The author seems to be arguing that since OSS is *too* successful, it doesn't work. Didn't he get the hint when he found out about the *TWO* major OSS desktop environments, with all their corresponding programs?
I agree that there is a lot of redundancy in Open Source projects, but so what? You don't have to use them all, just try out the top five or so. For text editors, of course there is vi and emacs, and you can use them all day. But if you don't want to, you could use joe or pico, or for that matter, ed or dd+sh! I like nano (an improved pico) and RHIDE (a Borland-style IDE), but I haven't used them much lately...
However, just to make these people happy, I think we should build a Unix distribution that only allows you to have ONE of each kind of application. I suggest X, twm, mail and ed to start with, and if they don't like that, they can build a distribution that only has mwm, CDE, Z-Mail and Visual Slickedit; then, if they don't like that, they can use only Enlightenment, StarOffice, and nedit; then...
Personally, I'll just install it all instead, and use what I want. Somehow, that doesn't bother me.
First off, I'd like to say that that was indeed a great response. While I'm not sold on IBM's approach just yet, at least I'm more informed now than I was when I read the article.
The main technique I've seen exploited in DSP programming (I know someone who does this for a living) is software pipelining, which often involves like loop unrolling, except that you have to pay attention to the instructions to make sure you use all your instruction units.
Compilers these days can do loop unrolling, and past that I guess you'd just hope to be able to reorder the instructions somewhat, to get a decent instruction mix out of the code in the loop, and maximize that magical "IPC" number. However, yes, it's hard to get rid of all the dependencies, and run-time profiling (a la Transmeta) will probably get more popular as research into VLIW systems gains in popularity.
SMT sounds interesting; does it refer to "threads" in the software sense, or just separate processes? For the moment I'll view it with skepticism, just as I did when Sun built streams into the kernel. Any multiprocessing Unix system should be able to run separate processes on separate processors, and some of them can surely do the same with different threads, depending on the implementation. I guess IBM would just have a more compact and possibly more scalable solution in this case.
And yes, I realize these are supposed to be server chips for now, and knowing IBM, they might just stay that way. But this sort of technology usually filters down into the PC market quicker than you'd think, especially if it improves the price/performance ratio, as this might do, eventually.
First off, the basic news here (that IBM's Power4 architecture will be two processors running at Gigahertz speeds) isn't news; that's been known for a while.
However, it *is* nice to have this depth of technical information to examine, and also it's good to know that they're still doing this.
I think the big advantage that VLIW instruction sets will have is strictly architectural, and I'm not sure how IBM's approach fits in yet, but it looks interesting. Throwing more chips at the problem is one approach, but remember that your competitors can do that too, *and* make the chips do more as well...
However, IBM will have to make sure people design their apps with more than one processor in mind, which will be a Good Idea for the future, since more people might have multiprocessor computers. --- pb Reply or e-mail; don't vaguely moderate.
Yeah, seriously, you tell them, man. Those Slashdot Moderators have NO sense of humor whatsoever, or we wouldn't have to have mod wars over funny posts, and whatnot.
Actually, wait a minute, isn't it "Trolling At +2 Day", also known in some countries as "Friday"? --- pb Reply or e-mail; don't vaguely moderate.
The man behind the magic--literally! It must be really cool to play the characters that inspired a lot of core D&D spells. (Bigby's Crushing Hand, anyone?:)
I could care less about the "violence in gaming", and it looks like Gary Gygax agrees, too. But the rest of the interview is great. (even if I liked The Realms better, because Elminster is the man, it was all still in the D&D universe...) --- pb Reply or e-mail; don't vaguely moderate.
Well, a couple of the old core members are still around, but I wouldn't be surprised if some of the others still contributed, and eventually got changes folded in.
Really, this isn't that different from how Linux kernel development works, or CVS access in most projects are handled: only a few people are allowed to commit changes. This is just a bit more formalized. --- pb Reply or e-mail; don't vaguely moderate.
I'd like to complain about Mozilla too: lately, Mozilla has gotten too fast and stable and standards-compliant. I miss the old days when everything was obviously broken, and it was easy to find bugs.
Therefore, here is my advice to you: install Mozilla for Windows. No, no, don't install Windows, unless you already have it. Just make sure that you run Mozilla for Windows under Wine on Linux. THEN you get all your old bugs back, for free! Oh yeah, it still browses the web, but at least you have some real, obvious, fixable bugs. And then you get to help out the Wine project, too!:) --- pb Reply or e-mail; don't vaguely moderate.
