For those of you who make the effort to see Episode II on DLP and regular... here's something to do:
(1) Try to see the "regular" version on one of the few theatres getting a Kodak VisionPremier print. They are awesome and faaaaar exceed the DLP quality.
(2) Whether you see it on a VisionPremier print or not, the differences you should look for (and will notice if you do) is that the colors are slightly washed out (like a bad mpeg) or faded looking - check the deep blues and greens for instance, look for motion blur in the encoding, blockiness, line clarity (like looking for anti-aliasing artifacts and such on sharp contrast diagonal surfaces).
If you are in or near Baltimore [especially since there are no (inferior) DLP projectors in the area] I'd suggest the Senator Theatre in Baltimore just south of Towson (on York Road/Rt 45). Largest screen in the area (42' if memory serves), 16,500 watts of sound system, Dobly extended surround and more (you can FEEL when the big ships fly by). Amazing projection as well - family owned theatre, and the same guy running the projector since (what seems like) the beginning of time (or at least the last 63 years - and the quality of that experience shows). Their website is at The Senator Theatre. Heck, if you are within 100 miles of the place, it's worth seeing the movie there. I think tickets are going on sale there this Saturday... and there will probably be a line forming Friday night (and a drum circle and maybe a fire pit... if you play some sort of percussion or other fitting instrument, bring it along)... I live 3 blocks away, and the Episode I crowd and ticket campout was sh|tloads of fun. They do some kick ass stuff every Star Wars movie.
Another very real thing he will have doing a co-lo, like he said, are the bandwidth costs... they can (on an adult site for instance, like a friend soon learned) exceed 5 figures... using the bandwidth 2 T1's would require. These are at last November's hosting prices.
Very good reason NOT to co-lo unless someone is willing to allow it unmetered (rare) or with a high cap b4 pay-per-byte time.
...and the beauty of OS/2 Warp is that IBM just
released yet another refreshed version only a
few months ago, and Serenity Systems is
scheduled to release the next release of eComStation
(with many optional and included features) in the next couple of months.
We still use (exclusively - except on our MACs) OS/2 Warp 3 Connect, Warp 4, Warp Server Advanced 4, Warp Server for e-Business (WSeB), WSeB Nov 2000 release and WSeB Dec 2001 release - as well as eComStation Pro.
If it works (and works and works...) why change to something else or get rid of it?
> How many crackers do you think will try to crack OS/2 boxes? Not many.
Actually, the number is pretty high. And, it's not script kiddies either. It's real "pros" interested at a way in the largely-run-on-OS/2 banking and insurance networks.
To date number of vulnerabilities found? One. A certain attack on the first release of Warp Server for e-Business would make it shut down the TCP/IP stack and possibly crash the machine (no security or data jeopardized). It was fixed in 18 hours of being reported and the patch was made available that Saturday.
It's not like people dont try... they just dont succeeed... and very few companies run around reporting failed hack attempts. My servers get attacked around the clock with every means I can imagine and many I cant even find references of to break them. They were attacked with Code Red like virii long before it was released... (some of our servers are adult in nature, and it seems serious efforts are made to kill the competition... in talking with other small site owners, they notice the same problems... as ours is getting decent exporse, the attacks have occassionally hit 6 digits worth in a day. Including IRC spawned: DDOS, password and OS exploit hacks the likes of what Steve Gibson of grc.com reported months ago.)
What do I do? Nothing. I watch. I zip the log files. I laugh.
Well, not entirely true. Twice they managed to accumulate multiple gigabytes of httpd error logs in a day filling the log drive, which the server is then set to stop sending data out (in the event of). Changed that.
And I am changing my authentication engine from DominoGoWebserver's to something of my own devising (using a MySQL back end if it can keep up, and if not, then using a DB/2 back end).
There are still definitely things OS/2 does better - like be more secure. And play DiVX's - at least better than comparable Win__ hardware. Dunno about how Linux plays them...
All of Sears point of sale (including registers), video kiosks, information kiosks (that are not manufacturer provided), servers, order entry systems, inventory systems, etc run OS/2. Most are Warp 4 [3,000 departments (not machines) worth].
In addition, the building control consoles (fire, security, etc) of Sears Tower, Empire State Building and many other large buildings run OS/2.
You can find info here:
http://www.johnsoncontrols.com/esusa/st.htm
http://people.wiesbaden.netsurf.de/~meile/los2cl _9 76.html
Microchannel was designed for OS/2 was designed for microchannel... get it? Then enhanced microchannel came out... then it went to the high end RS/6000's and to this day, the PC bus in it's best incarnation cant truly touch it.
