X apps are pretty diverse. If you want to port a low-level X app that uses something like Xt/Motif or otherwise talks pretty closely to the X server, you're probably hosed.
On the other hand, if you are talking about porting a KDE app that uses the Qt/KDE APIs to the exclusion of lower-level junk, then it might be easier -- while there probably wont be a MacOS X version of Qt for a while, you'll be dealing with more abstract (semantically cleaner/simpler) APIs that map better to what's available on MacOS X.
Remember that you'll probably want to produce an app that fits in nicely and feels like a native application -- that involves a lot of work no matter where you're coming from. Doing a straight port of low-level X stuff to MacOS X graphics primitives (*shudder*) would give you an app with a pretty horrendous user experience...
The colo facilities will also have excellent fire supression systems that wont destroy your equipment. (I.E. if a server in a neighboring rack/cage catches on fire, your servers don't get doused in water)
Most office-type sprinkler systems are likely going to be water-based...
I'd dispute that having someone on-hand 24/7 is as big a need as you think, with a properly designed network. In a proper network you eliminate *all* single points of failure, and a non-redundant router/firewall/load balancer (or one that has to be manually switched over) is BAD BAD BAD. With the right equipment, you can even remotely power-cycle a hosed machine.
It's all about how much you want to spend really...
Also, to add my $0.02 on the power issue: The colo my daytime employer uses has seperate connections to two different power grids. Then, they have an entire *floor* filled with battery banks to provide 2-3 hours of backup. The batteries feed into line conditioners before the power ever gets to the servers. If the batteries start to get low, there's a 1,000 gallon hot-fillable generator in the basement which can feed the batteries for 24 hours before it needs more fuel.
And on the redundant links issue: This gets one more advantage -- growth. The lead time on a T1 is about a month. I'm sure a T3 is at least that. If you start to experience sudden growth, you want room *now*. Most colo's wont exceed 50% capacity before they add even more fiber, and you're talking upwards of 30Gb to begin with.
What user is MySQL running as? Make sure you're not hitting the per-user process limit. Also, make sure you're not near the MAX_CONNECTIONS setting.
My daytime employer handles millions of database driven page views every day with MySQL on Solaris... We have our connection limit set to around 4,500 right now...
We have a bunch of NetApps where I work (don't ask) -- they are blindingly fast (something like 6,000 NFS ops/sec with 10ms latency) with oodles of space (hundreds of GB or more) and loas of redundancy. You can get them with dual fiber channels from each head unit to the disk shelves, and you can cluster two of them together so that if a head unit fails, the other just takes over...
Very impressive, relatively cheap... And oh yeah -- no Sun box required since they just hook straight up to your LAN. (100Mb, or 1,000Mb Ethernet...)
Not quite. With things like CORBA and RMI, it's very good. But if you're talking sockets there are some serious problems.
Fundamentally, Java I/O is very limited -- and intentionally so -- by the need for 2 threads to handle bi-directional communication. This is of course because of a lack of asynchronous I/O support. Add the lack of support for multiplexing I/O and you're basically screwed.
Just *try* writing a server in Java that can handle 100,000 simultaneous connections. 200,000 threads is going to break pretty much any server. Yeah, you can resort to clustering and have your servers communicate but that adds a *lot* of complexity in some instances.
A chat server with featues like rooms and private messages and all those goodies that can handle tens of thousands of simultaneous connections on a single machine would be relatively simple to write if one had multiplexing and asynchronous I/O. But without it you have to add your clustering system and implement things like message routing/forwarding and all other sorts of ickiness.
But for simpler stuff, Java makes network programming VERY simple and straight-forward unlike other languages -- and that's a VERY good thing.
IIS is known to be superior to Apache for *static content* -- but what about dynamically generated content, which is what most companies really care about *anyway*.
The kernel httpd will help for performance with static content, but it's not out of beta just yet...
My daytime employer handled 300,000,000 page views in August -- almost all dynamically generated, most on Apache/Linux, the rest on Apache/Solaris.
Our server farm is around 60-ish servers I think... That includes the two E4500's at the back-end. (one live, one hot-spare)
Between this and his previous comments, the impression I get is that he's saying "censor yourself, or be censored."
Remember kiddies: The advancement of society comes from struggle and strife. The most offensive art is the most important: It is the art that tells us the most about who we are, who we think we are, and where we are going. Everything else is elevator music.
