On a side note.. My squid operates transparent and I use some filtering as well. I tried to go to a site I know is blocked and it was still blocked with the Google Accelerator running. Once again, as expected because the Google proxy is running on the local machine. I wanted to check and verify for myself though.

Re:Smart. Scary.

[agentkhaki]

Just a note on the time aspect, Horde's actually pretty easy to install and config. We're talking an hour, tops.

Re:Smart. Scary.

[Cromac]

They already know where you go. Adsense is everywhere. Remember DoubleClick?

DoubleClick? What does 127.0.0.1 know that it didn't know already?

IE users: it DOES work with a corporate proxy

[Steven Reddie]

Posting here to be noticed, since a number of people have complained it doesn't work behind a corporate proxy/firewall. It seems that on first use it trashes the proxy setting, but to fix it:
1. Stop GWA
2. Fix your connection settings in IE
3. Start GWA
4. See connection settings in IE redirect to localhost:9100
5. Stop GWA
6. See connection settings as before (if not, repeat from step 2)
7. Start GWA
8. GWA should be using your proper proxy

Re:IE users: it DOES work with a corporate proxy

[Steven Reddie]

I may have spoken too quickly. After following the steps above IE works through GWA, however as far as I can tell from netstat the connections are still being made directly from IE to the corporate proxy -- perhaps GWA is simply passing the real .pac file through to IE which then parses it an uses an IP address as specified in the file. That would explain why I'm still seeing "0.0 seconds saved" even after doing some browsing.

Re:IE users: it DOES work with a corporate proxy

[Steven Reddie]

Hmm, trying a fixed proxy address, the .pac returned from localhost:9100 looks ok, ie. looks like it returned localhost:9100 as the proxy for non-local/non-bypassed urls. However, netstat still shows that IE is accessing the corp proxy directly. Don't know what's going on, maybe IE choking on the regex in the .pac file.

Re:Smart. Scary.

[Kris_J]

they'd have to do it lossy, and so now pics look like crap.

No, they wouldn't.

Re:Smart. Scary.

[deasach]

Pretty much similar to the way Marketscore used to work. (They've changed tack in the last month, and no longer proxy infected users' traffic, just replay the HTTP requests back to Marketscore instead.) Except Google have specifically stated that they're not proxying https traffic, which is good (unlike Marketscore, which not only proxied the https traffic but also decrypted it through an ingenious man in the middle attack).

And the reason why Marketscore did that was that their parent, Comscore, could use the data gathered for website ratings. It's extremely valuable data in many ways.

What are Google up to?

free proxy?

[twohorse]

I wonder how long before this will be blocked by firewalls/proxy servers etc? And if it isnt blocked it'll be pretty tricky to filter as its already compressed...

slashdot effect

[Danathar]

What will be interesting is how well it works when we get a story where the site is slashdotted. If google's web cache works correctly it could be as effective or more effective than the coral web cache system

You can probably centralize this!

[Guspaz]

The GWA does three things that save bandwidth:

1) Compress all HTML content passed between Google and client
2) Use local cacheing
3) Send diffs (just what changed) of files that are in the cache but out of date.

I could refresh Slashdot over and over, and the only thing that I would have to download when the page changes is a compressed diff, probably a savings of at least an order of magnitude.

In the situation you describe though there are many computers using one connection. Since GWA interfaces with browsers via a simple HTTP proxy (The IE/Firefox integration is just for the "x seconds saved" display), you should be able to install GWA on one computer and set multiple other computers to use that proxy.

Of course Google might check to see if the machine making the request is on localhost.

Re:You can probably centralize this!

[Andrewkov]

Plus DNS lookups will be faster, I didn't think of that before.. Slow DNS servers can really slow down the perceived speed of the line.

Re:You can probably centralize this!

[fuzza]

3) Send diffs (just what changed) of files that are in the cache but out of date.

This sounds suspiciously close to the intent of RProxy. Might it have the same unfortunate patent problems? (One would hope that Google would have checked that sort of thing, though...)

Re:You can probably centralize this!

[Guspaz]

Indeed it is, but of course it also does more.

There are different ways of generating diffs. You could do a per-line diff, or a per-byte diff. Or Rsync, which I understand works differently entirely.

Hopefully whatever method Google uses is not covered by the patents.

Web banking.

[ElAurian]

From Google's privacy policy page:

"To enhance Google Web Accelerator's performance, Google temporarily caches cookies from third party sites that are used in your Web requests. For more information, please see our FAQ."

I know that I'm certainly not logging into my online banking site while using Google.

Re:Web banking.

[The Cisco Kid]

As long as your bank uses https, that isnt a problem. https cannot be proxied or cached. Either their proxy recognizes https requests and lets them go direct, or it uses the tunneling mode of allowing the (encrypted) data to be channeled thru the proxy unmodified (and uninspected)

Re:Web banking.

[ElAurian]

Awesome. Thanks for pointing that out, I need to read more closely. :)

Re:Smart. Scary.

[YoungHack]

I can't offer what you ask for, but I have really liked Plans (http://www.planscalendar.com/) myself. I'm pretty sure there are others out there who would to.

Re:I keed! I keed!

[AyeRoxor!]

and their's no privacy policy

their's no gramur eyethur

Looks like people forgot one thing

[Anonymous Coward]

This would help Google screw all other online Ad companies because the click info will no longer go to the Ad company but will be going to the Google cache. Pay-per-click will no longer work. If the Google Web accelerator doesn't set the referrer right (I doubt this would happen), it could screw other kinds of billing terms used by online Ad companies. Hmmm... improve the value of AdSense by bringing down other companies or their products!!

What's your take on this?

-Itsme

Some overzealous modder needs a sense of humour...

[_undan]

I was trying to point something out. The odds of google releasing their tools as open-source is remote. This makes the possibility of most Linux users installing it even remoter.

Maybe if I hadn't used the word Zealot...

Re:I keed! I keed!

[PKPerson]

Just wait till the accelerator gets slashdotted

Somebody needs to tell them...

[amrittuladhar]

Look at the screenshot at http://webaccelerator.google.com/done_ff.html . Somebody needs to tell them that their Firefox needs to be updated. And, I think it's doing pretty well, specially with websites from servers that I have visited before, such as Fark Photoshop contests. But it didn't work so well for a site hosted outside the US (I tried http://www.nepalitimes.com/ ).

Why you "can't really see"

[seanadams.com]

I can't really see what google (or anyone for that matter) can really do to accelerate web content on broadband connections. [...] There is no good reason to sign up for this.

The reason you're skeptical is because you don't know as much about the Internet as google does.

When you download a web page on your 6Mbps cable modem, do you think it instanly goes to 6Mbps throughput, transfers the page, and then drops to zero? It doesn't. The efficiency *decreases* as your connection gets faster (which is why google does not claim to speed up slow connections - there's little room for improvement). Here's why:

The TCP stack under your browser starts by establishing a connection (3 way handshake). Then it sends a packet with the HTTP request. Finally after those long round trip times of basically doing nothing, your browser starts receiving HTML. As the HTML comes in, the process repeats for the embedded stuff (images). If you have a fast link (and especially if the server is far away), your link spends a lot of time doing nothing while connections are established and transactions take place.

By routing your connection through google, many efficiencies can be gained. These are listed in, of all places TFA. It's not just caching, either. Prefetching, for example, is a trick where their servers will start requesting and transferring the images within a web page, even before your browser has requested them. Since the HTML already went through google's proxy, they know what your browser is going to request before your browser does.

So instead of just pooh-poohing it because you don't understand the technology, why don't you go download a copy of Ethereal, which will let you see these tricks in action. Then you can offer us a more educated opinion based on empirical fact, instead of a long diatribe amounting to "I don't understand how it works, therefore it sucks".

Re:Why you "can't really see"

[jp10558]

All that is great, except for 2 things.

One, you can do prefetching without selling your soul to google. Allegrosurf is good at this.

Two, pipelining. All modern browsers use pipelining, which severely limits the amount of handshaking that needs to be done to a server.

Final comment, from what I've seen, the people who are using this program (at least with Opera) seem to see no improvement in the first hour or so of use. In fact, some are reporting slowdowns.

I maintain my reservations about this being able to offer a significant boost to browsing, especially when contrasted to the major privacy intrusion.

Re:Why you "can't really see"

[radish]

HTTP/1.1 persistent connections have already vastly reduced the overhead associated with a typical page request. I also don't understand your point about prefetching images within a page - according to the HTTP spec most browsers limit the number of connections to a server. This can of course lead to a slow down in some cases (typically where bandwidth is high but connection latency is also high). But the browser would have to limit the connections to google's cache in the same way, and for google to be a good internet citizen it would have to limit it's connections to the original host too. I really don't see any dramatic savings coming out of all this.

Re:Why you "can't really see"

[Trifthen]

Yeah, I was about to say that. HTTP/1.1 basically gained a foothold in 1997, so we've had the ability to do subrequests sans handshaking for at least the last 8 years. In addition, even if Google prefetches, you still have to get the images/whatever from Google, meaning more requests.