Thus, yes, OS/2 flies on microchannel. Any true multitasking task, or bus intensive task, or many needlessly CPU intensive task (ie: could be performed with better bus mastering and device to device communication) can be handled at 2-5 times the speed.
I'm looking for a few of the ancient Pentium MCA machines to run as OS/2 servers... amazingly solid, and amazingly fast... especially as servers where CPU speed isnt that relevant (unless you are running Windows).
Those enhanced MCA machines (bus design wise - not CPU wise, unless of course you are talking about the RS/6000's that came with enhanced MCA) are pretty cuick for today - not just "for their day". And one day, maybe PCI with all it's extensions and AGP, etc may catch up.
IBM cant get rid of OS/2 quite just yet even if they wanted to as some of their big or mid sized metal run "controllers" inside of them that run OS/2 (I think the z/Series for one, including it's previous incarnation). The "controller" is actually a very specialized PC that runs all the busses (hardware, memory, disk, networking, "BIOS", etc) of the machine... it sort of acts like a bus mastering chip and I/O controller. It runs OS/2 and has for ages.
I doubt IBM is yet ready (especially since the other OS alternatives out there have been as mature as they are long enough for the deed to have been done already) to switch those controllers from OS/2 to something else. The machines need 100% uptime (or at least IBM's guaranteeed 99.997%) so the controllers that make them run need to be neat little boxes that sit inside the machine, keep running and nobody needs to know about, running an OS that they have full control over to interact with their proprietary hardware and big metal OS.
I think they'll be keeping it around at least till the promised 2007 via maintenance, etc. And many OS/2 divisions in IBM seem to have decided it's worth more than just keeping it around... OS/2 just got fingerprint login recognition last week from IBM Germany who has been regulary cranking out OS/2 related things (and just recently started training seminars on it, and the new networking components... not things you'd expect for an OS you'd think they are trying to kill...)
Odin made more progress under OS/2 than Wine for the following reasons...
Odin incorporates Wine code - meaning (assuming they've ported all of it, Odin starts out equivalent with Wine)
OS/2 has hundreds of WinXX APIs supported for use as native calls. A very large base of the major ones according to IBM. Odin started off using them, replaced some of those calls with native code, etc.
Remapping WinAPI calls is thus easier with less code writing.
Innotek has a great deal of experience with such things, including on the graphical end, hence Flash Player 4 and 5 for OS/2 - which I personally think phenomenal - and according to them, there's not much of anything they cant get running well (if the proce is right).
Just a few reasons I can think of... there are others as well. A lot of it is on the odin.netlabs.org page if it's fully up again (either their directly or via links).
Do a search for restaurant guide on teoma and see what comes up... looks like they are suggesting that you use other's search engines to find restaurant guides... how neat! At least their search engine is set up to send you elsewhere (to their competition) when it doesnt have a clue.
The results were:
1. Webcrawler
WebCrawler.com Web News Photos Yellow Pages White Pages Search Zoom In Escape! airfare discount travel hotels Autos...
http://www.webcrawler.com/
[More results like this]
2. Yahoo!
Yahoo Yahoo! Personals find the one for you Click Here ! Yahoo! Messenger 5.0 download now! advanced search Y! Shopping - Computers ? Video Games...
http://www.yahoo.com/
3. HotBot
HotBot Search Smarter Look for: all the words any of the words exact phrase the page title the person links to this URL boolean phrase Date:...
http://www.hotbot.com/
[More results like this]
5. Google
Happy Holidays from Google Web Images Groups Directory ? Advanced Search ? Preferences ? Language Tools Advertise with Us - Add...
http://www.google.com/
6. Lycos
Lycos Home Site Map My Lycos TOOLS Chat Clubs Downloads Email/IM Horoscopes Maps MP3s Multimedia Personals Photos Stocks Translate WAP/SMS...
http://www.lycos.com/
7. AltaVista - The Search Company
AltaVista provides the most comprehensive search experience on the Web!... AltaVista Try your search in: Images ? Video ? MP3/Audio ? News...
http://www.altavista.com/
8. AllTheWeb.com
AllTheWeb Search All The Web, All the Time :... Help | Customize | Advanced Search Any language Afrikaans Albanian Arabic Basque Byelorussian...
http://www.alltheweb.com/
9. Dogpile. All results, no mess.
Dogpile gives you more result with less effort. Find the best results on the Web with the help of more than 15 of the top search engines in the...
http://www.dogpile.com/
[More results like this]
10. Direct Hit. One Search Engine. Millions of Minds.
direct hit direct hit November 26, 2001 Search for... web sites categories...or ask a question Tragedy in the US: News and Resources...
http://www.directhit.com/
[More results like this]
Neat, huh?
