I have run many of my own benchmarks on Apple's machines. Apple blatantly lies about their performance (we're way past distortion here) and everybody knows it. A PowerPC G4 chip is _about_ 15% faster per clock than a Pentium III, on both fp and int. The P4 is actually slower than a PIII per clock. Apple _could_ claim that their new 867mhz G4 is the same speed as a 1.3ghz P4, without distorting too much.
ASN.1 is the basis of a great many protocols, LDAP among them. What is not mentioned in the article is that ASN.1 is a binary protocol and is therefore not human-readable. It may save space for bandwidth-constrained applications. However, bandwidth has a tendency to increase over time. When all wireless handhelds have a megabit of bandwidth, we would sorely regret being tied to ASN.1, as LDAP regrets it now.
Not to mention, ASN.1 does not generally reduce the document size by more than 40% compared to XML. Think about it: how much space is really taken by tags?
It's also worth noting that there is lots of documentation surrounding XML. With ASN.1 you have to download the spec from ITU which is an INCREDIBLY annoying organization and their specs are barely readable and they charge money to look at them, despite the fact that they are supposedly an open organization. The IETF and the W3C are actually open organizations; ITU just pretends to be. ITU does whatever it can to restrict the distribution of their specifications.
Take a second processor, with more pipelines available for instruction issue. Since it has more pipelines available it is able to issue more instructions while waiting for the branch to the calculated.
He was referring to pipeline length, not width. In a 20 stage processor at the same clock rate, it takes longer to fill a pipeline and consequently the branch misprediction penalty is worse.
Suppose you have two processors, each at the same clock speed. One has a 5-stage pipeline, the second a 20-stage pipeline. Suppose that there is a branch every 6 instructions (which is typical). For every mispredicted branch, the first processor need only throw away 4 instructions, but the second 19. If most branches were mispredicted, it would kill the second processor.
Pipeline length and clock speed are closely related design parameters. Longer pipes allow faster clock rates (because less is done per cycle per stage), but they increase the branch misprediction penalty. Generally there is a "happy compromise" for a processor, between pipeline length and clock speed. Most recent chips have found that happy medium to be around 10 stages. The Pentium-4 is unusual in the regard that it has 20 stages. Branch prediction therefore becomes extremely important.
Long pipelines tend to benefit Floating Point code more than Integer code, because FP is more loop-intensive, and the branches are therefore more easily predicted. This is why the P4, with its extremely long pipelines, performes poorly on integer performance compared to the PIII, but well on FP.
It's worth noting that Sun's MAJC architecture already has an implementation that is commercially available (MAJC 5200) and employs several of the multithreading strategies outlined in the article (including CMP and CMT).
Sun has an excellent in-depth explanation of those multithreading technologies in it's
whitepaper.
Your post was emotional, distorted, incendiary, and silly. I felt like I was reading Atlas Shrugged.
That the left-wing media and so-called "scientific" establishments attack all forms of progress is nothing new. The anti-Human forces have been at it for decades, sickly raging at those who are better than them.
Umm, the article only mentioned a new theory about what caused the extinction of some species. All of this about attacking progress in the name of sickly fascism etc etc is just a figment of your paranoid imagination, and you likely imagine yourself to be the lone crusader against it. Nothing about "anti-humanism" was mentioned in the post; it was just a theory about extinction.
In our present day, the whining, half-human statists wish to use their leverage inside the corrupt, reeking organ of fascism known as the modern nation to bind as they could not bind our ancient forefathers.
I'm glad you're capable of being dispassionate and avoiding loaded terms.
It is the liberals who have passed their point of usefulness. A political Darwinism will sweep the land, raging like holy fire across a peoples weary of the lies, slander, and weakness of the liberal ideals.
You accuse your opponents of fascism?
Try to relax and think for awhile. When you find yourself getting upset, you aren't thinking.