Consider this: With the exception of those of us with old Sun IPX boxes lying around (*grin*), Sparc customers tend to be people who need "big iron." I.E. 8, 16, or more CPUs and oodles of memory.
Solaris is well tuned to this sort of hardware environment, and the needs that produce demand for that kind of hardware. Linux still has some deficiencies in this area. Consider Linux's 1-1 thread/process model. Linus considers this to be a Good Thing(tm) and just wants to reduce the cost of a context switch as much as possible. Cetainly admirable but Solaris's 1-m thread/process model tends to be the Right Thing(tm) for this type of environment. Our MySQL server running on a Solaris E4500 is serving 4,500 active connections at any given moment during peak usage hours. It would probably need more but that's what we have our MAX_CONNECTIONS set to right now. We're considering upping that as high as 6,000 in the not too distant future.
Between the relatively high cost of a thread on Linux (in terms of spawning and context switching primarily -- I'm less concerned about the 16KB or so of overhead per thread since that's actually quite reasonable) and the remaining coarse-grained kernel locks, plus the somewhat flawed I/O in 2.2 (ladder-I/O starvation anyone?) Linux isn't quite ready to handle 8+ CPUs in a high demand environment just yet.
But even when Linux *does* get to that point (and it *is* getting there -- the kernel team is doing an incredible job with 2.4, and with companies like IBM helping out at the high-end they certainly have a lot of information at their disposal...) you must consider that usually Sparc hardware comes as a bundle with the software and support. This provides a strong disincentive to use Linux -- you've already paid for a generally excellent OS and top-notch support for it.
This tends to relegate Linux/Sparc to older castaway Sparc hardware which has the distinction of not being cost effective to keep up anymore. Performance vs. space and the cost of maintaining one more box (read: TCO) tend to make it cheaper to replace two of those old Sparc machines with 1 brand-new Intel machine.
Personally, I'm more interested to see Linux on IBM's NUMA-Q based machines with 64 processors... IBM is investing a HUGE amount of energy into making Linux *the* OS for such machines. *drool* (Yes, I'm aware that Linus thinks NUMA is fundamentally the Wrong Thing(tm)... It's still VERY fast though...:>)
I'd give Everyone.net a shot. Ok ok ok -- they are my daytime employer... But I'm signing up with Plug-in-Ad-Sales for my personal site as soon as I meet the 5,000 pageview/month requirement.:)
I'd like to think we're not evil -- but make up your own mind: Take a look, see if you think it's a good deal and give it a shot...
Good point... Most of the co-brandable e-mail systems provide filters on inbound mail (both user definable, and generic "spam" filters), but that of course wouldn't be adequate...
Perhaps parental controls is the Next Big Thing for such providers?
Well... Except that we don't allow users under 13...
Frankly, it's a pain in the butt -- more work for you.
Why not just use one of the free hosted solutions that let you co-brand everything? You can have it be at your domain (whoever@fubar.com) and on your hostname (mail.fubar.com or whatever...) and have it look however you want.
There's a number of such providers... I happen to work for Everyone.net -- one such provider...
One potential counterbalance however, in particular if the viewer's habits are stored on the server, is that the defense could subpoena statistics showing you to be "normal"... If 80% of single white males between the ages of 18-24 watch softcore porn, then it would be hard to use your history of viewing softcore porn against you in a rape trial -- unless the prosecutors want to argue that 80% of 18-24yo SWM's are rapists, or that that particular demographic is noticably more likely to commit rape.
This counterpoint falls down in the case of the truly unique individual -- one whose viewing habits cannot be correlated with his/her demographic profile. However, I think that such isn't really that much of a problem: It isn't financially worthwhile for companies to sub-divide their users *too* much. Then they don't have enough viewers to show the ads to. It only makes sense to divide your userbase up so far.
For example, if the company sells ads based strictly on whether or not you watch porn. Your viewing profile consists strictly of "72% of material viewed was pornographic, 28% was not." Why waste space noting how much of that porn involved S&M/B&D/rape fantasies/whatever?
If it is meaningful to track down to that level of detail, then there will be a large enough userbase for you to use as a defense against such an attack.
Essentially, what it comes down to is that you, as an individual are meaningless to the company, and that is your defense against abuses. Well, that combined with privacy policies. The AOL staffer involved in the Tim McVeigh example probably just screwed up. I'd guess he/she was reprimanded for it... It doesn't make sense financially for companies to abuse your privacy like that... It's bad press and could cost them customers.