Besides often irrelevant results, or results such as above, it's slower, and still has bugs that need to be worked out (clicking certain NEXT navigation will put you at results 1-10 of who knows what or where - like click the >> at the end of the 1 2 3... section - leaving no other page options other than NEXT >> after that...).
I see it getting slower and slower if it gets adopted by any decent amount of people, and then eventually either in need of a server farm of immense proportions or lots of IIS's famous "Server too busy" errors...
Robert
FoodPlaces.com - The Largest Restaurant Guide in the Nation!
There is a lot more to this than everyone seems to be discussing (unless I missed a post or two...)
First, this is at least the second time Yahoo has done this...
Second, and most importantly, once the damage is done, it's permanent. Some of the settings allow Yahoo to sell and/or give away your email address, so going back to Yahoo and changing the settings back is an "already too late" action. The damage is done, and who knows how many companies your email address has been given and sold to? Yahoo refused to discuss that issue in the few emails they and I sent back and forth.
Third, this so far has happened with each "major" system change (3 times according to a friend of mine who has a Yahoo ID longer than I - I only know of 2 that I caught on to)... so how many more times will they do this? Hmmmm... every time they need to make a few more bucks and sell their entire subscriber list? Perhaps...
Fourth, why hasnt anyone sued them yet? This spells class action to me...
>It doesnt NEED to. A proper L4 balanced proxy
>implementation will select the proxy member from
>hashing enabled based on source/dest port,
>source/dest IP Address AND on the URL. This will
>ensure that connections to www.foo.com from your
>source will always go the to the same server.
>
Which is NOT what AOL nor MSN nor WebTV do. Nor do many other large ISPs.
My argument was never that it cannot be done, nor that it should not be done. My argument is the big guys dont do it, even though they can and should, thus wasting the bandwidth and then some that they'd like to claim they are saving.
"you idiot"... never ceases to amaze me when the kids come out to play on their parent's computers...
Oh well, let me give YOU a CLUE then. AOL is arguably the largest ISP in the nation. MSN/WebTV conglomerate are trying to fight their way up to that position. So, that's a whole LOT of misconfigured proxy servers. You have a clue yet? Add up all the rest of the ISPs and compare their userbase to those 3 combined and you have less - or so they want us to believe from their marketing - which if that's the case, means your measly efforts are just that in comparison to their poor implementations.
Go get a clue and go grow up.
Why dont you come back when you can post like an adult instead of resorting to childish name calling?
This isnt DoS'ing their server... once every 5 minutes, a simple GET garbage or other command wont do anything major... if a whopping 10-20 bytes kills their proxy, it shouldnt be running anyway.
Besides, what right do they have spying on their customers activities anyway? Let's see, At Home, AOL, Verizon and Qwest have been sued for such as that was (at least one of) their apparent reason(s) for proxy servers... (At Home's replacement in the ComCast world was threatened with a suit for similar reasons if they didnt stop - which they did).
So then, for the person in question, it should be as simple as to bounce a malformed request back TO his machine to disable the router from trying to proxy his outgoing httpd traffic... so your answer should still apply!
(Remember, he is trying to serve web pages if I remember correctly... then again, been a long time since i had a cup of coffee... but I think that was the original scenario...)
I think you may have just solved his problem. A simple script, bounce some bad requests to his web server for whatever address (or each) that he is hosting on it and viola!
They probably havent set up their filters or IP allow list properly yet then. We had to be added to our ISP's list (not really theirs) as they license out someone else's news server... if they've changed over to another server, highly likely they need to update their files to allow all the gte customers (or the subsets they have forgotten) back on. Possibly, they had been blocked in the past from access to the verizon servers due to issues, but now that verizon owns it, needs to be fixed... best guess I have... and of course with verizon, any guess is probably the best you'll get, no matter how off the wall since it is almost always more than you will get out of them...:-(
Best of luck to you...
Robert
www.WebBinaries.com
Try www.WebBinaries.com or Re:On a related note...
on
Usenet Encoding: yEnc
·
· Score: 1
or www.Pictureview.com (I'd suggest WebBinaries...)
You are probably being blocked from it... I could access it, though since I am not in the correct IP pool, the 200 (OK) message states I need a login (authinfo stuff)
200 This connection requires username/password authentication (Twister v2.0.5.1)
You post something they dont like or that someone complained about?
It's a direct grab from our main web server. The firewall is part of our TCP/IP stack and we only serve about 100,000 pages and 7 million images a day, so we dont do load balancing or use proxy servers as our load on our server is nonexistent.