The greatest cost to any large IT department, by far, is personnel. Every $80,000 employee costs the corporation over $120,000 in salary, taxes, benefits, etc, PER YEAR. The cost of Web development tools, web servers, etc, is maybe a few hundred or a few thousand dollars. If a company is paying over $100k a year for a developer, they don't mind paying a few thousand for a development tool.
Open source tools cost an IT deparment more _in salary_ because they are more difficult to learn. MCSEs cannot charge as much as unix admins.
There are some pieces of software that are prohibitively expensive, like Oracle. However there are no open source equivalents.
Not to mention, corporations have not shown themselves to be very eager to reduce software costs. Otherwise they would buy IBM's DB2 at $20,000 rather than Oracle8i at several million dollars. Both have roughly the same features.
The article clearly states that the technology can only differentiate between 20 or so people. As such it will only be used to tell which family member is using the computer at a given time. Therefore, the technology will only be useful to a company that is already tracking your computer but does not know which family member is currently using it. Thus, this technology will be (at best) an adjunct to more serious tracking technologies.
It is unlikely that a technology could ever be developed that could pick a single person out of the entire population, from keystrokes alone. Not that much information is revealed from keystroke timing. Therefore, this technology is far less of a threat to privacy than, say, fingerprints.
People constantly make arguments about the superior reliability of the Linux OS, especially when compared to NT. Linux has a more reliable _kernel_, but it unquestionably does not have a more reliable gui, or more reliable desktop apps. Gnome with enlightenment has constant glitches, and crashes regularly. My new Red Hat 7.1 installation (Gnome/Sawfish) hangs the entire box within 3 days no matter what box I install it on. Even the Linux mail clients, which are comparatively simple pieces of software, are very much in beta (Balsa & Evolution, for example). The desktop software (diagramming tools, documentation tools, etc) are in an extremely primitive state: development on AbiWord appears to have ceased, and Dia is not even in beta.
The sole redeeming feature is the newest release of Mozilla (0.9), which _is nice._ It has gotten dramatically better. However, this is no vindication of the Open Source development model: Mozilla development is done in a traditional "cathedral" way with a paid, professional development staff.
Aside from that one exception, the Linux desktop environment is vastly inferior to its commercial alternatives (OS X & Windows 2000).
Most game consoles are sold for well below their cost of manufacture. The console makers do this so that there is a strong enticement to buy the console. The loss generated by the console sale is compensated by increasing the price of the games: in the long run there is a hefty profit.
This works because most consumers only look at the up-front price rather than the total cost of ownership. A cheap console with expensive games sells better than an expensive console with cheap games, because the entry is so much easier.
An open platform cannot exploit this advantage, so the up-front hardware cost will necessarily be much greater. For this reason, an open console standard is unlikely to succeed.
This benchmark was not that useful
on
Kernel Benchmarks
·
· Score: 4
First, the university benchmarking team simply ran lmbench (a free, popular, old kernel benchmarking utility) on a variety of kernels. Claiming that:
Three students and a professor from Northern Michigan University spent the semester benchmarking a bunch of Linux kernels
...somewhat exaggerates this accomplishment
Second, no data were presented on the main areas of the kernel that were improved. How is SMP performance in kernel space? Did the finer grained locks help? How is the performance from the threaded IP stack? Does it prevent IO blocking?
THAT kind of information would have been interesting. They tested only things that the kernel has done forever.
Unions are unnecessary in IT
on
IT Unions?
·
· Score: 1
Tech is a field where employees have a marketable skill which they leverage to increase their salaries and benefits. This system works very well: most tech workers are paid much better than average and are treated much better than average.
The main complaint of the article is that timelines today are too short, so tech workers feel rushed. This unionization effort seems to stem from a desire of the tech workers to usurp the role of determining timelines from management. Presumably they feel the product should take much longer to produce and correspondingly have fewer bugs. Perhaps they would feel a better sense of craftsmanship if their product was more carefully engineered. However the purpose of an organization is not to further feelings of craftsmanship in their employees, but to sell products for money.