As a web developer, I've actually found that the best way to reduce page load time is highly aggressive caching - like telling the browser with headers to use a cached copy if it has one, and closing the connection. Most browser like this just fine, and it can be overridden if they *really* want a new copy.

Now what would make the web *really* fast would be some kind of CVS system, so the browser can say which version it has cached, and you send a diff to update to the most recent version. ^_^

Re:Why you "can't really see"

[ZephyrXero]

How is this any different than the about:config pipeling trick for Firefox? Instructions on how to speed up your Firefox pageviewing (for broadband users)

Re:Why you "can't really see"

[Anonymous Coward]

The reason you're skeptical is because you don't know as much about the Internet as google does.

I think what you meant to say was: "The reason you're skeptical is because you don't know as much about the Internet as I DO! And here I will prove that I am so much smarter than you! nah nah nah boo boo."

Re:Why you "can't really see"

[Anonymous Coward]

I don't want google caching or prefetching my content, if they want to serve my content then they are welcome to provide free hosting! For my purposes I need root because I'm running custom httpd's and backends coded in C for performance.

At this rate, it wont be much longer until Googlebot is unwelcome.

Re:Why you "can't really see"

[EXrider]

I kinda remember reading somewhere how some web servers don't support HTTP Pipelining, and by turning it on you might screw up browsing some how if you come across one of these websites, I might just be imagining things.

But I did just turn pipelining on Firefox here at work, and so far slashdot seems way faster.

Re:Why you "can't really see"

[Jellybob]

Now what would make the web *really* fast would be some kind of CVS system, so the browser can say which version it has cached, and you send a diff to update to the most recent version. ^_^

Isn't this what the Google proxy does?

From what I've read it appears that it only sends you a diff between the current version, and the last version you downloaded.

Re:Why you "can't really see"

[seanadams.com]

One, you can do prefetching without selling your soul to google. Allegrosurf is good at this.

Please explain how a piece of software on your computer can begin requesting the images in a web page before it has received the html for the web page itself. I don't think you understand the concept of prefetching as applied to embedded objects (not links).

Two, pipelining. All modern browsers use pipelining, which severely limits the amount of handshaking that needs to be done to a server.

Pipelining is one optimization - this is another. There is some overlap in what latency they eliminate, but not 100%.

Re:Why you "can't really see"

[Anonymous Coward]

All modern browsers use pipelining, which severely limits the amount of handshaking that needs to be done to a server.

Yes, but in any given surfing session, you'll be handshaking with dozens or hundreds of different hosts. The TCP slow start has to start from scratch with each of them. Compare that with maintaining a single (or small pool of) connection(s) with Google.

I maintain my reservations about this being able to offer a significant boost to browsing, especially when contrasted to the major privacy intrusion.

You do realise that your ISP can see everything Google sees?

Re:Why you "can't really see"

[Anonymous Coward]

As a web developer, I've actually found that the best way to reduce page load time is highly aggressive caching - like telling the browser with headers to use a cached copy if it has one, and closing the connection.

Why would you close the connection? If you have a 304 Not Modified response for an HTML document, chances are the client will still want that connection around to get the stylesheets, images, etc. All you are doing by closing the connection is forcing the client to open a new connection.

Now what would make the web *really* fast would be some kind of CVS system, so the browser can say which version it has cached, and you send a diff to update to the most recent version.

See RFC 3229. It's not usually feasible because hosts and caches have to store revision histories, and it's rarely worth the complexity.

Re:Why you "can't really see"

[Anonymous Coward]

I don't want google caching or prefetching my content, if they want to serve my content then they are welcome to provide free hosting!

Caching is free hosting. And if you want to disable it, there are HTTP headers you can send so that they won't.

Re:Why you "can't really see"

[Trifthen]

I misspoke. I really meant to quit processing after sending the 304, which is basically the same thing in stuff like PHP.

So far as the diffs, I knew that would basically be impossible due to the currently dynamic nature of the web - I was just being flippant. ^_^ Just because something isn't feasable doesn't mean it's not better. Hehe.

I don't own a tin foil hat... but...

[Eric_Cartman_South_P]

Many people will use many, if not all, of Googles services. That means one single company can aggregate the data of a persons:

Website visits
Emails
Web Searches
Photos
Hard Disk Drive contents
Hard Disk Drive searches ...and now everything about every page they visit, cookies and all, since they are acting as a proxy!

Just the aggregation of this data on people who use all of their services could make their current income seam like pennies. This is the type of think that governments like a lot, not just large corporations. I know they have a "don't be evil" pholosophy (their words) but shit, even Skynet was nice at one point.

Re:I don't own a tin foil hat... but...

[Anonymous Coward]

I zip content when delivering to compatible browsers. It's no big deal. As for cacheing, many big ISPs use it to. We designed a system for shared access, condominiums, with SQUID in. When 500 Yuppies get up in the morning to read Dilbert and check their stocks, it makes a big difference.

Google taking over the world

[n2networksolutions]

I really think google has plans to take over the world. Jeremy Whittaker MCSE MCSA CCNA http://www.n2networksolutions.com/ Arizona Computer consulting

Movielink Yes, Showtime No

[wynand1004]

I have the same problem. I guess Showtime's geographical locator is better than movielink's.

Interesting...

[DaPhoenix]

During the install Peerguardian detected communication to/from:

Sitefinder (ip: 12.158.80.10:80) on local ports 3901->3906

Upon installing the google web accelerator, during web browsing my box immediately started trying to send information to/from

http://www.ctyme.com/ (ip address: 209.237.228.10:80) from ports 3978->4108

I have run peerguardian for a LONG time now and have never seen communication to either of these sites until installing the Google Web Accelerator.

Can anyone else confirm/deny?

Re:Interesting...

[Anonymous Coward]

I have it too! Anyone else see this? Also, upon rebooting Firefox refused to load!

Re:Interesting...

[Anonymous Coward]

It happens for me too! wtf is google up to?

Re:Smart. Scary.

[michrech]

I have it setup for my home email.

Horde is the framework

IMP is the mail

Kronolith is the calendar

Turba is your contact list

There are other modules. They integrate VERY nicely with eachother. I'm unsure if there is a place on the interweb where you can log in and demo the process, but if you are really nice to me, I might give you a temporary spot on my setup so you can see how it all looks.

---
telnet://sinep.gotdns.com -- it's a BBS for those of us that still like playing all the old games

The one that isn't

[Anonymous Coward]

evil.

Re:Smart. Scary.

[Atryn]

Great, so now they'll know what you search for, what you visit *and* what you are going to do in the future.

That's not very imaginative... no, they'll just put thing onto your calendar and make you do things in the future!!! And when you go back to your email to figure our what happened you'll see you sent an email RSVP for that event... And if you search it, you'll get a result with your name on the attendee list!

Now try and prove you aren't a terrorist if they say you are...

Re:Smart. Scary.

[mmarshall]

Online calender? Who would need that? What would really be cool is an online weatherbug thingy.

Re:Smart. Scary.

[Guspaz]

Nice try, but Yahoo! Mail works perfectly through the Google Web Accelerator. Better even.

Re:Smart. Scary.

[RevAaron]

Isn't there already a web version of Outlook?

Looks like there is. Either from Microsoft, or this outfit which isn't too expensive. $15 for the latter vs untold MILLIONS (of pesos) for the former. But still, it's there if you want it.

TrustRank related?

[chrima]

Is it possible that this service would simply be used to provide the 'seed' sites for a search oriented Trust Rank implementation? If it enabled them to build a consistent and well balanced set of weights for a wide range of sites, it would certainly pay for itself simply by making their search options that much more effective and comprehensive.

It could also add another way for them to calculate page rank. People going directly to a site without searching for it, would be able to add their 'vote' for the site without having to add a link to it from a website.

Just a couple of random thoughts.

Hell, let's just have GoogleNet

[birge]

Why do we even need an Internet? Why don't we all just have a connection to Google, sort of a star network with Google at the center. Google has a copy of everything anyway, 98% of the Internet is just a waste. The content people should just cut the bullshit and upload their stuff to Google, and then we'll all get it from them. For all the beauty of the original distributed Internet, the whole idea of everybody connected to everybody didn't anticipate our proclivity for monopolies and cheap hardrives. Except for P2P illegal file sharing, all of the connections that don't go between Google and someplace else are really pointless.

Re:Hell, let's just have GoogleNet

[The Cisco Kid]

I'm not sure if you are trying to be funny, sarcastic, or just plain dont have a clue. The "Web" and "The Internet" are entirely different things. One is an application that uses the connectivity provided by the other (But one of many applications, I will note)

Just becuase on your computer, your web browser is labeled "The Internet", doesnt make it so.

Re:Hell, let's just have GoogleNet

[birge]

I was being sarcastic, pointing out that despite your obviously correct understanding of the "web" vs the "internet", as things are evolving, the internet is mostly used for the web, and we're mostly all just connecting to the same few places. It's not the point of the internet as originally intended or engineered.

Re:Smart. Scary.

[incognitopoet]

I was actually kinda hoping they would create a way for me to enter my social security number, medical recards, and banking information. Now THAT would be convenient (and all that for free!!).