Our (html) pages each have 20 thumbnail images linked to 20 full sized images, and 14 miscellaneous images, of which 4 are repeated (once on top, once on bottom), and even for the same request, from the same server, their proxies will re-request that document. WebTV's and MSN's proxy servers will often do that off and on as well (depending on how many serve a request). AOL will do that 65% of the time (based on a pretty large log sampling) as that many users will end up being served via multiple proxy servers that do not check each other for the documents.
For us, between them, they waste an additional 25% bandwidth. Add to that, versions of IE that try to "pre-cache" (dont know what else to call it) wrong documents (as versions of their proxy servers do as well), and we've calculated 32% wasted bandwidth (in the "pre-cache" case, lets say you are looking at/sportsimages/images/picture12.jpg from document/sportsimages/pictures0001.html - their crap will also try to retrieve index.html from that same directory and the main directory. This issue at least seems to be on the decline... but poorly set up proxy pools seem to be increasing).
Gee, I am sorry... I thought that since I specified two examples, it was obvious that what I was saying is that (a) both in the examples I cited and (b) in most similar cases, "there is no mechanism (in use by those ISPs) to check to see if proxy 12 has the document the user is requesting that now proxy23 needs to get."
I apologize, as I thought it was clear. Fact is, no such mechanism is in use by AOL, or formerly AtHome, and in all probability other ISP's. To name a few, the virtually only ISP in Australia under any of it's 3 major names, btinternet, webtv, and on and on and on and on...
You claim that proxy servers help alleviate bandwidth usage at ISP's? I am not sure how you come up with this figure.
The honest truth is in numerous large ISPs utilizing proxy servers, they INCREASE bandwidth usage. The reason quite simply is (for instance, in the case of AOL), a user is often served by a half dozen to 2 dozen proxy servers for a single request. There is no mechanism to check to see if proxy12 has the document the user is requesting that now proxy23 needs to get. Thus, often it increases traffic because the proxy servers request documents that were already cached by other proxy servers.
I've done tests with AOL and the former AtHome Network and reloaded the same page and watch a different set of proxy servers [or the same ones retrieving different documents (read: images) from the same page] meaning that roughly 8 out of 10 requests for an image or document, it was being re-retrieved, even though it was already cached by an earlier request by that same user (but via a different one of their ISPs proxy servers).
So... explain how that makes more of their bandwidth available... not just does it decrease their available bandwidth, but mine (for my servers) as well.
They dont provide better service, they dont save anyone money (except the ISP who sells the demographics - which, btw, is one of two reasons why AOL still does it - the other is image compression - which anyone with any AOL intelligence turns off ASAP). AOL has already been in lawsuits about this, so it's not guesswork on my part. Including with a company that was working with doubleclick to use all the data that they (two) were accumulating.
First, on Slashdot, I'd have presumed that a regular poster would understand the difference between Napster and Usenet... if not, perhaps this one difference alone will clarify some things. Napster allows peer to peer access to people's machines to share files, while Usenet, for it's functioning, allows 100 million users (or more) worldwide, to post to over 15,000 servers in a variety of "newsgroups" ranging from 1 to 86,000 carried per server. Every server that carries a group generally replicates (or is supposed to) all posts being posted to that group regardless of point of origin (with variances due to installed spam filters, etc).
Because of that, it makes it impossible for Usenet to be policed by online service providers for either copyrighted or illegal content - and even in the event of copyrighted content, a number of copyrighted posts are in Usenet because the owners want them there. The sheer size of Usenet and the inability to distinguish between legitimate posts and intellectual property theft is thus an impossible task.
The DMCA TITLE II in this matter DOES indeed protect both the OSP (including ISP, which is a subset of OSP) AND the copyright holder. The OSP is required to take down the material after sufficient notification has been given (within a reasonable time period), while the copyright holder can sue the original poster for damages.
The copyright owner CAN sue the OSP for contributory damges - IF - the OSP does not honor the (if valid) takedown request.
Seems perfectly reasonable to me. With millions of posts a day, in 80,000+ groups coming from 15,000+ servers, I dont see any other reasonable method, so in this aspect at least, I think the DMCA is adequate to the task.
So, how is the copyright holder getting screwed by this? Because AOL-TW is a big bad company that should be sued without valid reason? I dont think so. Find valid reasons to slam AOL - dont try to rewrite good sections of the law just so you can use it against them.