[[If you got a phone number at a party and then lost the matchbook, it's frankly not terribly useful to know that in fact she still has a phone number -- you just can't call her because your knowledge of the number was tightly bound to that matchbook.]]
The analogy today is more along the lines of copies of that matchbook being scattered liberally throughout the city. Poke around a bit, and you'd find that phone number again. So would 10,000 other guys.
[[Bottom line is that Plato wrote things we will never read.]]
But had Plato published his works today, it would be relatively trivial to guarantee that they would not be lost. Especially if he tried to limit access to the works. You'd quickly see sites mirroring cracked copies of the works...:)
If nothing else, he could post his works to a dozen of the web-disks (FreeDrive, X-Drive, etc...) and a few of the free hosting companies and be assured that
[[It's not as if the ease of digital duplication and distribution guarantees that it will happen. How many times have you heard, "Well, I usually do a backup, but..."? ]]
Again, I never said it was *guaranteed* that it would happen, or that information would be immortal. The real crux of "information wants to be free" is that as technology improves, it becomes harder and harder to *restrict* access to it, or wipe it out -- if it is valuable enough that people do not want it to be wiped out.
DeCSS just got mass-posted to Usenet. It's in DNS servers, MP3s, MIDI files, a zillion web sites... People want to restrict it, so other people are taking great pains to guarantee it does not get restricted.
[[In fact, there is a case to be made that the churning of technology may cost us information, as older formats become less and less practical to read. If somebody walked up to you right now with an 8" H-DOS disk containing the draft of his unpublished novel, would you be able to read it? There are services that could, if the disk were in readable condition (also not guaranteed), but how often would the cost be prohibitive and the novel simply lost? I can read books printed a thousand years ago, but I may not be able to read floppies from a thousand weeks ago. ]]
That's a bit of a red herring argument. Again, the "information wants to be free" observation is not the same as "information wants to be immortal." Those thousand-year-old books, by the way, have a tendency to crumble into dust. Duplicating them is very difficult. Transferring media is a much easier problem.
But the transferring media issue is a non-issue. If I wish to see some piece of information preserved, and I publish it as mentioned above in the Plato example, then it matters not what media the individual locations choose to store it on. If one of them fails to keep up, the others certainly wont make the same mistake.
[[It's not as if we have a mechanism in place to distribute and transfer more than a fraction of the information out there, and a skewed fraction at that. A lot of people have mirrored DeCSS, but how many have offered to mirror, say, the Census data? Probably more significant (I mean, our kids will find DeCSS the equivalent of a skate key), but much more burdensome and not nearly as sexy.]]
People don't mirror the census data because the census beaurau isn't trying to quash the data. "Information wants to be free" is just an observation on the futility of *restricting* access to information.
The *information* was not lost or destroyed, the *books* were. The information happened to be bound to those books.
The point of the "information wants to be free" observation is that as technology improves, this tight binding between information and the physical medium tends to decrease, resulting in the increasing impossibility of containing or destroying information. You can't burn down a library and have DeCSS be lost forever.
The truth of the statement is a function of the state of techology, and by extension, of time.
You miss my point again. In a "large enough" city, such as Los Angeles, someone will be murdered. Maybe a few dozen people. But that is the *exception*, rather than the rule. (Translation: The VAST majority of people *will not* be murdered.)
I never said information was *immortal*. My point is that it cannot be effectively *restrained*. If it lacks value, it will be lost. If it has value, you will not stop people from getting it.
Inevitability is a good word to use here. It is inevitable that information will become "free." This is merely a factor of the innate impossibility of reliably and permanently *restricting* the flow of information (witness copy protection mechanisms), combined with the virtually nil cost of reproduction, and the ever decreasing cost of dissemination.
The statement "information wants to be free" is an observation that information cannot be effectively restrained. The "water wants to run downhill" analogy is superb. The "people want to be murdered" analog is not, because even in a crowd, it is the exception, rather than the rule that someone will be murdered.
Ads aren't the *only* revenue stream Google has. But they are a neccesary one. Partnerships, licensing, and cobranding (click "Everything Else" on their main page) are another strong source of revenue.
I suppose it's possible also that they can increase their CPM beyond even $20 due to the targeting they do (read their advertising pitch) but I don't know enough to be sure about that.
Essentially, if they *didn't* leverage this revenue stream, investors would question why they were wasting money on something (their end-user site) that didn't make any money. The benefits (exposure for their B2B products primarily) would seriously outweigh the costs.
When they see the cash "rolling in"? Frankly, you don't understand the economics of the web.