From the Terms and Conditions:

[gophish]

Intellectual Property You acknowledge that Google or third parties own all right, title and interest in and to Google Web Accelerator, portions thereof, or software provided through or in conjunction with Google Web Accelerator, including without limitation all Intellectual Property Rights. "Intellectual Property Rights" means any and all rights existing from time to time under patent law, copyright law, trade secret law, trademark law, unfair competition law, and any and all other proprietary rights, and any and all applications, renewals, extensions and restorations thereof, now or hereafter in force and effect worldwide. You agree not to modify, adapt, translate, prepare derivative works from, decompile, reverse engineer, disassemble or otherwise attempt to derive source code from Google Web Accelerator, or to extract significant portions of Google Web Accelerator's files for use in other applications. You also agree to not remove, obscure, or alter Google's or any third party's copyright notice, trademarks, or other proprietary rights notices affixed to or contained within or accessed in conjunction with or through Google Web Accelerator. -- emphasis mine

OT: Posting in order to get it recorded: May 16

[Anonymous Coward]

Call me crazy. Just had a premonition, and am posting it here to get it recorded prior to the event.

As follows:

In Jerusalem
Large Kill
Black Gas
May 16

That's all there is. Crazy. But wanted it out there in a place that is public, tracked, timestamped, etc.

Re:OT: Posting in order to get it recorded: May 16

[gimpimp]

erm, i'll just make a note of your IP, Mr Bin Laden

Would this work like those anonymiser sites?

[Anonymous Coward]

So would work web filters block inappropriate sites correctly still?

Re:Smart. Scary.

[CAIMLAS]

What, are you daft? This should be bloody fucking obvious to anyone that reads slashdot.

Two words: slashdot effect

Instead of having to manually create a site mirror for everything, google will do it automatically for you - if you have their accelerator plugin. That's pretty nice.

Usable in China?

[kristan]

Hello,

Does anyone know if the service will work in China? The web-page says that for the period of the beta it is "For users in North America and Europe" only; the service would be particularly useful in China where net connections to the rest of the world are notoriously slow. My suspicion is it won't work, as currently all connections to the Google cached pages are blocked, but if it did it would be incredibly useful.

--Kristan

Re:Smart. Scary.

[jp10558]

Isn't this the point of coral though? And you don't need any special plugin or extra program.

YESS! FASTER INTERNET!! GOOGLE PWN----oh.

[zoogies]

Just when I was jumping up, ELATED at this new find and development, OVERFLOWING with all sorts of lovely feelings for Google, FRUSTRATED by the overwhelming slowness of my connection, I read the catchline:

"Designed for Broadband."

-signed miserably,
a 56ker

Re:I keed! I keed!

[bmrh]

No privacy policy?

I clicked on the "Pricay Policy" link and saw this:
http://www.google.com/privacy.html

my experience

[Anonymous Coward]

For the most part, I only see about .1 second saved on each page.

The big time saver comes when I do a search on Google. I have my preferences set to return a 100 search results, and when I run a search now, it saves me up to 10 seconds a search.

GWA with Opera-and any browser with proxy support

[Anonymous Coward]

Can I run Google Web Accelerator on a browser other than Internet Explorer or Firefox?

WFM with Opera 8.0

Re:Smart. Scary.

[The Wicked Priest]

Blind faith is not a virtue.

Does not work in under Windows XP 64bit Edition

[Anonymous Coward]

google web accelerator does not work under Windows XP 64bit edition

Genius

[adolfojp]

Now Google can keep track of the websites we visit! :-D

difference

[etzel]

I wonder how long before GOOGLE becomes an official synonym of GAIN.

Great tool, but check this out, file storage gmail

[bigbinc]

This is a good tool, but we need google to implement internet filestorage, possibly through gmail. http://www.petitiononline.com/google45/petition.ht ml/

No benefit if your connection is fast?

[cgenman]

If one decides to visit a page, unless Google has a recent cache (not like a Google cache ever goes out-of-date) they'll have to re-download it from the webserver, compress it, and send it to your client. Unless you're on a painfully slow connection, chances are you could have downloaded it yourself basically instantaneously. And downloading a page to Google, diffing with Google's cache of your last recieved file, and sending you the update seems like it would just be faster to go to you directly. If they have a massive pipe, which I'm sure they do, and they can avoid outstripping the servers it feeds from, they might be able to save you a small percentage of time, based upon how much the diff requires, compressed, and how quickly google can download the page and perform the calculation, or if it's local copy is the latest.

Now, prefetching potential branch pages is cool, but I have yet to see an implementation of this tech that I'm impressed with.

All of this leads me to believe that dial-up users, the traditional market for such compress-and-prefetch schemes, will be the ones who actually benefit from this tech. That is, aside from those who side-benefit is better page searches on google with their suddenly dynamic DB.

Possible MAJOR Problem - not Privacy

[nigelthellama]

So, after using Mike's Ad-Blocker for the past year or so, I am kind of used to not seeing any ads on the web when I surf. One of the first thing I noticed right after I began surfing with the web accelerator was that all of those annoying ads were back (or at least most of them, especially the Flash ads - and I hate the flash ads...) So, after a quick experiment, the Google web accelerator is making its exit from my install of Firefox. Nice idea, but with Google acting as your proxy, you get the ads again... Oh, and I'm pretty sure that loading all the ads I'm not used to is actually slowing my web experience.

No catch!?

[Syre]

Everything Google does lately is designed to
monitor your surfing habits and email^h^h^h^h^h^h^h^h^h^h^h^h^h^h^h^h^h^h^h
make your life easier!

Re:No catch!?

[Syre]

But seriously, I just looked through all of Google's privacy policies dealing with this Web Accelerator, including the Google Web Accelerator Privacy Policy and the Google Privacy Policy and there are problems with them.

They say:

Google collects limited non-personally identifying information your browser makes available whenever you visit a website. This log information includes your Internet Protocol address, browser type, browser language, the date and time of your query and one or more cookies that may uniquely identify your browser.

The problem is that this information, when correlated with information from web sites you're using such as user names, passwords, etc. (all of which would be routed through their proxy and caches except for https information which goes through the proxy but not the caches), can tell them, or anyone else who has access, exactly who you are, where you surf and what you do.

Their privacy policies completely fail to address this issue.

Re:No catch!?

[Anonymous Coward]

How is this any different than what an ISP, say Road Runner, collects when we hit the net?

Re:No catch!?

[gad_zuki!]

Exactly. People think google is a "web search" company. Whatever that means. They are data miners. The toolbar tracked you also. The non-expiring google cookie is for tracking also. This is how they make money.

Re:No catch!?

[tzot]

This is important too (from their Terms And Conditions):

Changes to Terms and Conditions Google reserves the right to modify these Terms and Conditions from time to time in its sole discretion, without notice or liability to you. You agree to be bound by these Terms and Conditions, as modified. Please review the most current version of the Terms and Conditions from time to time, located at http://webaccelerator.google.com/eula (or such successor URL as Google may provide), so that you will be apprised of any changes.

Re:No catch!?

[Anonymous Coward]

Lately?
For us conspiracy theorists, google has ALWAYS fit perfectly as CIA branch. For every aspect of the company and its business.

Re:No catch!?

[Anonymous Coward]

The money isnt important all CIA/TIA is interested in is the information.

Re:No catch!?

[Infernal Device]

They probably won't tell your mom about the porn stash on your computer, so stop worrying.

Re:No catch!?

[that IT girl]

If they can monitor your surfing, they can pummel you with more ads generated based on where you go. Looks like yet another program designed to support adware/spyware.

Speed up your browser without the help of Google..

[b3rs3rk3r]

Some simple tweeks to your browser, or registry (if you are using IE). http://www.swedegoingberserk.com/2005/04/speeding- up-firefox.html Basicly you increase the number of connections your browser can use to download content, ie your browser can download everything on a page at once instaed of doing a couple items at a time. It really makes a differece on "busy" sites with lot of content. IE is set to defualt to use 2 measly connections...

haha

[Anonymous Coward]

HAHAAHA Totally Blows all those ISPs away that provide the same thing but charge you way to much! XD YAY FREEE!!

And what we have next...

[Semyazza]

Google TGP! (What else are they going to do with all that cached porn?)

Seems to Really Help Database Driven Sites

[monkeyman_67156]

From my rather breif experience with this, it seems to speed up sites that are mostly database driven. Sites like Interface Lift and anything like PHPBB or PHP-Nuke.

Since we're talking SF ...

[dbIII]

conveniently forget to update their cached version?

It would be a little like the MiniTruth*, wouldn't it?

No - more like a despair squid.

FBI Google nub

[D_Lehman(at)ISPAN.or]

Of course they would, because they "have" to. Let's not get in the mindset that this makes Google evil, just the FBI. ;)

But, let's not forget that they can tap your ISP directly if they want to. Use Google or not, if you live in the US, the FBI owns you... unless you use tools like Tor (http://tor.eff.org/ or Freenet (http://freenet.sourceforge.net/). In as far as unencrypted information, Google really gives the FBI no more advantage than if you used nothing.