Warning: MySQL Connection Failed: Can't connect to MySQL server on 'mysql.csoft.net' (111) in/export/home0/webtrade/www/dbconnect.inc on line 1
Unable to connect to database
Other than that, it looks like a really cool site, but you're not getting my 286!:) --- pb Reply or e-mail; don't vaguely moderate.
Yeah, me too! I hate PDF. "Open Standard", my ass.
Unfortunately, though, I found this stuff so fascinating that I couldn't resist! I'm a sucker for ciphers, I wrote a Vinegre cipher encoder/decoder in BASIC like 8 years ago...:)
This just builds the table, because I'm sure no one wants to see all this old code, but I'm just getting nostalgic here. I guess I was 14, reading through some tiny book on ciphers throughout history, wondering how the modern stuff worked, and later why PGP was so slow on my 386SX25...
DIM ALPHABET$(25, 25)
FOR Y = 0 TO 25
FOR Z = 0 TO 25
A = Y + Z + 65
IF Y + Z > 25 THEN A = Y + Z - 26 + 65
ALPHABET$(Y, Z) = CHR$(A)
NEXT Z
NEXT Y
I wish Microsoft would apologize to me, at the least for the spam they send me occasionally about their "Freedom to Innovate" crud.
Also, I want to see more of these web pads! I could care less about laptops, but I think surfing the web remotely on a slim little pad would be *the* killer app. (and it would definitely be a better use of a Crusoe processor, too:) --- pb Reply or e-mail; don't vaguely moderate.
Well, we've just gone full circle now, haven't we?
Starting with the punch cards inspired by the textile industry, and using the icons inspired by embroidery... Now we're using a pocket gaming system to do the original functions we copied!
I guess that's a tribute to our history, albeit a sick one. --- pb Reply or e-mail; don't vaguely moderate.
I've been wondering for a while what "Digital DNA" is; I guess that's it! It's "DNA of the Digital Kind"--personal information that identifies you, except digitally.
I thought it was just a dumb marketing ploy for Motorola technology, but now I know it's an even dumber ploy for Motorola marketing. Well, my answer is the same for both kinds of DNA: it's mine; hands off. --- pb Reply or e-mail; don't vaguely moderate.
I've been interested in this project for a while now, and it looks like (according to the TODO list) they've fixed a lot of gotchas, and now it's time to hammer out the bugs and optimize.
However, I still like the idea of running a virtual x86 machine better, for the OS compatibility. Running a virtualized Linux would be cool for security, (run sendmail and wuftpd and whatnot on top of that; if someone hacks it, just kill and respawn:) and of course development, but that's about it. --- pb Reply or e-mail; don't vaguely moderate.
Actually, I was wondering if 666 came from the old Greek dice games where they would roll three dice; it was considered lucky to roll "triple-sixes", as that was a high score.
Of course, if you're looking for conspiracy theories, then you can find 666 in role-playing, "www", UPC bar-codes, and pretty much everything else, too. (29Ah; ha ha ha!) --- pb Reply or e-mail; don't vaguely moderate.
First, if there's a hole in a product that has been found, the company has been notified, and nothing has been done about it for 45 days, I'm sure someone has and is using an exploit by then.
Second, by publishing the details, and saying "hey, you knew about this for 45 days; don't you think your customers should know?", you're encouraging software companies to get their act together.
If left to themselves, no, they won't fix bugs out of the goodness of their hearts. The only people this sort of thing affects are the consumers; big businesses have firewalls and probably use better, more expensive products internally, wherever possible.
I think it's pretty sad that it has to be this way, but that's the way it is. Taking a laissez-faire attitude to big software companies doesn't seem to work, because there is too much potential for abuse.
I also don't like the whole "unauthorized negative reviews aren't allowed" business. Who cares about freedom of speech, eh? We can have unauthorized biographies of Bill Gates, and not unauthorized reviews of Front Page. Whatever you say, guys. For example, I did a lot of benchmarking between DOS and Linux's DOSEmu; my findings at the time were that DOSEmu was about 3% slower in raw CPU than actual native DOS (testing using the DOS 32-bit BYTEMarks), but that defragging a native DOS partition was *much* faster under DOSEmu, due to Linux's cache subsystem. Those are the facts; why should they be censored just because someone else isn't happy about it? --- pb Reply or e-mail; don't vaguely moderate.