It is management that is best able to determine what is the best tradeoff between development time and bug count. A sense of craftsmanship is absolutely irrelevant. It is incredibly presumptuous for tech workers to demand that they should be able to determine product timelines, that the purpose of their employer is to make them feel craftsmanship, and that they should be paid well for doing it.
There are two fundamentally different contentions in the MS speech, which are not very well separated. They are:
1. OSS is not a viable model, because you need intellectualy property to charge money and pay for R&D.
2. OSS poses a threat to commercial software even for the companies that do not participate in it.
The second contention is found in this quote:
It [OSS] also fundamentally undermines the independent commercial software sector because it effectively makes it impossible to distribute software on a basis where recipients pay for the product rather than just the cost of distribution.
...this implies that OSS should be banned or restricted.
Oddly, the two contentions blatantly contradict each other, when they are juxtaposed explicitly. If OSS is not a viable development model, then it poses no threat to commercial software. Because if OSS is not a viable model, then it will be unable to produce a competitive product; if it does produce a competitive product, then contention #1 is mistaken.
It is also odd that it is Microsoft, not OSS, that wants to violate property rights. Fundamental to the notion of property rights is the ability to dispose of your own property in any way that pleases you, including giving it away or even destroying it. It would be a serious violation of intellectual property to prevent the owners of code from distributing it as they see fit.
This article was in the vein of CNN's crossfire: calculated to be polemical, provocative, and irrational, so as to incite discussion and readership.
For example, here is a quote about the FSF:
He [RMS] was aware of the phenomenon codified by Abraham Maslow: there are lots of people who will sign on to just about any movement in exchange for the sense of belongingness that being the proud member of a group imparts. Fair enough. Nothing wrong with that. As long as you live it.
This kind of unsupported pop psychoanalysis could be levelled against any group or organization. In this case, the evidence weighs heavily against it: whatever RMS' faults, he almost certainly believes in what he preaches. I doubt very much that RMS started the FSF to acquire needy followers, and I doubt very much that people join for a sense of belongingness. Writing code in your basement for a compiler with other people you've never met is not a sure a path to belongingness. Anyone looking for a sense of beloning could far more easily find it in a church.
The other claims are similarly weak:
Gnome is controlled -- c'mon, don't kid yourself -- by two companies
The parenthetical clause ("c'mon, don't kid yourself") is the only support offered for this statement. The statement implies that RMS is a corporate lackey, which I seriously doubt.
It's tragic that this kind of talk-show commentary has eclipsed real argumentation.
Tech support sucks because people aren't willing to pay for it.
It costs money to provide tech support. Therefore, companies that have good support must charge more for their products. This gives consumers a choice: buy from the cheaper vendor with no support, or from the more expensive vendor with adequate support. Consumers routinely choose the cheaper alternative, even when they know that the support will be terrible.
Witness the runaway success of vendors that have a reputation for having flagrantly bad service. Fry's electronics is a perfect example. Everyone who goes to Fry's ends up saying that they will never go back. Then Fry's runs an ad for RAM at $5 less than their competitors, and all those consumers who swore them off go right back to Fry's.
If people were willing to pay money for support, and people flocked to the vendors offering more service, then companies would be climbing on top of each other to offer more support. People don't actually want support, despite what they say. People say one thing and do another. What they prefer is to save the $5 and forgo the support, then bitch for 2 hours about how badly they were wronged and how bad support is nowadays, then they pocket the $5 and repeat the process.
With a commercial database, you pay for proven reliability, proven scalability, support, thorough documentation, and features.
MySql is not a database in the same sense that Oracle is. The four basic criteria that need to be met by any database are summarized by the ACID acronym (Atomicity, Consistency, Isolation, Durability). MySql satisfies none of them and lacks a huge number of very basic features. It is not a database, but a simple SQL parser on top of a DBM file.
The ACID test assures that the data will always remain in a consistent state, despite failures. Most enterprises (banks, businesses, etc) would pay any amount for this feature.