Banner ads provide very little revenue. Let's do the math... We'll say a $10 CPM. (the amount Google gets for every 1,000 ads they show) Google had 6.6 million unique visitors in July according to PCData Online. Since it's a search engine we'll assume about 2.5 page views per user. (although that's probably a bit on the high side)
(6,600,000 * 2.5) * (10/1000) = $165,000/mo.
Even if you make more favorable assumptions you don't wind up with much money:
(7,000,000 * 3) * (20/1000) = $420,000/mo.
Keeping in mind that Google's operating cost is probably pretty high if you just consider colocation (they are hosted at AboveNet), bandwidth, and hardware. Then there is the cost of employees -- Google goes to great lengths to attract and keep top talent, even going so far as to hire an on-site gourmet chef.
Banner ads alone are insufficient to keep all but the lowest-overhead companies going. Google has partnerships involving licensing of their engine or cobranding of it but if that is all they did it would not make sense to keep the end-user site (the one you and I use for searches) going -- it would be a money pit. Banner ads reduce the loss incurred from the end-user site.
Anyone who talks about Google "selling out" is, frankly, an idiot. To run their search engine, Google needs hundreds, perhaps thousands of servers (you don't think a desktop PC running off a DSL line indexes 1 billion pages and serves the searching needs of 6.6 million distinct people I hope?) using tons of bandwidth and space at a pricey colocation facility. Do you honestly think that they started this company with no intention of being *profitable*? Do you honestly think they threw all this money, time, and energy into making Google out of some sense of philanthropy?
I can't speak for other services, but the only way you'd be able to do this with Everyone.net (disclosure: my employer) would be to use the web-based client instead of setting up a forwarding account.
You'd be able to send and receive e-mail but you have to use our client, and you'd be limited to 6MB of space.
Everyone.net (disclosure: my employer) provides free web based e-mail, branded to how your site looks and to your domain. You can however just sign up, point the MX to us, and then set up forwarding accounts to alternative e-mail addresses...
The end result is that you'll get your "me@foo.com" e-mails sent to "me@suckycablemodemprovider.com"...
Were you using NFSv2 or NFSv3 with Linux? At my company we've had severe performance problems with NFSv2 but have gotten a *huge* performance increase with NFSv3...
Admittedly, we're using Linux for clients and our server is a NetApp box, but if Linux's NFSv2 is screwed up for clients I'd guess it's screwed up for servers...
That's an HTML comment, not a JavaScript comment. It is there for browsers that don't understand JavaScript, so they wont display it to users. This is a very common practice.
That's the argument Napster made in court essentially, but reversed. Betamax was ruled to be legal even though it can be used for piracy, so why should Napster be illegal?
The answer is simple: Betamax (and Windows file sharing) have significant legitimate, legal uses. Napster does not.
While you and your buddies may have traded a few dozen MP3s via your college dorm's LAN the vast majority of Windows file-sharing usage is by corporations sharing internal documents.
Yes, Napster *can* be used to promote independent artists, but only peripherally. The legitimate use is an *anomoly* within Napster. According to the suit, something like 89% of MP3 files traded on Napster are *illegal*. Personally, I expected that number to be higher. Most people are just going to hit the search engine -- how do you search for artists you've never heard of?
The reality is that the folks at Napster created a tool to promote the *illegal* copying of music and they got called on it. Period.
Peripheral issues such as this being better for the RIAA's sales are irrelevent: The RIAA owns the rights to these works and as long as the copyright system is in place, the copyright holder sets the terms.
What really gets me is the *hypocrisy* of this matter. As Open Source/Free Software developers we depend on copyright law to enforce the terms that we as copyright holders have defined. (the GPL et al) Someone comes along and distributes a modified version of one of our OS/FS programs and doesn't redistribute the source. What do we do? We get angry. We talk about calling in the lawyers to fight these parasites and bloodsuckers. And we are fully justified in our anger because the users of our copyrighted works are not playing by the rules which we rightfully defined.
The RIAA is doing the same thing here. They defined the rules, and all you Napster users broke them. Now all you Napster users want to get angry at *them*? The arrogance and hypocrisy is just absurd!
Slashdot Mirror
X apps are pretty diverse. If you want to port a low-level X app that uses something like Xt/Motif or otherwise talks pretty closely to the X server, you're probably hosed.