Why, when you get to make a standard, would you make it *intentionally* incompatible everywhere? It's bad enough with TVs and cellphones, since the US and Europe consistently manage to pick different standards, but why do this when you *have* a choice?
I know this is about control and greed, and not about common sense. Therefore, why do we let these people make the standards? We need an independent group of experts deciding these things, a standards board releasing drafts, and companies implementing the drafts. Period. Also, there can be no patents held on the standards, or if there are patents, they can't be used to restrict an implementation of a draft or standard.
I don't care if this has to be a government regulated activity, or if it takes a little longer, (but not too much longer; that's why we need the drafts) but we can't let the corporations mess up important technology for their own personal gain. The web browser standards war was useless enough, and that didn't have nearly as much money riding on it. --- pb Reply or e-mail; don't vaguely moderate.
Oh, indeed. There's nothing wrong with testing that, but the out of the box settings often suck for both of them, and anyone doing real work will have to tune something. For instance, you never get security out of the box, and you often don't get speed; the initial settings are pretty newbie-friendly.
So, yes, it would be good to use the out-of-the-box settings for some sort of ease-of-use test, but not for a real benchmarking test. --- pb Reply or e-mail; don't vaguely moderate.
Heck, lets go back to flaming ZD-Net, and telling Jesse Berst "I bet you're getting paid by Microsoft to say that"...:)
Well, yeah; arla definitely needs a lot of work before that happens, but it makes an okay client now; I understand OpenLDAP is good, but I never had to implement it. However, the distributions have the resources to sink into these projects, easily. And back in the day, Caldera would have just paid for the real stuff, from Netscape or Transarc... --- pb Reply or e-mail; don't vaguely moderate.
I seem to remember that, actually, but I've gotten too lazy on slashdot to post links lately.
However, there you have it folks, Linux rules on a real test--a mix of static and dynamic content--even on the evil mindcraft test platform.
Mad props to IntlHarvester for his MLP skillz; mod him up for bothering to find the link!
And since I'm also too lazy to reply to your other post: if I want a scalable networked filesystem with ACLs on Linux, I'd use AFS, or just use arla, and get it to work with AFS. If I wanted security, I'd use kerberos on top of that. And since I'm at NCSU, I currently do all that anyhow to integrate with the network.
The NT boxes on the network do the same, but the interface is lousy and not available to the students, the users have to type their passwords in twice and wait an ungodly amount of time for everything, and the software is more expensive and takes up more resources. Hopefully they'll switch those boxes back over, too, now that Linux is a supported platform here. --- pb Reply or e-mail; don't vaguely moderate.
That page wasn't accurate *then*; I'm sure you can look up several refutations of it.
However, I'd be interested in seeing another round of benchmarks, especially between Linux 2.4.0 (preferably whenever it's officially released) and Windows 2000. Also, a comparison between the many actual journaling filesystems that run under Linux and the one (not really) journaling filesystem that runs on NT/2000 would be interesting; heck, include BeOS in that one, too!:)
Ground rules: use identical hardware, use stable, recent software, tune it all as much as possible, and test with multiple hardware configurations, (i.e. not just 4 Processors and 4 NIC cards; that's not terribly realistic in the first place) to figure out what is the best overall solution.
Also, when determining web server performance, make sure there are some tasks that are CPU-bound or IO-bound as well as simply network-bound; you want to test everything.
That all having been said, I'm pretty sure that a lot of the Mindcraft-specific differences have been fixed in 2.4, and therefore Linux should perform significantly better on those sorts of tests than it did in the past, as well. I'd rather see some more well-rounded tests constructed to go along with that, though. --- pb Reply or e-mail; don't vaguely moderate.
Slashdot Mirror
The author seems to be arguing that since OSS is *too* successful, it doesn't work. Didn't he get the hint when he found out about the *TWO* major OSS desktop environments, with all their corresponding programs?
I agree that there is a lot of redundancy in Open Source projects, but so what? You don't have to use them all, just try out the top five or so. For text editors, of course there is vi and emacs, and you can use them all day. But if you don't want to, you could use joe or pico, or for that matter, ed or dd+sh! I like nano (an improved pico) and RHIDE (a Borland-style IDE), but I haven't used them much lately...