Postgres is a real database. However, it is still _very_ slow and scales poorly. The documentation for Postgres has glaring omissions, and the database lacks some features.
It is often worth paying for a commercial database. There is not a substantial difference between Oracle the "other" commerical databases (DB2, Informix, Sybase, etc). All of those databases are enterprise-class and offer features and performance similar to Oracle. Oracle charges 3-4x as much money because they are the market leader and market leaders always do that. Managers are extremely comfortable with market leaders, programmers are familiar with it, and DBAs are easy to find. That justifies the extra cost for most corporations.
It is strange that this "ask slashdot" column lumped MSSQL in with Postgres and MySql. MSSQL is not free, and an enterprise license is not cheap (>$20k if I recall). I have very little experience with MSSQL but it is generally not well regarded.
If you're just serving out static Web pages, it doesn't matter what kind of machine you have, so long as it's reliable.
I took my PowerPC 601/75mhz (LinuxPPC) with 16M of Ram and hit it with a few million httpd requests. Note that this computer is worth about $50 on the open market. It could saturate more than five T1s and serve out many millions of pages per day.
The only place a Web Server would get bogged down is on CGIs and database accesses. To what extent it gets bogged down is entirely dependent on the CGI or database; the Web serving software is not a significant part of the equation.
CGIs don't tend to be _that_ big of a deal, though. Really the rough spots are around database access. That's the only place where real money should be spent.
Uhh... I've installed Linux on seven of my machines, I never even bothered to check hardware compatibility, and it always worked just fine. All the boxes were non-brand name small-manufacturer clones.
... You saw prices of EIGHTY DOLLARS for a copy of RH 5.2?!?! I walked down to my local Software Etc. and got an _official_ RH for $39.99.
... If you don't need the box and manual you can order the unofficial version for $1.99 from www.cheapbytes.com. That's not _free_ but it's very close.
If you have anything even resembling normal hardware, there will be drivers. The problem is easily avoided by not buying extremely obscure hardware.
Slashdot Mirror
I have run many of my own benchmarks on Apple's machines. Apple blatantly lies about their performance (we're way past distortion here) and everybody knows it. A PowerPC G4 chip is _about_ 15% faster per clock than a Pentium III, on both fp and int. The P4 is actually slower than a PIII per clock. Apple _could_ claim that their new 867mhz G4 is the same speed as a 1.3ghz P4, without distorting too much.
ASN.1 is the basis of a great many protocols, LDAP among them. What is not mentioned in the article is that ASN.1 is a binary protocol and is therefore not human-readable. It may save space for bandwidth-constrained applications. However, bandwidth has a tendency to increase over time. When all wireless handhelds have a megabit of bandwidth, we would sorely regret being tied to ASN.1, as LDAP regrets it now.
Not to mention, ASN.1 does not generally reduce the document size by more than 40% compared to XML. Think about it: how much space is really taken by tags?
It's also worth noting that there is lots of documentation surrounding XML. With ASN.1 you have to download the spec from ITU which is an INCREDIBLY annoying organization and their specs are barely readable and they charge money to look at them, despite the fact that they are supposedly an open organization. The IETF and the W3C are actually open organizations; ITU just pretends to be. ITU does whatever it can to restrict the distribution of their specifications.
Take a second processor, with more pipelines available for instruction issue. Since it has more pipelines available it is able to issue more instructions while waiting for the branch to the calculated.
He was referring to pipeline length, not width. In a 20 stage processor at the same clock rate, it takes longer to fill a pipeline and consequently the branch misprediction penalty is worse.
Suppose you have two processors, each at the same clock speed. One has a 5-stage pipeline, the second a 20-stage pipeline. Suppose that there is a branch every 6 instructions (which is typical). For every mispredicted branch, the first processor need only throw away 4 instructions, but the second 19. If most branches were mispredicted, it would kill the second processor.