What does have to be watched is how Google uses that information, and that is where the difference lies. But, just use https or turn if off if you both a) want to use it and b) don't want it to see "everything".

Google Desktop already reads your browser cache anyways... what, like you thought they couldn't already know what you were looking at if they really wanted to. ;)

So Now

[ArchAngel21x]

People who are paying a premium to their dial up ISP for this feature can now get it free from Google and save a few bucks every year. Nice.

oops

[D_Lehman(at)ISPAN.or]

The title of that post was supposed to be FBI > Google > nub... but /. editted it to make it look like I was calling you names. Sorry 'bout that. :(

Re:I keed! I keed!

[line.at.infinity]

Google web accelerator privacy policy page perhaps you meant to say something else.

Re:I keed! I keed!

[Anonymous Coward]

I submitted a few stories before and got rejected. No problem there. What really pissed me off is when a story rejected 10 days ago showed on the front page today using another source. What's the catch? /. don't like scoops?

Well, /. seems to always be the last one to talk about anything anyway.

not correct for ranking pages

[Anonymous Coward]

> Not only that, but it is also a beautiful solution to all the googlebombing, keyword-linking pages.

How is this a solution to googlebombing? If google will use it for ranking pages, rank-pumpers just need to create a bot or hire a person to click around in the necessary pages through google-cache. And so again google will not be able to tell 'good' pages appart from 'bad'.

I wouldnt be surprised if rank-pushers are already investigating google-cache and possible abuse possibilities.

Re:Smart. Scary.

[TheoGB]

Though bizarrely (and this isn't limited to me) both Horde and Neomail seem to fail after a certain point.

Luckily Squirrelmail is probably the best of the three packages and it has never failed (touch wood).

Why does it need a plug-in?

[gsasha]

For everything they might possibly want:

• Monitor everybody's traffic,
• Speed up sites,
• Use traffic for better page rank calculaiton,

They could just plain simple set up a proxy, and then ALL the brwsers, on ALL the OSs would be able to access it. I just don't get it.

Re:Why does it need a plug-in?

[anno1602]

Perversly, a lot of people have less trouble installing a plugin than configuring their web browsers correctly. Plus, a plugin is better for marketing.

Secondly, it seems that this will leverage (*grin*) Google's existing caching infrastructure. I don't know how well that maps to the standard proxy server model - it might not.

Re:Smart. Scary.

[HD Webdev]

Remember that they are already retrieving a huge amount of the web on a regular basis for their search engine. They may well have a "faster connection to the web server", since they might have already retrieved a copy of the resource beforehand. Also remember that, assuming it's not dynamic content, they can serve copies of the resource to hundreds of thousands of people without having to download it themselves.

I really can't stand assuming.

For instance, how does Google REALLY know if a page is dynamic or not. Many web programmers (as I do) disguise PHP pages as 'straight-up' html pages for various reasons. It could be to increase security, or to keep Google rankings high, or to deliver content optimized for a particular browser, or many other reasons.

What Google sees as a static web page could actually be a page that was optimized for the browser the server thought was 'looking' at that page. In that case, the majority of pages that Google displays would most likely be rendered as IE compatible ones.

It should improve their search results

[LadyLucky]

Currently they use static analysis to decide on popular websites - who is linking to who.

This information will give them dynamic data on who is actually browsing which websites. This will improve search quality, and make it even harder to googlebomb. It's great for us that won't install the Accelerator!

This can turn into a p2p CDN

[Diomidis Spinellis]

In the future the client could fetch material, not from Google's cache, but from an idle Web Accelerator client of a nearby computer. Thus Google can gain traction and momentum to deploy a Google-controlled peer-to-peer content delivery network. This can then be used for delivering content that really requires a lot of bandwdith, like video. By aggregating web requests from millions of users Google can not only create smarter prefetch algorithms, but also deliver better search results, and even suggest personalized interesting content. Think of it like the web's TiVo.

Google seems to have found a way to use network effects to gather the visiting habits of all the world's web surfers. The value of this data is priceless. Google's army of Computer Science PhDs is definitely not sitting idle.

Mouseover activation

[Mardak]

One thing I noticed when looking through apache logs on my computer, Google fetches files if I access my computer by IP.

But what's even more interesting is that when I was in a directory listing, the web accelerator would prefetch links that my mouse is hovering over. I followed my access log and indeed there were requests to my computer when I put my mouse over a link.

But then I'm not sure how accurate that seconds saved can be. It's reporting .5 seconds saved for each page request from apache on my own computer!

Re:Smart. Scary.

[leo_the_loin]

well it's quite simple. you install and test it out. If you like it, wonderful..go take a dump in a corner and laugh you head of at how smart you've been to install it If you don't like it..then uninstall the damn thing already..

SPOC - Single Point Of Control

[flibberdi]

Well... any Psychopathic controlfreak will get very exited (as in "oh...yeahh...BABY.....yesss"), this cant end well....

Better buy some stocks right now, sell them when RockerGates takes control, move to some obscure Iland and spend the rest of the days writing poetry.

I tried the Google Accelerator and guess what?

[HD Webdev]

It's pretty slick.

Anti-Advertisement software gets bypassed and the ads I haven't seen for years are now displayed for my 'enjoyment'.

Thanks, but no thanks. I'm off to uninstall this POS.

Open re-implementations?

[Nurgled]

I wonder how long it'll be before people create alternative implementations of the client portion of this. I think it'd be pretty neat, for example, to have it as a network-wide proxy on my server here so that everyone in the house can share the benefits of the local caching. We're all geeks, so we all load things like slashdot all the time. Of course, a ghetto solution for now would be to run it under Wine and arrange for Squid to forward requests to it.

It would also be nice to see other implementations of the server part. Sure, it won't be as fast as running it at Google because they've got a crawler cache and a fat pipe, but it would still be cool to have a bunch of different servers to choose from. ISPs could run such a server for their customers to speed up the DSL or Cable bit of the transfer, for example.

Re:Smart. Scary.

[imsabbel]

You know, microsoft was a a startup run by REAL firt generation people, and they had no problems getting evil even without the founders leaving.

Going public was just the first step, being billionaires will do the sufficient attitude adjustment to the founders.

No reason to be worried

[xtracto]

Every new product Google unveils its always the same, will Google become evil? what about my privacy? Will they stop being good?.

There is nothing to be worried about, the information google has is so SO big that it would be really difficult to concentrate in the granularity of one person, if you look close at the google service, it is all about tendencies the adds, the search (and that's all), it is called data mining.

And they will always be cautios in not pissing of the users because their customers (advertisers) would loss interest in them (less people to whom advertise) so, I think we can stop being paranoid there is no problem or at least I do not see it.

Re:Smart. Scary.

[hweimer]

Next they modify the data you receive to influence your opinion.

You mean like insert_coin? They run a proxy on proxy.odem.org:7007 which makes reading Slashdot quite an interesting experience.

Jeez

[Lord Kano]

I'm expecting Google soda pop and laundry soap next.

It seems that Google is getting its hands into everything these days. Search, Mail, News, Maps, Shopping, Web Accelleration...

When do we get 'Google The Flamethrower'?

LK

Re:I keed! I keed!

[jack_call]

isn't that 'slashdot got googled' in this case...

FUCK GOOGLE

[Anonymous Coward]

This is a PIECE OF SHIT!

When I installed it they disabled a bunch of my extensions in FireFox. I am STILL trying to get them all back. Thanks a lot you fucking assholes at Google! What gives you the right to disable my existing extensions. Had I known that you were going to do this I never would have installed your shitty software.

Google is (getting) EVIL!!!!

Re:FUCK GOOGLE

[Anonymous Coward]

What's more, they didn't disable the extensions the way Firefox typically disables them. Some of the extensions just don't work. They STILL don't work, even after the piece of shit Google shit is gone. You have to manually diable the (already disabled) extensions and then enable them again, its just that you have no idea which have been disabled.

I am not happy. Larry and Serge, go fuck yourselves!

Prefetching dangerous?

[nmg196]

Isn't prefetching a little dangerous?

What if google prefetches (for example) the advertiser sites which are linked to from it's own results pages? That would cause effective "click-thoughs" for sites which you haven't even clicked on, therefor causing them to waste their advertising budget on nothing more than a proxy. OK google has probably disabled this for it's own advert links, but what about everyone elses? And what about messing up the stats that the webmasters use? They might get the totally incorrect idea of what pages on their site are popular because Google has prefetched some links that people never normally click on, just because they're near the top or contain a certain keyword or something.

If usage of tools like Google Web Accellerator becomes widespread, I see many webmasters becoming particularly annoyed with them and possibly firewalling out requests from the Google Proxy.

Re:Prefetching dangerous?

[pbreit]

GWA only pre-fetches links that explicitly request to be pre-featched using {link rel="prefetch" href="/"}.

HOSTS file was semi-useless before...