However, just to make these people happy, I think we should build a Unix distribution that only allows you to have ONE of each kind of application. I suggest X, twm, mail and ed to start with, and if they don't like that, they can build a distribution that only has mwm, CDE, Z-Mail and Visual Slickedit; then, if they don't like that, they can use only Enlightenment, StarOffice, and nedit; then...
Personally, I'll just install it all instead, and use what I want. Somehow, that doesn't bother me.
---
pb Reply or e-mail; don't vaguely moderate.
First off, I'd like to say that that was indeed a great response. While I'm not sold on IBM's approach just yet, at least I'm more informed now than I was when I read the article.
The main technique I've seen exploited in DSP programming (I know someone who does this for a living) is software pipelining, which often involves like loop unrolling, except that you have to pay attention to the instructions to make sure you use all your instruction units.
Compilers these days can do loop unrolling, and past that I guess you'd just hope to be able to reorder the instructions somewhat, to get a decent instruction mix out of the code in the loop, and maximize that magical "IPC" number. However, yes, it's hard to get rid of all the dependencies, and run-time profiling (a la Transmeta) will probably get more popular as research into VLIW systems gains in popularity.
SMT sounds interesting; does it refer to "threads" in the software sense, or just separate processes? For the moment I'll view it with skepticism, just as I did when Sun built streams into the kernel. Any multiprocessing Unix system should be able to run separate processes on separate processors, and some of them can surely do the same with different threads, depending on the implementation. I guess IBM would just have a more compact and possibly more scalable solution in this case.
And yes, I realize these are supposed to be server chips for now, and knowing IBM, they might just stay that way. But this sort of technology usually filters down into the PC market quicker than you'd think, especially if it improves the price/performance ratio, as this might do, eventually.
---
pb Reply or e-mail; don't vaguely moderate.
Well, the "NEW INFORMATION" you seek would obviously be my opinion on the subject.
However, since you mention another source of information, would you be so kind as to post a link, perhaps?
---
pb Reply or e-mail; don't vaguely moderate.
First off, the basic news here (that IBM's Power4 architecture will be two processors running at Gigahertz speeds) isn't news; that's been known for a while.
However, it *is* nice to have this depth of technical information to examine, and also it's good to know that they're still doing this.
I think the big advantage that VLIW instruction sets will have is strictly architectural, and I'm not sure how IBM's approach fits in yet, but it looks interesting. Throwing more chips at the problem is one approach, but remember that your competitors can do that too, *and* make the chips do more as well...
However, IBM will have to make sure people design their apps with more than one processor in mind, which will be a Good Idea for the future, since more people might have multiprocessor computers.
---
pb Reply or e-mail; don't vaguely moderate.
"AS it says in the bible, ' there shall be no ice cube wide enough, and no ice cube tall enough to shut up a nerd. And there was much rejoicing' "
:)
Amen, Brother!
Even if slashdot keeps posting crap like this, and even though these posts are at -1, it's good to still see that MEEPT!! is still around.
...but now it's time to go to kuro5hin, and see what the news is.
---
pb Reply or e-mail; don't vaguely moderate.
Real hackers will still just use a chainsaw.
:)
Let me know when use this technique on something a little more permanent, like plastic.
---
pb Reply or e-mail; don't vaguely moderate.
Yeah, seriously, you tell them, man. Those Slashdot Moderators have NO sense of humor whatsoever, or we wouldn't have to have mod wars over funny posts, and whatnot.
Actually, wait a minute, isn't it "Trolling At +2 Day", also known in some countries as "Friday"?
---
pb Reply or e-mail; don't vaguely moderate.
Well, that's actually pretty cool... Thanks for the info!
---
pb Reply or e-mail; don't vaguely moderate.
The man behind the magic--literally! It must be really cool to play the characters that inspired a lot of core D&D spells. (Bigby's Crushing Hand, anyone? :)
I could care less about the "violence in gaming", and it looks like Gary Gygax agrees, too. But the rest of the interview is great. (even if I liked The Realms better, because Elminster is the man, it was all still in the D&D universe...)
---
pb Reply or e-mail; don't vaguely moderate.
Well, a couple of the old core members are still around, but I wouldn't be surprised if some of the others still contributed, and eventually got changes folded in.