Pipeline length and clock speed are closely related design parameters. Longer pipes allow faster clock rates (because less is done per cycle per stage), but they increase the branch misprediction penalty. Generally there is a "happy compromise" for a processor, between pipeline length and clock speed. Most recent chips have found that happy medium to be around 10 stages. The Pentium-4 is unusual in the regard that it has 20 stages. Branch prediction therefore becomes extremely important.
Long pipelines tend to benefit Floating Point code more than Integer code, because FP is more loop-intensive, and the branches are therefore more easily predicted. This is why the P4, with its extremely long pipelines, performes poorly on integer performance compared to the PIII, but well on FP.
Oops. Sorry.
It's worth noting that Sun's MAJC architecture already has an implementation that is commercially available (MAJC 5200) and employs several of the multithreading strategies outlined in the article (including CMP and CMT).
Sun has an excellent in-depth explanation of those multithreading technologies in it's whitepaper.
Your post was emotional, distorted, incendiary, and silly. I felt like I was reading Atlas Shrugged.
That the left-wing media and so-called "scientific" establishments attack all forms of progress is nothing new. The anti-Human forces have been at it for decades, sickly raging at those who are better than them.
Umm, the article only mentioned a new theory about what caused the extinction of some species. All of this about attacking progress in the name of sickly fascism etc etc is just a figment of your paranoid imagination, and you likely imagine yourself to be the lone crusader against it. Nothing about "anti-humanism" was mentioned in the post; it was just a theory about extinction.
In our present day, the whining, half-human statists wish to use their leverage inside the corrupt, reeking organ of fascism known as the modern nation to bind as they could not bind our ancient forefathers.
I'm glad you're capable of being dispassionate and avoiding loaded terms.
It is the liberals who have passed their point of usefulness. A political Darwinism will sweep the land, raging like holy fire across a peoples weary of the lies, slander, and weakness of the liberal ideals.
You accuse your opponents of fascism?
Try to relax and think for awhile. When you find yourself getting upset, you aren't thinking.
The greatest cost to any large IT department, by far, is personnel. Every $80,000 employee costs the corporation over $120,000 in salary, taxes, benefits, etc, PER YEAR. The cost of Web development tools, web servers, etc, is maybe a few hundred or a few thousand dollars. If a company is paying over $100k a year for a developer, they don't mind paying a few thousand for a development tool.
Open source tools cost an IT deparment more _in salary_ because they are more difficult to learn. MCSEs cannot charge as much as unix admins.
There are some pieces of software that are prohibitively expensive, like Oracle. However there are no open source equivalents.
Not to mention, corporations have not shown themselves to be very eager to reduce software costs. Otherwise they would buy IBM's DB2 at $20,000 rather than Oracle8i at several million dollars. Both have roughly the same features.
The article clearly states that the technology can only differentiate between 20 or so people. As such it will only be used to tell which family member is using the computer at a given time. Therefore, the technology will only be useful to a company that is already tracking your computer but does not know which family member is currently using it. Thus, this technology will be (at best) an adjunct to more serious tracking technologies.
It is unlikely that a technology could ever be developed that could pick a single person out of the entire population, from keystrokes alone. Not that much information is revealed from keystroke timing. Therefore, this technology is far less of a threat to privacy than, say, fingerprints.
People constantly make arguments about the superior reliability of the Linux OS, especially when compared to NT. Linux has a more reliable _kernel_, but it unquestionably does not have a more reliable gui, or more reliable desktop apps. Gnome with enlightenment has constant glitches, and crashes regularly. My new Red Hat 7.1 installation (Gnome/Sawfish) hangs the entire box within 3 days no matter what box I install it on. Even the Linux mail clients, which are comparatively simple pieces of software, are very much in beta (Balsa & Evolution, for example). The desktop software (diagramming tools, documentation tools, etc) are in an extremely primitive state: development on AbiWord appears to have ceased, and Dia is not even in beta.