Oh, but wait! How can the poor copyright owner catch the original poster? Gee, that's an easy one, but most people dont understand enough about the Internet to know that yet... here's some enlightenment (and now Verizon will hate me even more - though not as much as most of the world hates them I am sure)... It's called a circuit trace. Not a traceroute... All you need to know is what IP address the post originated from, and what time, contact the ISP, have them contact their NOC or Verizon to pull the telephone company circuit logs that they will pretend they know nothing about. Those logs are of the Telephone switching network activity that maps the "digital" data for Internet communication (and related TCPIP traffic) over the telco network to a physical telco circuit mounted to the person's house or apartment. When I worked for AOL's backbone provider (they dont own their network - nor does MSN - the same big company does), we'd acquire physical addresses of serious abusers in about 5 minutes through this method - but then again, the company I worked for had more dialup numbers than the next 5 ISP's combined - so Verizon was more than happy to work with them.
If it's a cable modem, the cable companies maintain similar logs which allow similar traces.
In some cases, if the poster is attempting to hide their address (spoofing, etc) that is a little more difficult, but since the telco has to know where to send the packets regardless of the IP address spoofing, they still have logs of where everything came from. Even though they claim otherwise.
If it's non-US, then yes, the copyright holder is SOL... but that's still not a reason to take it out on AOL.
Considering the description, it seems like the person in question is a mass advertiser who went to the cost of purchasing an opt-in email list, didnt receive what they were promised and is suing because of that - that would mean "spammer" is a far from applicable term for them. There are many opt-in email list companies online - the problem is there is no standard (and possibly no good way?) of determining which company is legit, or which email addresses are legit. Many, allow anyone to submit an address.
Ack... dont misunderstand me - laserprinter motors are not the problem - their surge draw is. Laser printers use "massive" (1 foot long or longer) halogen bulbs to heat the fuser roller from inside the roller, and they draw a lot of power during their heatup cycle.
The AC units have a high surge when the compressor is turned on.
Don't forget... if you plug anything with a motor into a UPS, better make sure it is rated for it and that it does not void the warranty. Laser printers, and air conditioners (compressor motors) draw waaaay too much surge current for most run of the mill SOHO UPS's and even for some mid sized ones...
Slashdot Mirror
(1) Try to see the "regular" version on one of the few theatres getting a Kodak VisionPremier print. They are awesome and faaaaar exceed the DLP quality.
(2) Whether you see it on a VisionPremier print or not, the differences you should look for (and will notice if you do) is that the colors are slightly washed out (like a bad mpeg) or faded looking - check the deep blues and greens for instance, look for motion blur in the encoding, blockiness, line clarity (like looking for anti-aliasing artifacts and such on sharp contrast diagonal surfaces).
If you are in or near Baltimore [especially since there are no (inferior) DLP projectors in the area] I'd suggest the Senator Theatre in Baltimore just south of Towson (on York Road/Rt 45). Largest screen in the area (42' if memory serves), 16,500 watts of sound system, Dobly extended surround and more (you can FEEL when the big ships fly by). Amazing projection as well - family owned theatre, and the same guy running the projector since (what seems like) the beginning of time (or at least the last 63 years - and the quality of that experience shows). Their website is at The Senator Theatre. Heck, if you are within 100 miles of the place, it's worth seeing the movie there. I think tickets are going on sale there this Saturday... and there will probably be a line forming Friday night (and a drum circle and maybe a fire pit... if you play some sort of percussion or other fitting instrument, bring it along)... I live 3 blocks away, and the Episode I crowd and ticket campout was sh|tloads of fun. They do some kick ass stuff every Star Wars movie.
Rob
www.Hyperforce.com
Another very real thing he will have doing a co-lo, like he said, are the bandwidth costs... they can (on an adult site for instance, like a friend soon learned) exceed 5 figures... using the bandwidth 2 T1's would require. These are at last November's hosting prices. Very good reason NOT to co-lo unless someone is willing to allow it unmetered (rare) or with a high cap b4 pay-per-byte time.
We still use (exclusively - except on our MACs) OS/2 Warp 3 Connect, Warp 4, Warp Server Advanced 4, Warp Server for e-Business (WSeB), WSeB Nov 2000 release and WSeB Dec 2001 release - as well as eComStation Pro.
If it works (and works and works...) why change to something else or get rid of it?
Robert
Actually, the number is pretty high. And, it's not script kiddies either. It's real "pros" interested at a way in the largely-run-on-OS/2 banking and insurance networks.
To date number of vulnerabilities found? One. A certain attack on the first release of Warp Server for e-Business would make it shut down the TCP/IP stack and possibly crash the machine (no security or data jeopardized). It was fixed in 18 hours of being reported and the patch was made available that Saturday.