On the other hand, if you are talking about porting a KDE app that uses the Qt/KDE APIs to the exclusion of lower-level junk, then it might be easier -- while there probably wont be a MacOS X version of Qt for a while, you'll be dealing with more abstract (semantically cleaner/simpler) APIs that map better to what's available on MacOS X.
Remember that you'll probably want to produce an app that fits in nicely and feels like a native application -- that involves a lot of work no matter where you're coming from. Doing a straight port of low-level X stuff to MacOS X graphics primitives (*shudder*) would give you an app with a pretty horrendous user experience...
-JF
The colo facilities will also have excellent fire supression systems that wont destroy your equipment. (I.E. if a server in a neighboring rack/cage catches on fire, your servers don't get doused in water)
Most office-type sprinkler systems are likely going to be water-based...
I'd dispute that having someone on-hand 24/7 is as big a need as you think, with a properly designed network. In a proper network you eliminate *all* single points of failure, and a non-redundant router/firewall/load balancer (or one that has to be manually switched over) is BAD BAD BAD. With the right equipment, you can even remotely power-cycle a hosed machine.
It's all about how much you want to spend really...
Also, to add my $0.02 on the power issue: The colo my daytime employer uses has seperate connections to two different power grids. Then, they have an entire *floor* filled with battery banks to provide 2-3 hours of backup. The batteries feed into line conditioners before the power ever gets to the servers. If the batteries start to get low, there's a 1,000 gallon hot-fillable generator in the basement which can feed the batteries for 24 hours before it needs more fuel.
And on the redundant links issue: This gets one more advantage -- growth. The lead time on a T1 is about a month. I'm sure a T3 is at least that. If you start to experience sudden growth, you want room *now*. Most colo's wont exceed 50% capacity before they add even more fiber, and you're talking upwards of 30Gb to begin with.
-JF
What user is MySQL running as? Make sure you're not hitting the per-user process limit. Also, make sure you're not near the MAX_CONNECTIONS setting.
My daytime employer handles millions of database driven page views every day with MySQL on Solaris... We have our connection limit set to around 4,500 right now...
-JF
We have a bunch of NetApps where I work (don't ask) -- they are blindingly fast (something like 6,000 NFS ops/sec with 10ms latency) with oodles of space (hundreds of GB or more) and loas of redundancy. You can get them with dual fiber channels from each head unit to the disk shelves, and you can cluster two of them together so that if a head unit fails, the other just takes over...
Very impressive, relatively cheap... And oh yeah -- no Sun box required since they just hook straight up to your LAN. (100Mb, or 1,000Mb Ethernet...)
-JF
(Java Network Programming *rocks*)
Not quite. With things like CORBA and RMI, it's very good. But if you're talking sockets there are some serious problems.
Fundamentally, Java I/O is very limited -- and intentionally so -- by the need for 2 threads to handle bi-directional communication. This is of course because of a lack of asynchronous I/O support. Add the lack of support for multiplexing I/O and you're basically screwed.
Just *try* writing a server in Java that can handle 100,000 simultaneous connections. 200,000 threads is going to break pretty much any server. Yeah, you can resort to clustering and have your servers communicate but that adds a *lot* of complexity in some instances.
A chat server with featues like rooms and private messages and all those goodies that can handle tens of thousands of simultaneous connections on a single machine would be relatively simple to write if one had multiplexing and asynchronous I/O. But without it you have to add your clustering system and implement things like message routing/forwarding and all other sorts of ickiness.
But for simpler stuff, Java makes network programming VERY simple and straight-forward unlike other languages -- and that's a VERY good thing.
-JF
IIS is known to be superior to Apache for *static content* -- but what about dynamically generated content, which is what most companies really care about *anyway*.
The kernel httpd will help for performance with static content, but it's not out of beta just yet...
My daytime employer handled 300,000,000 page views in August -- almost all dynamically generated, most on Apache/Linux, the rest on Apache/Solaris.
Our server farm is around 60-ish servers I think... That includes the two E4500's at the back-end. (one live, one hot-spare)
-JF
Between this and his previous comments, the impression I get is that he's saying "censor yourself, or be censored."
Remember kiddies: The advancement of society comes from struggle and strife. The most offensive art is the most important: It is the art that tells us the most about who we are, who we think we are, and where we are going. Everything else is elevator music.
-JF
BG said that the new nVidia chip would be three *generations* ahead of current chips, not that it would be 3x as fast. There's a VERY big difference.
Abrash said it would be 1.5-2 *generations* ahead of current chips.