The sole redeeming feature is the newest release of Mozilla (0.9), which _is nice._ It has gotten dramatically better. However, this is no vindication of the Open Source development model: Mozilla development is done in a traditional "cathedral" way with a paid, professional development staff.
Aside from that one exception, the Linux desktop environment is vastly inferior to its commercial alternatives (OS X & Windows 2000).
Most game consoles are sold for well below their cost of manufacture. The console makers do this so that there is a strong enticement to buy the console. The loss generated by the console sale is compensated by increasing the price of the games: in the long run there is a hefty profit.
This works because most consumers only look at the up-front price rather than the total cost of ownership. A cheap console with expensive games sells better than an expensive console with cheap games, because the entry is so much easier.
An open platform cannot exploit this advantage, so the up-front hardware cost will necessarily be much greater. For this reason, an open console standard is unlikely to succeed.
First, the university benchmarking team simply ran lmbench (a free, popular, old kernel benchmarking utility) on a variety of kernels. Claiming that:
...somewhat exaggerates this accomplishment
Three students and a professor from Northern Michigan University spent the semester benchmarking a bunch of Linux kernels
Second, no data were presented on the main areas of the kernel that were improved. How is SMP performance in kernel space? Did the finer grained locks help? How is the performance from the threaded IP stack? Does it prevent IO blocking?
THAT kind of information would have been interesting. They tested only things that the kernel has done forever.
Tech is a field where employees have a marketable skill which they leverage to increase their salaries and benefits. This system works very well: most tech workers are paid much better than average and are treated much better than average.
The main complaint of the article is that timelines today are too short, so tech workers feel rushed. This unionization effort seems to stem from a desire of the tech workers to usurp the role of determining timelines from management. Presumably they feel the product should take much longer to produce and correspondingly have fewer bugs. Perhaps they would feel a better sense of craftsmanship if their product was more carefully engineered. However the purpose of an organization is not to further feelings of craftsmanship in their employees, but to sell products for money.
It is management that is best able to determine what is the best tradeoff between development time and bug count. A sense of craftsmanship is absolutely irrelevant. It is incredibly presumptuous for tech workers to demand that they should be able to determine product timelines, that the purpose of their employer is to make them feel craftsmanship, and that they should be paid well for doing it.
There are two fundamentally different contentions in the MS speech, which are not very well separated. They are:
...this implies that OSS should be banned or restricted.
1. OSS is not a viable model, because you need intellectualy property to charge money and pay for R&D.
2. OSS poses a threat to commercial software even for the companies that do not participate in it.
The second contention is found in this quote:
It [OSS] also fundamentally undermines the independent commercial software sector because it effectively makes it impossible to distribute software on a basis where recipients pay for the product rather than just the cost of distribution.
Oddly, the two contentions blatantly contradict each other, when they are juxtaposed explicitly. If OSS is not a viable development model, then it poses no threat to commercial software. Because if OSS is not a viable model, then it will be unable to produce a competitive product; if it does produce a competitive product, then contention #1 is mistaken.
It is also odd that it is Microsoft, not OSS, that wants to violate property rights. Fundamental to the notion of property rights is the ability to dispose of your own property in any way that pleases you, including giving it away or even destroying it. It would be a serious violation of intellectual property to prevent the owners of code from distributing it as they see fit.
This article was in the vein of CNN's crossfire: calculated to be polemical, provocative, and irrational, so as to incite discussion and readership.
For example, here is a quote about the FSF:
He [RMS] was aware of the phenomenon codified by Abraham Maslow: there are lots of people who will sign on to just about any movement in exchange for the sense of belongingness that being the proud member of a group imparts. Fair enough. Nothing wrong with that. As long as you live it.
This kind of unsupported pop psychoanalysis could be levelled against any group or organization. In this case, the evidence weighs heavily against it: whatever RMS' faults, he almost certainly believes in what he preaches. I doubt very much that RMS started the FSF to acquire needy followers, and I doubt very much that people join for a sense of belongingness. Writing code in your basement for a compiler with other people you've never met is not a sure a path to belongingness. Anyone looking for a sense of beloning could far more easily find it in a church.