It's not like people dont try... they just dont succeeed... and very few companies run around reporting failed hack attempts. My servers get attacked around the clock with every means I can imagine and many I cant even find references of to break them. They were attacked with Code Red like virii long before it was released... (some of our servers are adult in nature, and it seems serious efforts are made to kill the competition... in talking with other small site owners, they notice the same problems... as ours is getting decent exporse, the attacks have occassionally hit 6 digits worth in a day. Including IRC spawned: DDOS, password and OS exploit hacks the likes of what Steve Gibson of grc.com reported months ago.)
What do I do? Nothing. I watch. I zip the log files. I laugh.
Well, not entirely true. Twice they managed to accumulate multiple gigabytes of httpd error logs in a day filling the log drive, which the server is then set to stop sending data out (in the event of). Changed that.
And I am changing my authentication engine from DominoGoWebserver's to something of my own devising (using a MySQL back end if it can keep up, and if not, then using a DB/2 back end).
There are still definitely things OS/2 does better - like be more secure. And play DiVX's - at least better than comparable Win__ hardware. Dunno about how Linux plays them...
Robert
In addition, the building control consoles (fire, security, etc) of Sears Tower, Empire State Building and many other large buildings run OS/2.
You can find info here:l _9 76.html
http://www.johnsoncontrols.com/esusa/st.htm
http://people.wiesbaden.netsurf.de/~meile/los2c
Robert
Thus, yes, OS/2 flies on microchannel. Any true multitasking task, or bus intensive task, or many needlessly CPU intensive task (ie: could be performed with better bus mastering and device to device communication) can be handled at 2-5 times the speed.
I'm looking for a few of the ancient Pentium MCA machines to run as OS/2 servers... amazingly solid, and amazingly fast... especially as servers where CPU speed isnt that relevant (unless you are running Windows).
Those enhanced MCA machines (bus design wise - not CPU wise, unless of course you are talking about the RS/6000's that came with enhanced MCA) are pretty cuick for today - not just "for their day". And one day, maybe PCI with all it's extensions and AGP, etc may catch up.
Robert
I doubt IBM is yet ready (especially since the other OS alternatives out there have been as mature as they are long enough for the deed to have been done already) to switch those controllers from OS/2 to something else. The machines need 100% uptime (or at least IBM's guaranteeed 99.997%) so the controllers that make them run need to be neat little boxes that sit inside the machine, keep running and nobody needs to know about, running an OS that they have full control over to interact with their proprietary hardware and big metal OS.
I think they'll be keeping it around at least till the promised 2007 via maintenance, etc. And many OS/2 divisions in IBM seem to have decided it's worth more than just keeping it around... OS/2 just got fingerprint login recognition last week from IBM Germany who has been regulary cranking out OS/2 related things (and just recently started training seminars on it, and the new networking components... not things you'd expect for an OS you'd think they are trying to kill...)
Just my 1/2 pence
Robert
- Odin incorporates Wine code - meaning (assuming they've ported all of it, Odin starts out equivalent with Wine)
- OS/2 has hundreds of WinXX APIs supported for use as native calls. A very large base of the major ones according to IBM. Odin started off using them, replaced some of those calls with native code, etc.
Remapping WinAPI calls is thus easier with less code writing.
- Innotek has a great deal of experience with such things, including on the graphical end, hence Flash Player 4 and 5 for OS/2 - which I personally think phenomenal - and according to them, there's not much of anything they cant get running well (if the proce is right).
Just a few reasons I can think of... there are others as well. A lot of it is on the odin.netlabs.org page if it's fully up again (either their directly or via links).Robert
The results were:
1. Webcrawler
WebCrawler.com Web News Photos Yellow Pages White Pages Search Zoom In Escape! airfare discount travel hotels Autos...
http://www.webcrawler.com/
[More results like this]
2. Yahoo!
Yahoo Yahoo! Personals find the one for you Click Here ! Yahoo! Messenger 5.0 download now! advanced search Y! Shopping - Computers ? Video Games...
http://www.yahoo.com/
3. HotBot
HotBot Search Smarter Look for: all the words any of the words exact phrase the page title the person links to this URL boolean phrase Date:...
http://www.hotbot.com/
[More results like this]
4. http://www.excite.com/
http://www.excite.com/...
http://www.excite.com/
5. Google
Happy Holidays from Google Web Images Groups Directory ? Advanced Search ? Preferences ? Language Tools Advertise with Us - Add...
http://www.google.com/
6. Lycos
Lycos Home Site Map My Lycos TOOLS Chat Clubs Downloads Email/IM Horoscopes Maps MP3s Multimedia Personals Photos Stocks Translate WAP/SMS...
http://www.lycos.com/
7. AltaVista - The Search Company ... AltaVista Try your search in: Images ? Video ? MP3/Audio ? News...
AltaVista provides the most comprehensive search experience on the Web!