[InvisiBill]

A .pac file is better suited to blocking ads by denying connections in a web browser. It lets you block/allow by URL rather than just hostname (i.e. block http://server.com/ads/ but still allow http://server.com/goodstuff/ through). It's a lot like the Ad-Block extension for Firefox, but not tied specifically to one browser. It works in all modern browsers and many other internet programs (email clients, etc.). See http://www.schooner.com.nyud.net:8090/~loverso/no- ads/ for details. As with Ad-Block, a few regular expressions in a no-ads.pac file will block most ads as well, even with no site-specific filters.

A no-ads.pac file will be slower than a HOSTS file (it uses Javascript inside the browser to process it, rather than being handled by the networking subsystem of the OS). However, any remotely-modern PC should have no problem with the tiny amount of extra processing needed. John LoVerso has claimed that he used to run it on a (sub-200MHz?) P1 and didn't notice any slowdown on that.

I do use my HOSTS file to block servers that I absolutely don't want any connections made to (in my browser, email, IM, or anything else). However, I found the HOSTS file to work very poorly for blocking ads in web pages, compared to these other methods.

scary

[CoolCat]

I usally say "yay google!" but this is getting scary. With all the data they now can collect, they will be like.. god? Have to use yahoo and msn more..

Could be improved alot

[Tarqwak]

Looking at tethereal capture while using the GWA showed that it doesn't compress HTTP request headers, and no encryption is used to talk to the GWA servers. It does send the X-Forwarded-For header so no real anonymizing is done.

I'll still prefer running encrypted OpenVPN tunnel over switched Ethernet to to a router that is connected to a Squid server that uses ISP proxy as cache_peer.

Also ping RTT to GWA European servers is ~73 ms (11 hops) while to ISP proxy ~19 ms (3 hops) that could count for something too.

Re:Smart. Scary.

[shutdown -p now]

First, they collect your search information. Next they collected your email. Now they collect your destination. You put it all together, that is quite a bit of information. What is next?

Kinda reminds me of Aquinas Hub from Deus Ex. The central point of the Internet - as long as you control it, you control the Net.

Re:Smart. Scary.

[Otto]

Knowledge is Power.

No, electricity is power.

Knowledge is not power until it is used in a way that helps one to exert more influence in a way that could not have been done without the knowledge.

Furthermore, I have yet for anybody to show me, despite repeated inquiries, how Google could gain large amounts of power or control over... well... anything, using knowledge they gained from me, in a way that is detrimental to me.

I mean, so they know I search for porn. Shit, everybody knows that! ;)

Puts on Tinfoil Hat...

[Shaper_pmp]

Much as I hate to come off as a member of the tinfoil hat brigade, Google is making me increasingly uneasy with the way they present and implement a lot of their offerings...

So far at least, Google has arguably successfully Done No Evil - they've offered a great search site, extended their great search system to the desktop, embedded it into browsers for convenience, offered webmail with unprecedented storage space and lovely features, and even revitalised the online advertising industry away from obnoxious graphical banners and popups towards relevant, discrete and unobtrusive text ads.

However, against this background of saintly behaviour, the potential for great evil lurks. Take the Google Search cookie not expiring until 2038 - there is no reason whatsoever for this, apart from to make it easy to track your searching habits. Of course, they could just do this by aggregating all queries that hit their servers, but that wouldn't uniquely identify you down to your specific machine, would it?

Take GMail - it's a lovely idea, and a lovely system, but it does mean that (theoretically), Google now has unfettered access to your entire inbox, and all the personal information therein. They also make a big deal of how you "never have to delete anything ever again" - handy for users maybe, but definitely handy if you're interested in data-mining vast volumes of personal information.

Google Desktop Search is a lovely tool (and very handy), but it does have an annoying (and downplayed) habit (IIRC) of by default echoing any local searches you make to Google, so it can return lists of "web" and "desktop" matches. Not such a big deal, unless you're searching your local machine for, oh, I dunno... company credit card details? Passwords? Rarely-used logins? Where you left the downloaded "Hot XXX teen sluts.mpeg"? Etc. Etc. Etc.

Now look at the Google Web Accelerator - not only your searches, but now every single page you visit (and even some you don't - are these differentiated between?) passes through Google's systems. Fair play to them for excluding HTTPS requests, but in all fairness they couldn't ever have got away with caching those as well anyway.

At this point, (assuming you use Google and don't take regular tinfoil-hat precautions like clearing cookies/deleting old mail/never searching your local machine for anything private/etc), Google potentially has access to:

• Your e-mail, including headers, full text and all your contacts.
• The text of every search you ever made, both on the web and on your local machine.
• The address and full text of every web page you ever visit.

Hmmm.

I have to stress here that I severely doubt there's any kind of deliberate conspiracy going on. For my money this is just a case of a bunch of overenthusiastic geeks with access to a huge database to mine, who are too busy having fun to write privacy policies because "we'd never do anything bad anyway, and people know that".

However, this still doesn't mean that it's a good thing - power corrupts, and Google now has one hell of a lot of power. Even if Larry, Serge et al stay true to their vision, Google's a public company now - it only takes the board to fire L&S and replace them with a marketing puppet and all of a sudden your trust in Google isn't worth shit - they hold all the cards, and they've got your entire life written on them.

In addition, this getting carried away with where they're going, and not listening to user-opinion is exactly the kind of attitude that is most publicly (and damagingly) exhibited by Microsoft. It's a small step from not taking five minutes to assuage people's concerns to not taking five seconds to even consider them. Both attitudes exhibit a certain "I know better than you" arrogance, one which tends to only get worse with time, and the more people start complaining about it, the worse it tends to get.

As I said, I severely doubt Google

Re:Puts on Tinfoil Hat...

[Anonymous Coward]

very nice comments. feel the same way

Re:Puts on Tinfoil Hat...

[Alioth]

As for HTTPS requests, it is simply not *possible* for them to cache them even if they wanted to. If you have a proxy, when getting an HTTPS page, your browser issues 'http connect www.example.com:443' to the proxy. All the proxy can do is pass the data unmolested since the data is encrypted end to end (i.e. server to your browser).

Re:Puts on Tinfoil Hat...

[SlimFlem]

I hear you man, but...how is using Google as a proxy any different than your normal ISP? Your ISP caches and logs all of your web browsing habits for sure, so what makes Google different. What if Google turned into an ISP as well and offered those types of services? Its no different. If you are doing things you don't want people to know about or find out about, its sort of hard these days unless you use someone else's computer while browsing.

Re:Puts on Tinfoil Hat...

[elemental23]

You're joking, right? Do you really think your ISP "caches and logs all of your web browsing habits"? Do you know the immense amounts of storage that would take for even a mid-sized ISP?

Not to mention that a lot of internet access these days is handled through third parties. My DSL line doesn't actually touch my ISP's network, it goes straight to SBC. Most dialup accounts I've used in the past actually used leased UUNET, PSI, Sprint, or Level 3 POPs, not equipment belonging to my actual ISP. The very most the average ISP could expect to log these days is DNS requests. The ISP I sysadmin for doesn't even do that, for any of our dialup, DSL, or T1/frame relay accounts.

Re:Puts on Tinfoil Hat...

[SlimFlem]

I was mainly thinking of DNS requests as I wrote that. In the end, I don't see a problem with Google's new tool. I think its great and do nothing to worry about. If you want to visit a site and not cache stuff like a bank, then update the preferences and stop worrying. Using their proxy is no different than someone else's in my opinion.

Not you elemental23, but someone also mentioned Google has complete access to people's email with gmail. And you don't think Yahoo, HotMail, LinuxMail, etc., etc. don't have this same access, come on.

"My DSL line doesn't actually touch my ISP's network, it goes straight to SBC." ...and you don't think SBC is doing some sort of logging for their DSL lines? Come on. Whoever provides the line service whether its dsl, cable, T-1, etc., logging is occuring at some level about something you have done. How do you think the RIAA and MPAA are able to identify and sue people over P2P usage? Logging my friend.

Re:Puts on Tinfoil Hat...

[Em Adespoton]

What is different is that Google is already a content aggregator, and makes money by selling targeted ads. I think your ISP would jeopardize its common carrier status if it tried to do either of those things with the data it logs.

Re:Puts on Tinfoil Hat...

[elemental23]

you don't think SBC is doing some sort of logging for their DSL lines?

Not really, no. See above re: massive amounts of storage and processing for logging this much traffic. There's really little to no benefit for them, as SBC isn't in the data mining business (whether Google is can be argued, but it's outside the scope of this comment).

How do you think the RIAA and MPAA are able to identify and sue people over P2P usage? Logging my friend.

They do that the same way I identified and whacked spammers during the years I spent in a large network abuse department: Logging of IP addresses and authentication information, not traffic. Traffic logs aren't need to ID someone. All the ISP needs to do is compare the IP address in question and the date/time of the incident with your RADIUS or DHCP logs. That gives you the account responsible, but no details on what they actually do on-line. For that you need logs or mail headers or similar from the reporting party.

Re:Puts on Tinfoil Hat...

[petermgreen]

if they have access to run arbitary code on your machine they can just install a new root cert in your browser

once thats done they can mitm your connections and your browser won't put up any warnings at all.

Re:Puts on Tinfoil Hat...