-JF
Consider this: With the exception of those of us with old Sun IPX boxes lying around (*grin*), Sparc customers tend to be people who need "big iron." I.E. 8, 16, or more CPUs and oodles of memory.
:>)
Solaris is well tuned to this sort of hardware environment, and the needs that produce demand for that kind of hardware. Linux still has some deficiencies in this area. Consider Linux's 1-1 thread/process model. Linus considers this to be a Good Thing(tm) and just wants to reduce the cost of a context switch as much as possible. Cetainly admirable but Solaris's 1-m thread/process model tends to be the Right Thing(tm) for this type of environment. Our MySQL server running on a Solaris E4500 is serving 4,500 active connections at any given moment during peak usage hours. It would probably need more but that's what we have our MAX_CONNECTIONS set to right now. We're considering upping that as high as 6,000 in the not too distant future.
Between the relatively high cost of a thread on Linux (in terms of spawning and context switching primarily -- I'm less concerned about the 16KB or so of overhead per thread since that's actually quite reasonable) and the remaining coarse-grained kernel locks, plus the somewhat flawed I/O in 2.2 (ladder-I/O starvation anyone?) Linux isn't quite ready to handle 8+ CPUs in a high demand environment just yet.
But even when Linux *does* get to that point (and it *is* getting there -- the kernel team is doing an incredible job with 2.4, and with companies like IBM helping out at the high-end they certainly have a lot of information at their disposal...) you must consider that usually Sparc hardware comes as a bundle with the software and support. This provides a strong disincentive to use Linux -- you've already paid for a generally excellent OS and top-notch support for it.
This tends to relegate Linux/Sparc to older castaway Sparc hardware which has the distinction of not being cost effective to keep up anymore. Performance vs. space and the cost of maintaining one more box (read: TCO) tend to make it cheaper to replace two of those old Sparc machines with 1 brand-new Intel machine.
Personally, I'm more interested to see Linux on IBM's NUMA-Q based machines with 64 processors... IBM is investing a HUGE amount of energy into making Linux *the* OS for such machines. *drool* (Yes, I'm aware that Linus thinks NUMA is fundamentally the Wrong Thing(tm)... It's still VERY fast though...
-JF
I'd give Everyone.net a shot. Ok ok ok -- they are my daytime employer... But I'm signing up with Plug-in-Ad-Sales for my personal site as soon as I meet the 5,000 pageview/month requirement. :)
I'd like to think we're not evil -- but make up your own mind: Take a look, see if you think it's a good deal and give it a shot...
-JF
Good point... Most of the co-brandable e-mail systems provide filters on inbound mail (both user definable, and generic "spam" filters), but that of course wouldn't be adequate...
Perhaps parental controls is the Next Big Thing for such providers?
Well... Except that we don't allow users under 13...
-JF
Frankly, it's a pain in the butt -- more work for you.
:)
Why not just use one of the free hosted solutions that let you co-brand everything? You can have it be at your domain (whoever@fubar.com) and on your hostname (mail.fubar.com or whatever...) and have it look however you want.
There's a number of such providers... I happen to work for Everyone.net -- one such provider...
Lower maintenance...
-JF
One potential counterbalance however, in particular if the viewer's habits are stored on the server, is that the defense could subpoena statistics showing you to be "normal"... If 80% of single white males between the ages of 18-24 watch softcore porn, then it would be hard to use your history of viewing softcore porn against you in a rape trial -- unless the prosecutors want to argue that 80% of 18-24yo SWM's are rapists, or that that particular demographic is noticably more likely to commit rape.
This counterpoint falls down in the case of the truly unique individual -- one whose viewing habits cannot be correlated with his/her demographic profile. However, I think that such isn't really that much of a problem: It isn't financially worthwhile for companies to sub-divide their users *too* much. Then they don't have enough viewers to show the ads to. It only makes sense to divide your userbase up so far.
For example, if the company sells ads based strictly on whether or not you watch porn. Your viewing profile consists strictly of "72% of material viewed was pornographic, 28% was not." Why waste space noting how much of that porn involved S&M/B&D/rape fantasies/whatever?
If it is meaningful to track down to that level of detail, then there will be a large enough userbase for you to use as a defense against such an attack.
Essentially, what it comes down to is that you, as an individual are meaningless to the company, and that is your defense against abuses. Well, that combined with privacy policies. The AOL staffer involved in the Tim McVeigh example probably just screwed up. I'd guess he/she was reprimanded for it... It doesn't make sense financially for companies to abuse your privacy like that... It's bad press and could cost them customers.