http://www.altavista.com/
8. AllTheWeb.com ... Help | Customize | Advanced Search Any language Afrikaans Albanian Arabic Basque Byelorussian...
AllTheWeb Search All The Web, All the Time :
http://www.alltheweb.com/
9. Dogpile. All results, no mess.
Dogpile gives you more result with less effort. Find the best results on the Web with the help of more than 15 of the top search engines in the...
http://www.dogpile.com/
[More results like this]
10. Direct Hit. One Search Engine. Millions of Minds. ...or ask a question Tragedy in the US: News and Resources...
direct hit direct hit November 26, 2001 Search for... web sites categories
http://www.directhit.com/
[More results like this] Neat, huh?
Besides often irrelevant results, or results such as above, it's slower, and still has bugs that need to be worked out (clicking certain NEXT navigation will put you at results 1-10 of who knows what or where - like click the >> at the end of the 1 2 3 ... section - leaving no other page options other than NEXT >> after that...).
I see it getting slower and slower if it gets adopted by any decent amount of people, and then eventually either in need of a server farm of immense proportions or lots of IIS's famous "Server too busy" errors...
Robert FoodPlaces.com - The Largest Restaurant Guide in the Nation!
First, this is at least the second time Yahoo has done this...
Second, and most importantly, once the damage is done, it's permanent. Some of the settings allow Yahoo to sell and/or give away your email address, so going back to Yahoo and changing the settings back is an "already too late" action. The damage is done, and who knows how many companies your email address has been given and sold to? Yahoo refused to discuss that issue in the few emails they and I sent back and forth.
Third, this so far has happened with each "major" system change (3 times according to a friend of mine who has a Yahoo ID longer than I - I only know of 2 that I caught on to)... so how many more times will they do this? Hmmmm... every time they need to make a few more bucks and sell their entire subscriber list? Perhaps...
Fourth, why hasnt anyone sued them yet? This spells class action to me...
Robert
>It doesnt NEED to. A proper L4 balanced proxy
>implementation will select the proxy member from
>hashing enabled based on source/dest port,
>source/dest IP Address AND on the URL. This will
>ensure that connections to www.foo.com from your
>source will always go the to the same server.
>
Which is NOT what AOL nor MSN nor WebTV do. Nor do many other large ISPs.
My argument was never that it cannot be done, nor that it should not be done. My argument is the big guys dont do it, even though they can and should, thus wasting the bandwidth and then some that they'd like to claim they are saving.
Oh well, let me give YOU a CLUE then. AOL is arguably the largest ISP in the nation. MSN/WebTV conglomerate are trying to fight their way up to that position. So, that's a whole LOT of misconfigured proxy servers. You have a clue yet? Add up all the rest of the ISPs and compare their userbase to those 3 combined and you have less - or so they want us to believe from their marketing - which if that's the case, means your measly efforts are just that in comparison to their poor implementations.
Go get a clue and go grow up.
Why dont you come back when you can post like an adult instead of resorting to childish name calling?
Besides, what right do they have spying on their customers activities anyway? Let's see, At Home, AOL, Verizon and Qwest have been sued for such as that was (at least one of) their apparent reason(s) for proxy servers... (At Home's replacement in the ComCast world was threatened with a suit for similar reasons if they didnt stop - which they did).
(Remember, he is trying to serve web pages if I remember correctly... then again, been a long time since i had a cup of coffee... but I think that was the original scenario...)
I think you may have just solved his problem. A simple script, bounce some bad requests to his web server for whatever address (or each) that he is hosting on it and viola!
Robert
Best of luck to you...
Robert
www.WebBinaries.com
or www.Pictureview.com (I'd suggest WebBinaries...)
- Where the !A means access to the host is prohibited. (I'm going to have one hell of a time trying to convey this to a tech support moron.)
That means you are being blocked on a firewall level someplace... either on the machine itself or before it.200 This connection requires username/password authentication (Twister v2.0.5.1)
You post something they dont like or that someone complained about?
- Robert
Our (html) pages each have 20 thumbnail images linked to 20 full sized images, and 14 miscellaneous images, of which 4 are repeated (once on top, once on bottom), and even for the same request, from the same server, their proxies will re-request that document. WebTV's and MSN's proxy servers will often do that off and on as well (depending on how many serve a request). AOL will do that 65% of the time (based on a pretty large log sampling) as that many users will end up being served via multiple proxy servers that do not check each other for the documents.