The other claims are similarly weak:
Gnome is controlled -- c'mon, don't kid yourself -- by two companies
The parenthetical clause ("c'mon, don't kid yourself") is the only support offered for this statement. The statement implies that RMS is a corporate lackey, which I seriously doubt.
It's tragic that this kind of talk-show commentary has eclipsed real argumentation.
Tech support sucks because people aren't willing to pay for it.
It costs money to provide tech support. Therefore, companies that have good support must charge more for their products. This gives consumers a choice: buy from the cheaper vendor with no support, or from the more expensive vendor with adequate support. Consumers routinely choose the cheaper alternative, even when they know that the support will be terrible.
Witness the runaway success of vendors that have a reputation for having flagrantly bad service. Fry's electronics is a perfect example. Everyone who goes to Fry's ends up saying that they will never go back. Then Fry's runs an ad for RAM at $5 less than their competitors, and all those consumers who swore them off go right back to Fry's.
If people were willing to pay money for support, and people flocked to the vendors offering more service, then companies would be climbing on top of each other to offer more support. People don't actually want support, despite what they say. People say one thing and do another. What they prefer is to save the $5 and forgo the support, then bitch for 2 hours about how badly they were wronged and how bad support is nowadays, then they pocket the $5 and repeat the process.
With a commercial database, you pay for proven reliability, proven scalability, support, thorough documentation, and features.
MySql is not a database in the same sense that Oracle is. The four basic criteria that need to be met by any database are summarized by the ACID acronym (Atomicity, Consistency, Isolation, Durability). MySql satisfies none of them and lacks a huge number of very basic features. It is not a database, but a simple SQL parser on top of a DBM file.
The ACID test assures that the data will always remain in a consistent state, despite failures. Most enterprises (banks, businesses, etc) would pay any amount for this feature.
Postgres is a real database. However, it is still _very_ slow and scales poorly. The documentation for Postgres has glaring omissions, and the database lacks some features.
It is often worth paying for a commercial database. There is not a substantial difference between Oracle the "other" commerical databases (DB2, Informix, Sybase, etc). All of those databases are enterprise-class and offer features and performance similar to Oracle. Oracle charges 3-4x as much money because they are the market leader and market leaders always do that. Managers are extremely comfortable with market leaders, programmers are familiar with it, and DBAs are easy to find. That justifies the extra cost for most corporations.
It is strange that this "ask slashdot" column lumped MSSQL in with Postgres and MySql. MSSQL is not free, and an enterprise license is not cheap (>$20k if I recall). I have very little experience with MSSQL but it is generally not well regarded.
If you're just serving out static Web pages, it doesn't matter what kind of machine you have, so long as it's reliable.
I took my PowerPC 601/75mhz (LinuxPPC) with 16M of Ram and hit it with a few million httpd requests. Note that this computer is worth about $50 on the open market. It could saturate more than five T1s and serve out many millions of pages per day.
The only place a Web Server would get bogged down is on CGIs and database accesses. To what extent it gets bogged down is entirely dependent on the CGI or database; the Web serving software is not a significant part of the equation.
CGIs don't tend to be _that_ big of a deal, though. Really the rough spots are around database access. That's the only place where real money should be spent.
Cartman
twerges@hotmail.com
Uhh... I've installed Linux on seven of my machines, I never even bothered to check hardware compatibility, and it always worked just fine. All the boxes were non-brand name small-manufacturer clones.
... You saw prices of EIGHTY DOLLARS for a copy of RH 5.2?!?! I walked down to my local Software Etc. and got an _official_ RH for $39.99.
... If you don't need the box and manual you can order the unofficial version for $1.99 from www.cheapbytes.com. That's not _free_ but it's very close.
If you have anything even resembling normal hardware, there will be drivers. The problem is easily avoided by not buying extremely obscure hardware.