-JF
[[If you got a phone number at a party and then lost the matchbook, it's frankly not terribly useful to know that in fact she still has a phone number -- you just can't call her because your knowledge of the number was tightly bound to that matchbook.]]
:)
The analogy today is more along the lines of copies of that matchbook being scattered liberally throughout the city. Poke around a bit, and you'd find that phone number again. So would 10,000 other guys.
[[Bottom line is that Plato wrote things we will never read.]]
But had Plato published his works today, it would be relatively trivial to guarantee that they would not be lost. Especially if he tried to limit access to the works. You'd quickly see sites mirroring cracked copies of the works...
If nothing else, he could post his works to a dozen of the web-disks (FreeDrive, X-Drive, etc...) and a few of the free hosting companies and be assured that
[[It's not as if the ease of digital duplication and distribution guarantees that it will happen. How many times have you heard, "Well, I usually do a backup, but..."? ]]
Again, I never said it was *guaranteed* that it would happen, or that information would be immortal. The real crux of "information wants to be free" is that as technology improves, it becomes harder and harder to *restrict* access to it, or wipe it out -- if it is valuable enough that people do not want it to be wiped out.
DeCSS just got mass-posted to Usenet. It's in DNS servers, MP3s, MIDI files, a zillion web sites... People want to restrict it, so other people are taking great pains to guarantee it does not get restricted.
[[In fact, there is a case to be made that the churning of technology may cost us information, as older formats become less and less practical to read. If somebody walked up to you right now with an 8" H-DOS disk containing the draft of his unpublished novel, would you be able to read it? There are services that could, if the disk were in readable condition (also not guaranteed), but how often would the cost be prohibitive and the novel simply lost? I can read books printed a thousand years ago, but I may not be able to read floppies from a thousand weeks ago. ]]
That's a bit of a red herring argument. Again, the "information wants to be free" observation is not the same as "information wants to be immortal." Those thousand-year-old books, by the way, have a tendency to crumble into dust. Duplicating them is very difficult. Transferring media is a much easier problem.
But the transferring media issue is a non-issue. If I wish to see some piece of information preserved, and I publish it as mentioned above in the Plato example, then it matters not what media the individual locations choose to store it on. If one of them fails to keep up, the others certainly wont make the same mistake.
[[It's not as if we have a mechanism in place to distribute and transfer more than a fraction of the information out there, and a skewed fraction at that. A lot of people have mirrored DeCSS, but how many have offered to mirror, say, the Census data? Probably more significant (I mean, our kids will find DeCSS the equivalent of a skate key), but much more burdensome and not nearly as sexy.]]
People don't mirror the census data because the census beaurau isn't trying to quash the data. "Information wants to be free" is just an observation on the futility of *restricting* access to information.
The *information* was not lost or destroyed, the *books* were. The information happened to be bound to those books.
The point of the "information wants to be free" observation is that as technology improves, this tight binding between information and the physical medium tends to decrease, resulting in the increasing impossibility of containing or destroying information. You can't burn down a library and have DeCSS be lost forever.
The truth of the statement is a function of the state of techology, and by extension, of time.
You miss my point again. In a "large enough" city, such as Los Angeles, someone will be murdered. Maybe a few dozen people. But that is the *exception*, rather than the rule. (Translation: The VAST majority of people *will not* be murdered.)
So again, you're analogy is wrong.
I never said information was *immortal*. My point is that it cannot be effectively *restrained*. If it lacks value, it will be lost. If it has value, you will not stop people from getting it.
Inevitability is a good word to use here. It is inevitable that information will become "free." This is merely a factor of the innate impossibility of reliably and permanently *restricting* the flow of information (witness copy protection mechanisms), combined with the virtually nil cost of reproduction, and the ever decreasing cost of dissemination.
The statement "information wants to be free" is an observation that information cannot be effectively restrained. The "water wants to run downhill" analogy is superb. The "people want to be murdered" analog is not, because even in a crowd, it is the exception, rather than the rule that someone will be murdered.
-JF
Ads aren't the *only* revenue stream Google has. But they are a neccesary one. Partnerships, licensing, and cobranding (click "Everything Else" on their main page) are another strong source of revenue.
I suppose it's possible also that they can increase their CPM beyond even $20 due to the targeting they do (read their advertising pitch) but I don't know enough to be sure about that.