For us, between them, they waste an additional 25% bandwidth. Add to that, versions of IE that try to "pre-cache" (dont know what else to call it) wrong documents (as versions of their proxy servers do as well), and we've calculated 32% wasted bandwidth (in the "pre-cache" case, lets say you are looking at /sportsimages/images/picture12.jpg from document /sportsimages/pictures0001.html - their crap will also try to retrieve index.html from that same directory and the main directory. This issue at least seems to be on the decline... but poorly set up proxy pools seem to be increasing).
Robert
I apologize, as I thought it was clear. Fact is, no such mechanism is in use by AOL, or formerly AtHome, and in all probability other ISP's. To name a few, the virtually only ISP in Australia under any of it's 3 major names, btinternet, webtv, and on and on and on and on...
Sorry for the confusion...
Robert
The honest truth is in numerous large ISPs utilizing proxy servers, they INCREASE bandwidth usage. The reason quite simply is (for instance, in the case of AOL), a user is often served by a half dozen to 2 dozen proxy servers for a single request. There is no mechanism to check to see if proxy12 has the document the user is requesting that now proxy23 needs to get. Thus, often it increases traffic because the proxy servers request documents that were already cached by other proxy servers.
I've done tests with AOL and the former AtHome Network and reloaded the same page and watch a different set of proxy servers [or the same ones retrieving different documents (read: images) from the same page] meaning that roughly 8 out of 10 requests for an image or document, it was being re-retrieved, even though it was already cached by an earlier request by that same user (but via a different one of their ISPs proxy servers).
So... explain how that makes more of their bandwidth available... not just does it decrease their available bandwidth, but mine (for my servers) as well.
They dont provide better service, they dont save anyone money (except the ISP who sells the demographics - which, btw, is one of two reasons why AOL still does it - the other is image compression - which anyone with any AOL intelligence turns off ASAP). AOL has already been in lawsuits about this, so it's not guesswork on my part. Including with a company that was working with doubleclick to use all the data that they (two) were accumulating.
Robert
Because of that, it makes it impossible for Usenet to be policed by online service providers for either copyrighted or illegal content - and even in the event of copyrighted content, a number of copyrighted posts are in Usenet because the owners want them there. The sheer size of Usenet and the inability to distinguish between legitimate posts and intellectual property theft is thus an impossible task.
The DMCA TITLE II in this matter DOES indeed protect both the OSP (including ISP, which is a subset of OSP) AND the copyright holder. The OSP is required to take down the material after sufficient notification has been given (within a reasonable time period), while the copyright holder can sue the original poster for damages.
The copyright owner CAN sue the OSP for contributory damges - IF - the OSP does not honor the (if valid) takedown request.
Seems perfectly reasonable to me. With millions of posts a day, in 80,000+ groups coming from 15,000+ servers, I dont see any other reasonable method, so in this aspect at least, I think the DMCA is adequate to the task.
So, how is the copyright holder getting screwed by this? Because AOL-TW is a big bad company that should be sued without valid reason? I dont think so. Find valid reasons to slam AOL - dont try to rewrite good sections of the law just so you can use it against them.
Oh, but wait! How can the poor copyright owner catch the original poster? Gee, that's an easy one, but most people dont understand enough about the Internet to know that yet... here's some enlightenment (and now Verizon will hate me even more - though not as much as most of the world hates them I am sure)... It's called a circuit trace. Not a traceroute... All you need to know is what IP address the post originated from, and what time, contact the ISP, have them contact their NOC or Verizon to pull the telephone company circuit logs that they will pretend they know nothing about. Those logs are of the Telephone switching network activity that maps the "digital" data for Internet communication (and related TCPIP traffic) over the telco network to a physical telco circuit mounted to the person's house or apartment. When I worked for AOL's backbone provider (they dont own their network - nor does MSN - the same big company does), we'd acquire physical addresses of serious abusers in about 5 minutes through this method - but then again, the company I worked for had more dialup numbers than the next 5 ISP's combined - so Verizon was more than happy to work with them.
If it's a cable modem, the cable companies maintain similar logs which allow similar traces.
In some cases, if the poster is attempting to hide their address (spoofing, etc) that is a little more difficult, but since the telco has to know where to send the packets regardless of the IP address spoofing, they still have logs of where everything came from. Even though they claim otherwise.
If it's non-US, then yes, the copyright holder is SOL... but that's still not a reason to take it out on AOL.
Robert
Robert
The AC units have a high surge when the compressor is turned on.
Robert
Don't forget... if you plug anything with a motor into a UPS, better make sure it is rated for it and that it does not void the warranty. Laser printers, and air conditioners (compressor motors) draw waaaay too much surge current for most run of the mill SOHO UPS's and even for some mid sized ones...