[Anonymous Coward]

Well, all of what you said is true - except for one detail about Google's corporate structure. Admittedly, it is a public company, and stockholders could convievably elect the Devil and his henchdemons to the board of directors. In fact, pure Evil could have control of 99% of the stock... but Serge and Page would still have total, absolute control. They possess unqiue "supershares." Basically, when they formed the corporation, they decided that their shares would be these supershares, which can outvote any number of normal shares. One supershare has more power than a million normal ones. Also, these shares can never belong to anyone other than Serge and Page - as soon as they sell such a share, it turns into a normal one. Funny, eh?

Re:Puts on Tinfoil Hat...

[Anonymous Coward]

They need data to crunch on SSL.

</tin foil hat>
<? wave(tin_foil_hat); ?>

Erm, no. No conspiracy, no chance. Move along. Nothing to see here.

Re:Puts on Tinfoil Hat...

[Shaper_pmp]

"Not you elemental23, but someone also mentioned Google has complete access to people's email with gmail. And you don't think Yahoo, HotMail, LinuxMail, etc., etc. don't have this same access, come on."

Granted, and I can deal with a company being able to read my mail. I can also deal with another separate company knowing what I search on, etc, etc.

But when it's one company that can read my e-mail, knows what I search on, knows what I search my local machine for, knows what I shop for, knows what map areas I look at (highly likely to be places I live/I'm visiting), and knows every page I browse on the net (and that's all they know currently), that makes me somewhat uncomfortable.

Here in the UK your mother's maiden name is a relatively useless piece of information, as is your National Insurance Number (SSN) or bank account number, but get any two of those pieces of information together, and identity-theft is relatively trivial.

In data-mining, the value of the information rises exponentially with the quantity of data, not linearly. It's all in the cross-referencing between separate data sources, not the data itself.

Re:Puts on Tinfoil Hat...

[Shaper_pmp]

If this is true, it soothes my paranoia hugely. Unless, of course, you believe www.googlewatch.com about Google's ties to the NSA, etc ;-)

However, I can't help but doubt it - wouldn't this unusual step have disasterously affected Google's share price in their IPO? I know nothing about stocks and shares, but surely non-voting shares are only worth a fraction of voting stock? And Google still seemed to be fighting potential buyers off with a stick, in spite of Wall St's best efforts.

Re:Puts on Tinfoil Hat...

[debiansid]

It is clearly mentioned in the Google Privacy Policy that they collect information about you (browser types, cookies, cached web pages). Infact it also warns that insecure submission of your data will be cached on their servers and that cookies will be saved and your browsing patterns are plain for them to see. It is upon you whether to trust them with it or not.

They should be given credit for being upfront about it and infact, educating people about how exactly their service will affect their privacy.

Re:Smart. Scary.

[bjelkeman]

When you use Google's services you are also exposing yourself to Googles future owners, whoever they may be. Ok so Google "isnt doing anything bad now", but who knows who has control of that data in the future.

When you sign a contract (or enter a deal, as you sort of do when you use Googles systems), you should never think that you are signing with the guy who is in charge now. Think of it as signing a contract with the next guy to own the contract. What is he going to do with the information... ?

Re:Smart. Scary.

[Shaper_pmp]

But the key thing here, you might say, is that God hasn't ever sent me 250 spam messages advertising "V1AGR4!!!!1!!!11!!! 4 J00!!!!!1!" in a single day.

Companies have.

(Nice attempt at a troll, BTW, but faaaar too obvious).

Re:Smart. Scary.

[adepali]

Considering that your opinion is vastly influenced by your web search results, they already do this.

Re:Smart. Scary.

[mst]

Next they modify the data you receive to influence your opinion.

In a sense they already do...

Does it help avoid corporate firewalls?

[Anonymous Coward]

For instance, Websense ?

Human-driven Googlebot Interface

[porneL]

It's useful for Google because:

• They know which search results people follow
• They find most active (interesting) sites and this might help them fight with spam and optimize ranking algorithm
• Sites cheating by sending garbage to Google's IPs are going to send garbage to end-users now
• Some files fetched for proxy may be fed to GoogleBot as well (fresh results)

It's useless for users becase:

• Transfer between User<->Google<->Site may not be faster than User<->Site except few pages that are anonymous and already cached by Google (and public cache only works for sites with no cookies, no browser sniffing, no no-cache headers)
• Non-IE browsers have pretty good HTTP/1.1 implementations (with pipelining) and there isn't much to improve here
• Browsers limit number of concurrent connections to proxy and this might throttle speed in some cases
• Only geekiest sites use <link rel="next/prefetch">

It's free but paid services are better

[Dexter.M]

Personally I think that Onspeed & Artera Turbo are MUCH better web accelerators (given that they have many servers across the globe and support image/ftp/mail compression..etc), but you just can't say no to Google now, can you?

I don't know though why Google designed this keeping broadband users in mind. Broadband by definition means > 256 kbps. I think that IS quite sufficient for ordinary web pages (given GWA's limited acceleration capabilities). I'm using dialup however, and the speed increase is quite noticable.

Although GWA is in a very nascent stage; once they add image compression and more global servers, only then it will become worthwile. But it will take quite some time for that to happen I guess.

For those who are looking for some SERIOUS web acceleration, I'd say stick with Onspeed or Artera Turbo... Until perhaps Google releases the final version of GWA.

** Security issue

[Anonymous Coward]

http://groups-beta.google.com/group/Google-Web-Acc elerator/browse_frm/thread/6c0290ab558802c4/a56327 4414ecafa7

Well it works...

[Anonymous Coward]

The links are prefetched as you idle on the page (only some pages) or as you hover your mouse over them. They load about 1 second faster (then again I have a 4mbps cable line, not dialup). I still don't think sending everything I do through google is a good idea. They already read my email and my searches....

I will stick with what I've got. If they had an upload accelerator then I might start seeding bittorrent's with google's help :)

Can you say "Marketscore" / "Netsetter" ?

[Anonymous Coward]

Funny how just recently Marketscore/Netsetter hit the news "'Researchware' watches where you click" Spyware or Researchware? and has been condemned for maybe tracking usernames and passwords for sites.

Now Google hits the news and is the darling when they explicitly state that they have access to track all your unencrypted pages and data, including email addresses, usernames, passwords (Google Web Accelerator Basics, part 5).

Quick survey for the interested:
How many sites do you log into that do not use a secure web page and a secure form submission?

Re:Smart. Scary.

[EvilGrin666]

Calender for firefox. Its based on Sunbird. (Or Sunbird is based on the Calender component, I'm not sure which came first).

Re:Smart. Scary.

[Thundertje]

Backpackit from the guys who brought you Basecamp HQ and Tadalist. Just launched a couple of days ago, it needs a little work but given their reputation I'd say it'll be the best EVAR! in a very short time :)

Re:Smart. Scary.

[Anonymous Coward]

They cant become annoying their owner forbids that. Google IS bigbrother under construction, Total Information Agency existed long before thye mistakedly put up the web page. Now I go watch some more Anthony J Hilder masterpieces.

Web accelerator?

[Anonymous Coward]

Well maybe - if only the pages had'nt been /.'ed seems url's are down and images broken.... How good was the product ???

Oh... I get it I'm not using it - That's why it's not working :o)

This service shouldn't need to exist

[baadger]

Another quality service from Google i'm sure. But this isn't really reducing transfer on the internet at large, it's 1xx% of what would be trasnferred with direct client->website transfers.

My ISP (NTL, UK) run transparent proxies, presumeably these proxies cache content and make my surfing experience better and reduce costs for my ISP (and therefore me, in theory). Shouldn't these proxies be introducing gzip/deflate compression (atleast for the most cached content) anyway?

I can see why this service exists, not all ISP's do cache, but on the ideal internet this just shouldn't need to exist.

If Google really wanted to be charitable and nice they could have approached ISP's directly and helped build up decent caching architecture. The fact of the matter is it isn't in their interests to do so.

Maybe instead of praising Google for doing this you should be hammering your ISP. The more reliant people become soley on Google for many web services the more painful it will be when something eventually pops and there is outcry. And there is a high chance that something will.

Unless of course you actually want to inform Google about your surfing habbits (which, I agree, you actually might. It may make surfing more useful etc depending what they do with this information - as many others have pointed out already)

Free Speech

[There is no cake]

Would is be possible for users in China to circumvent the blocking of site by using the download accelerator as a proxy?

Transparent cascading caching proxy accelorator?

[Halvard]

Hmmm. I wonder if a transparent caching proxy server benefits by having end users install this. Potentially less work for the transparent proxy. Time for some network analysis work.

Re:Smart. Scary.

[eille-la]

Hey the parent post is not funny, what is said is the fscking truth.
Not only adsense and search results is a way to do it, but now with this accelerator service, how many connection and transfer "errors" will be caused depending on what they want you to think?
Google IS as evil as microsoft, the karma they had is getting low.
Whatever you can think about google, they want to FUCK you and get MORE money, nothing else matter for them.
Get a look on www.fravia.com and remember to not only use google as you internet search engine.