Essentially, if they *didn't* leverage this revenue stream, investors would question why they were wasting money on something (their end-user site) that didn't make any money. The benefits (exposure for their B2B products primarily) would seriously outweigh the costs.
-JF
When they see the cash "rolling in"? Frankly, you don't understand the economics of the web.
Banner ads provide very little revenue. Let's do the math... We'll say a $10 CPM. (the amount Google gets for every 1,000 ads they show) Google had 6.6 million unique visitors in July according to PCData Online. Since it's a search engine we'll assume about 2.5 page views per user. (although that's probably a bit on the high side)
(6,600,000 * 2.5) * (10/1000) = $165,000/mo.
Even if you make more favorable assumptions you don't wind up with much money:
(7,000,000 * 3) * (20/1000) = $420,000/mo.
Keeping in mind that Google's operating cost is probably pretty high if you just consider colocation (they are hosted at AboveNet), bandwidth, and hardware. Then there is the cost of employees -- Google goes to great lengths to attract and keep top talent, even going so far as to hire an on-site gourmet chef.
Banner ads alone are insufficient to keep all but the lowest-overhead companies going. Google has partnerships involving licensing of their engine or cobranding of it but if that is all they did it would not make sense to keep the end-user site (the one you and I use for searches) going -- it would be a money pit. Banner ads reduce the loss incurred from the end-user site.
Anyone who talks about Google "selling out" is, frankly, an idiot. To run their search engine, Google needs hundreds, perhaps thousands of servers (you don't think a desktop PC running off a DSL line indexes 1 billion pages and serves the searching needs of 6.6 million distinct people I hope?) using tons of bandwidth and space at a pricey colocation facility. Do you honestly think that they started this company with no intention of being *profitable*? Do you honestly think they threw all this money, time, and energy into making Google out of some sense of philanthropy?
Please!
-JF
I can't speak for other services, but the only way you'd be able to do this with Everyone.net (disclosure: my employer) would be to use the web-based client instead of setting up a forwarding account.
You'd be able to send and receive e-mail but you have to use our client, and you'd be limited to 6MB of space.
-JF
Everyone.net (disclosure: my employer) provides free web based e-mail, branded to how your site looks and to your domain. You can however just sign up, point the MX to us, and then set up forwarding accounts to alternative e-mail addresses...
The end result is that you'll get your "me@foo.com" e-mails sent to "me@suckycablemodemprovider.com"...
-JF
Were you using NFSv2 or NFSv3 with Linux? At my company we've had severe performance problems with NFSv2 but have gotten a *huge* performance increase with NFSv3...
Admittedly, we're using Linux for clients and our server is a NetApp box, but if Linux's NFSv2 is screwed up for clients I'd guess it's screwed up for servers...
-JF
That's an HTML comment, not a JavaScript comment. It is there for browsers that don't understand JavaScript, so they wont display it to users. This is a very common practice.
The JavaScript is still executed.
-JF
That's the argument Napster made in court essentially, but reversed. Betamax was ruled to be legal even though it can be used for piracy, so why should Napster be illegal?
The answer is simple: Betamax (and Windows file sharing) have significant legitimate, legal uses. Napster does not.
While you and your buddies may have traded a few dozen MP3s via your college dorm's LAN the vast majority of Windows file-sharing usage is by corporations sharing internal documents.
Yes, Napster *can* be used to promote independent artists, but only peripherally. The legitimate use is an *anomoly* within Napster. According to the suit, something like 89% of MP3 files traded on Napster are *illegal*. Personally, I expected that number to be higher. Most people are just going to hit the search engine -- how do you search for artists you've never heard of?
The reality is that the folks at Napster created a tool to promote the *illegal* copying of music and they got called on it. Period.
Peripheral issues such as this being better for the RIAA's sales are irrelevent: The RIAA owns the rights to these works and as long as the copyright system is in place, the copyright holder sets the terms.
What really gets me is the *hypocrisy* of this matter. As Open Source/Free Software developers we depend on copyright law to enforce the terms that we as copyright holders have defined. (the GPL et al) Someone comes along and distributes a modified version of one of our OS/FS programs and doesn't redistribute the source. What do we do? We get angry. We talk about calling in the lawyers to fight these parasites and bloodsuckers. And we are fully justified in our anger because the users of our copyrighted works are not playing by the rules which we rightfully defined.
The RIAA is doing the same thing here. They defined the rules, and all you Napster users broke them. Now all you Napster users want to get angry at *them*? The arrogance and hypocrisy is just absurd!
-JF