The new international governement is made of corporations, and THEY DON'T want you free.

onspeed

[Anonymous Coward]

and the difference between this and onspeed is what exactly?

Spammer Reply

[TheKnave]

So - how are the spammers going to respond to this one? 100 billion AI Bots using GWA to find www.wesleycrusherdoesvenus.com?

MOD PARENT UP

[scenestar]

Running a bit nbehind, but this is one of the most insightfull posts ive seen for a while. To all the guys modding it down as a troll, I hope you live next to a chemical plant.

This one was a Google interview question

[gorbachev]

They have a question about a design of a system like this in their tech interview questions arsenal.

In the interview I was in the question was framed to address the problem of serving Google content to developing countries and other places with poor network connectivity. I wonder, if the purpose of the web accelerator is to make Google more accessible in those kind of environments than their (graphics-heavy) competitors.

They are selling YOU

[Anonymous Coward]

to advertisers. That is their main business, as is any media company that sells ads -- including your neighborhood theater.

Welcome to the road to the future. There is no exit.

Why not squid?

[ajs318]

Why not just apt-get update && apt-get install squid && /etc/init.d/squid start instead? Set your proxy in Konqueror to the IP address of the machine you installed Squid on, port 3128, for all protocols. Then write some Squid rules for blocking adverts.

Incomplete example: (note to put a \ before each ., because these are actually regular expressions)

acl adverts url_regex doubleclick
acl adverts url_regex advertising\.com
acl adverts url_regex fastclick\.net
acl adverts url_regex tribalfusion\.com
acl adverts url_regex falkag
acl adverts url_regex valueclick\.com
acl adverts url_regex burstnet\.com
acl adverts url_regex floppybank\.com
acl adverts url_regex freepush\.com
acl adverts url_regex atdmt\.com
acl adverts url_regex ads\.osdn\.com

http_access deny adverts

This will work right across your LAN, of course -- and beyond, if you configure your broadband router / IPtables to pass on port 3128 to the outside world. And only people with root access on the proxy server -- which is behind a locked door in your home -- will know what you're surfing.

Re:Smart. Scary.

[Kent Recal]

In that case, the majority of pages that Google displays would most likely be rendered as IE compatible ones.

See? Yet another reason to not design for a browser.
Stick to the standard.

Re:I keed! I keed!

[chrisnewbie]

that's what is scary! Everybody is raving about all those nice tools and saying that they are revolutionizing the internet but nobody ask themselves if it's going to turn like a BIG BROTHER project at some point! 'Watch out for the brain plug for Virtual Reality by Google'

Reminds of Rab Corbett's "How much is an internet"

[redeye69]

Direct link (real audio):
http://www.bawbag.com/cgi-bin/phonecalls.cgi?site= 1&filename=howmuchisaninternet.ram

Containing page:
http://www.bawbag.com/funnyphonecalls/index.php

Is this proxy a security hole?

[Anonymous Coward]

Assuming no firewall, can someone scan for the local google proxy port and use a connected machine as an anonymous proxy?

Re:Speed up your browser without the help of Googl

[BrowserCapsGuy]

And as a webmaster of some very popular sites I take steps to ensure that each visitor only gets one connection at a time to said websites regardless of what you've told your computer to do. It's things like this, plus WebAccelerator and other caching schemes that can slow an otherwise robust web server to a crawl.

Re:Smart. Scary.

[MarkusQ]

Next they modify the data you receive to influence your opinion.

Oh yeah? I just Googled for this very topic and there is absolutly no proof of that sort of thing. Ever. To Anyone. You'd think that if it were true somebody would have blogged about it. So you must just be parinoid.

--MarkusQ

Re:Smart. Scary.

[Anonymous Coward]

For instance, how does Google REALLY know if a page is dynamic or not.

The same way every other cache on the planet knows. By applying the normal HTTP rules - inspect the Cache-Control, Pragma and Expires headers (among others).

Caching is a very well-defined part of HTTP 1.1 and the mechanisms are described in detail in RFC 2616.

Many web programmers (as I do) disguise PHP pages as 'straight-up' html pages for various reasons.

Do you explicitly tell clients that your content is cachable by replacing the Cache-Control headers that PHP sends by default?

What Google sees as a static web page could actually be a page that was optimized for the browser the server thought was 'looking' at that page.

You can have that and still have caching. Look into the Vary header.

Hopefully..

[Snaller]

What is next?

Hopefully that the fix Googlegroups to be readable for all...

issues with accelerator & optimized firefox

[Greenrider]

I'm using a Moox-optimized version of Firefox that doesn't actually install itself normally through windows and works as a stand-alone .EXE.

The accelerator seems to work for me, but I can't get the toolbar to show up or see any stats. Anyone have any ideas?

Re : Google Web Accelerator

[geoff43230]

As a quick reminder to the agape observers, re : Google's latest offering, remember : as has been true with (almost) every previous other thing the company has offered, Google is not "forcing" you to download this. They are merely offering it. It won't even come up as a "suggested update".

Unalike, say Internet Explorer, if you don't like something from the company you can simply remove it from your system / desktop. Tracking non-personal information, if even not to any significant degree here, is not only by no means limited to Google, it has been largely - cookies aside, and even then - voluntary.

in a few years, we'll have a google-gw

[Anonymous Coward]

... where we surf through. google-virus and google-spam and google-ids protects us from the bad one on the net...

Re:Speed up your browser without the help of Googl

[Anonymous Coward]

I modded you down for the bad advice, because in reality what you say here can cause quite quickly a site to go down or become so freaking slow that hurts.

Re:Smart. Scary.

[Snaller]

I bet John Titor would tell you not to trust Google ;)

Google accelerator

[Anonymous Coward]

Seems to be little to no speed improvement on my system and connection. Takes up too much (about 14MB, more than quadruple any other resident non-OS processes) memory on my memory-poor system. About 5% of processor time when loading a webpage, that's fine. Then there's the privacy concerns....little or no improvement + excessive memory consumption + tinfoil hat = uninstalled.

Interferes with ActiveX

[unk1911]

The Google accelerator works great in terms of speeding up the web but
it seems to crash Internet Explorer when I try to play WorldWinner
games. Wonder if I can tell it ignore certain website, if that will
help matters.

Has anyone else experienced similar types of issues -- where ActiveX
sites malfunctioned with the accelerator?

--
http://unk1911.blogspot.com/

I block flash, not ads, no difference here.

[guidryp]

I have nothing against tastefull ads, that is what supports the net after all. But I have flashblock and anidisable installed, so all my pages are static. I have adblock installed, but so seldom use it that I didn't notice any difference when enabling the google acellerator, then again, I also notice no benefit.

Wickedness!

[theseeria]

0.3 seconds saved! w00t

Re:Smart. Scary.

[SunSaw]

Well, I was originally getting connect error messages while trying to connect to Yahoo! Mail but only when the Google Web Accelerator was active. Go figure!

Proxy of Proxy

[Spez]

I know someone posted something about "If my company decided i can't view some sites because they think its... unacceptable, will I be able to see them through the Google Web Accelerator?"

Since during the time i installed it and tried it, the post just vanished in the sea of posts, i'll just answer here

Where I work, we already have a proxy to limit internet access. They do not allow web sites that start with IP addresses and other stranges limitations. Because of this, i can't even play my favorite webgames. I also thought that with the GWA, i could.

But since there is no way to use a... proxy of proxy, or 2 level of proxies, and there is no place in the GWA to set up an other proxy, i could not. I think the only way would be to contact the ITs here and tell them to use GWA!

So, if you're using a proxy at work, don't even think about the GWA, it useless

Re:Smart. Scary.

[nmx]

See? Yet another reason to not design for a browser.
Stick to the standard.

Actually, he wasn't "designing for a browser," he was trying to make his page accessible to all browsers. He might generate different HTML depending on which browser is looking at it. Unfortunately, different browsers implement the standards differently, and sometimes you have to do hacks to get your pages to look the same in all browsers.

Re:Smart. Scary.

[FlimFlamboyant]

You know... we could just, dare I say ... NOT USE IT *gasp*

Seriously, though, what do you expect? They're either going to collect information that they can either sell to other companies or use for themselves, or they are going to charge you a service fee for using the software. It's just that simple.

What I don't get is why more ISPs haven't jumped on this type of technology long ago? I can remember wondering why ISPs weren't doing this very thing a good 10 years ago. Now Earthlink, Netzero, and no doubt other dial-up providers are FINALLY beginning to provide this. Hopefully this will expand to broadband providers at some point. And as far as privacy goes; well, every page you visit already passes through your ISP, so no harm done.

Affecting Firefox RSS extension.....

[bossj52]

Anyone else notice that the caching function of the web accelerator automatically refreshes headlines faster than every 30 minutes? Got banned by /. Besides that I think I might like this little app...

Re:Affecting Firefox RSS extension.....

[aristofeles]

Great... using Opera, and I just got banned too :(

this is only a test

[Anonymous Coward]

http://www.test.com/

Ooh shiny!

Re:this is only a test

[Anonymous Coward]

http://www.test.com/ craz

Re:I keed! I keed!

[uberjoe]

Hey I thought this was appledot?

Earthlinks Web Accellerator is WAY Better...

[Anonymous Coward]

who agrees?

GWA + /.RSS == WTF????

[Dexter.M]

I just got banned for 72 hours!!!!!

First of all, I'm on dialup and my IP changes everytime I log in.

Secondly, my Opera downloads RSS feeds only hourly.

Then I realised! Its the stupid GWA I've been using! Hardly a couple of hours and I'm already banned? Damn! So much for bad PR..

Could /. please exclude GWA from the RSS ban??

What is next? Hopefully folders for Gmail.

[glrotate]

I'm sure they'll figure it out one of these days. Labels just don't play well with POP.

Re:Smart. Scary.

[Kent Recal]

That's exactly what I was trying to say; don't do that.
Design your pages so they work in all browsers.

There's plenty of free knowlegde available online about how to properly design
accessible pages that work in all browsers. Do yourself a favor and learn to do it right.

Re:Smart. Scary. Idiots.

[Exocet]

From an old user to a new user:

What if, what if, what if? Who fucking cares. They don't have your SSN, your credit cards, etc. At best they have your name, IP and email address plus browsing habits and maybe interests.

So MAYBE in 10 years (when everything has completely changed, BTW) when the founders leave and the Big Evil Board of Directors does something... Evil... you will get some MORE spam. Boohoo!

Here are the facts: google doesn't spam you. They haven't spammed you. There's no goddamned obvious fact that you WILL be spammed or your information sold. However, when you shop at discounts-r-us.com or freeipod.com or every OTHER place ...yes, they will fucking spam you! So go bitch up a storm about those places and give google a rest. :P

Re:Speed up your browser without the help of Googl

[b3rs3rk3r]

I have a hard time understanding how cashing of your said popular website can bring it to a crawl? If google chooses to show a cashed version of your website, your website is not even hit, hence it will save you processing power.

Re:I keed! I keed!

[Anonymous Coward]

Don't worry about that. We all know your surfing habits already.

First you go to doggiesex.com, then you send an email to anonymousSixtyYearOldMan@nambla.com, then you have smallsex with a 35 year old cop pretending to be a 12 year-old girl.

Apple has a good solution

[SethJohnson]

Have you checked out iCal?

Seth

Oh! My brain.

[Anonymous Coward]

It hurts, it hurts.

Re:Smart. Scary.

[eille-la]

How can my parent post be overrated when it havent been rated before this first moderation?
I feel paranoid today and I accuse google of having robots to mod down the posts who talk against them.

Uh-Oh...

[webhed123]

Looks like the Google accelerator has resulted in my ip being banned from viewing slash rss feeds! Thats unfortunate!

Re:Smart. Scary.

[SenorChuck]

Personally, I'd be more worried about the Wayback Machine.. but maybe that's just me.

Re:I keed! I keed!

[Anonymous Coward]

Your computer is broadcasting an IP address!

Big problems over at Something Awful

[Tar Baby Says Nothin]

The SA goons server admins seem to have discovered some serious flaws in the program. http://forums.somethingawful.com/showthread.php?s= &threadid=1550986

Money and more business.

[ddebrito]

Remember the recent artilcle about gambling web site fighting off a denial of service attack? A web site could hire Google as verification engine. Let Google do the filtering of what are valid web page requests and what are zombie attacks. Perhaps Google will go into the ISP business. With Googles massive pipes, server farms and brains it could probably handle any denial of service attack.

but at least they warn you

[Jafa]

I installed GWA (goddamn, that's close to GWB, which ties in all kinds of other implications... :) and on the EULA clicky screen, in bright read letters, reads in part:
"This is not the usual Yada Yada. It is different from the Google Toolbar Yada Yada."

Awesome. So with google's weight behind it, we'll stop using the term EULA and start referring to "Yada yada agreements".
J

Is it worth it?

[whitehatlurker]

I tried it out for a while, I saved no time with it with several minutes of loading. (Not really sure how any of this would be measured.) I'm not sure if prefetching helped at all. Given my browsing patterns, I would doubt it - I tend to load up a set of tabs, then go through them and load pages in the background, if I want to see other information.

Most of the sites I visited were primarily text, and I was anticipating a good savings through compression. There did not seem to be any used, though I did not try to snoop the traffic between the local Google proxy and the main proxy server.

Seeing that I was using a remote proxy, several of the sites which give me special access based on my source address did not grant me that access. (Yeah, I could turn it off for those sites, but that seems like work.)

From my short test, I can't see any advantage for me to continue using this. (Of course if I had a nasty firewall censorship policy to bypass, I might change my mind. Or I might pick a different remote proxy.)

When I first saw this, I was reminded of other "web accelerators" which were basically spyware.

Your performance may vary, but I'll just continue with only Proxomitron between me and the web.

Re:Smart. Scary.

[HD Webdev]

Actually, he wasn't "designing for a browser," he was trying to make his page accessible to all browsers. He might generate different HTML depending on which browser is looking at it. Unfortunately, different browsers implement the standards differently, and sometimes you have to do hacks to get your pages to look the same in all browsers.

Yes. All of the browsers deviate from the standard. Some people are a bit anal about hacking up web pages until they are identical in all browsers. This isn't cost effective and only impresses 'gotta make them look identical in all browsers!' Zealots. I do it if I have the spare time, but usually I don't. I DO follow current standards and when I'm finished with a set of pages I check them in several browsers. If there is a problem with a few browsers and there isn't an relatively easy way to fix it, I make pages for those other browsers and the server 'invisibly' displays them. Making other pages with a few changes is much easier than often having to dumb down all of the pages 1 or more HTML versions. Also, if the client wants changes, I can easily do it without going through all of that other tedious process again.

There are many other reasons I like to identify browsers. For instance, I code so that 'get Firefox' images and links don't show up for Firefox users. Also, I can let people know if browser X is out of date.

I can select different content for different browsers/platforms if I have the information. For instance, it wouldn't be efficient for me to advertise a bunch of Windows software for someone who is browsing the site with an OS/X browser.

So what if many of us do this sort of thing for the above and dozens of other reasons? It's our coding and no one can tell us what to do. Not to mention, many major sites do the same thing. They customize content for depending on what they know or can guess about the user on the other end.

rss problem with /. and accel.

[demmer]

just for the record:
i get a "your headline reader has been banned" when using gwa and livebookmarks...

IP address, non-personally identifying?

[cbreaker]

Since when? RIAA seems to have no problem using IP addresses to sue people.

IP addresses are certainly personally identifying. ISP's know who owned what IP when, they log this kind of thing.

Now, google can't really get your name right NOW, but who's to say some ruling won't allow it later on, or make the information easily obtainable.

Not to mention, many of us retain our IP's for quite awhile, and I could see a company like Google not only offering users a PageRank system, but a UserRank system for web masters.

Re:Smart. Scary. Idiots.

[jparp]

interesting point.

However, the danger is that a devide is being created between those who have access to massive amounts of personal information, and those who do not.

This creates an enormous political, financial, and fundementaly social advantage for those who have this access. Corporations with this kind of access, can easilty spy on other corprations, steal there intellectual assets. A quick example: the stock market. By google or whoever, could very easily modify there investmetn stratagies by data mining for buseness realted discusions in personal gmails of there most successful clients.

With this web accelorator, google is putting themself in a prime position for all sorts of juicy man in the middle attacks.
http://en.wikipedia.org/wiki/Man_in_the_middle_att ack

Unless of course, people just shared all there information openly, the, the situation would be equalized.

Some guy named David Brin wrote about something this in a book "The Tranparent Society" But I haven't read it yet, so Im not sure how good it is. Fantastic topic though.
http://www.davidbrin.com/tschp1.html

Web-Based Outlook

[Acer500]

I agree. For performance reasons (we are in the process of getting more bandwidth), my entire division is now using Outlook Web Access (OWA), and it works very well, up to the point that some users don't even notice the difference at first glance.

But it only works for Microsoft-centric companies with Exchange, etc. See here for more info http://www.microsoft.com/exchange/evaluation/clien ts.mspx

Re:I keed! I keed!

[sbeashwar]

Sounds like the big "Internet" bypass; though it saves time seems a little "Evil" to me.

Re:Free Advertising for Penny Arcade

[Kaorimoch]

Maybe Google is us a favour from having our noodles overheat while reading their front page.

Maybe Its Useful

[Brother Grifter]

Maybe you can bypass all the crap your mega-corp employer is doing to monitor your internet surfing by just using this google proxy. Ok, google is monitoring you now, but what's worse. If you visit Ars Technica and browse forums all day, you're employer could catch on and repremand you. What if you're browing google all damn day long. You can claim you're doing research.

Of course, no one (well maybe no one), really browses the web all day, but sometimes when you have nothing to do, you'd like to be able to pass time by without drawing heat to yourself from those damn corporate types.

Any ideas if this is basically a free proxy that essentially masks what you're doing as just traffic to and from google? That's what I'm assuming it could do.

Download Mirror

[Anonymous Coward]

Since google pulled it, here is a download mirror http://home.insightbb.com/~google/