Really, this isn't that different from how Linux kernel development works, or CVS access in most projects are handled: only a few people are allowed to commit changes. This is just a bit more formalized.
---
pb Reply or e-mail; don't vaguely moderate.
That is some UGLY crap!
:)
I'm *so* glad that the NES didn't look like this back in the day, or I *never* would have gotten one.
P.S. Mad props out to post #25, the first Sneakers reference I saw!
---
pb Reply or e-mail; don't vaguely moderate.
I'd like to complain about Mozilla too: lately, Mozilla has gotten too fast and stable and standards-compliant. I miss the old days when everything was obviously broken, and it was easy to find bugs.
:)
Therefore, here is my advice to you: install Mozilla for Windows. No, no, don't install Windows, unless you already have it. Just make sure that you run Mozilla for Windows under Wine on Linux. THEN you get all your old bugs back, for free! Oh yeah, it still browses the web, but at least you have some real, obvious, fixable bugs. And then you get to help out the Wine project, too!
---
pb Reply or e-mail; don't vaguely moderate.
Warning: MySQL Connection Failed: Can't connect to MySQL server on 'mysql.csoft.net' (111) in /export/home0/webtrade/www/dbconnect.inc on line 1
:)
Unable to connect to database
Other than that, it looks like a really cool site, but you're not getting my 286!
---
pb Reply or e-mail; don't vaguely moderate.
Yeah, me too! I hate PDF. "Open Standard", my ass.
:)
Unfortunately, though, I found this stuff so fascinating that I couldn't resist! I'm a sucker for ciphers, I wrote a Vinegre cipher encoder/decoder in BASIC like 8 years ago...
This just builds the table, because I'm sure no one wants to see all this old code, but I'm just getting nostalgic here. I guess I was 14, reading through some tiny book on ciphers throughout history, wondering how the modern stuff worked, and later why PGP was so slow on my 386SX25...
DIM ALPHABET$(25, 25)
FOR Y = 0 TO 25
FOR Z = 0 TO 25
A = Y + Z + 65
IF Y + Z > 25 THEN A = Y + Z - 26 + 65
ALPHABET$(Y, Z) = CHR$(A)
NEXT Z
NEXT Y
---
pb Reply or e-mail; don't vaguely moderate.
A company apologizes for Spam. What a concept!
:)
I wish Microsoft would apologize to me, at the least for the spam they send me occasionally about their "Freedom to Innovate" crud.
Also, I want to see more of these web pads! I could care less about laptops, but I think surfing the web remotely on a slim little pad would be *the* killer app. (and it would definitely be a better use of a Crusoe processor, too
---
pb Reply or e-mail; don't vaguely moderate.
Well, we've just gone full circle now, haven't we?
Starting with the punch cards inspired by the textile industry, and using the icons inspired by embroidery... Now we're using a pocket gaming system to do the original functions we copied!
I guess that's a tribute to our history, albeit a sick one.
---
pb Reply or e-mail; don't vaguely moderate.
I've been wondering for a while what "Digital DNA" is; I guess that's it! It's "DNA of the Digital Kind"--personal information that identifies you, except digitally.
I thought it was just a dumb marketing ploy for Motorola technology, but now I know it's an even dumber ploy for Motorola marketing. Well, my answer is the same for both kinds of DNA: it's mine; hands off.
---
pb Reply or e-mail; don't vaguely moderate.
I've been interested in this project for a while now, and it looks like (according to the TODO list) they've fixed a lot of gotchas, and now it's time to hammer out the bugs and optimize.
:) and of course development, but that's about it.
However, I still like the idea of running a virtual x86 machine better, for the OS compatibility. Running a virtualized Linux would be cool for security, (run sendmail and wuftpd and whatnot on top of that; if someone hacks it, just kill and respawn
---
pb Reply or e-mail; don't vaguely moderate.
Actually, I was wondering if 666 came from the old Greek dice games where they would roll three dice; it was considered lucky to roll "triple-sixes", as that was a high score.
Of course, if you're looking for conspiracy theories, then you can find 666 in role-playing, "www", UPC bar-codes, and pretty much everything else, too. (29Ah; ha ha ha!)
---
pb Reply or e-mail; don't vaguely moderate.
First, if there's a hole in a product that has been found, the company has been notified, and nothing has been done about it for 45 days, I'm sure someone has and is using an exploit by then.
Second, by publishing the details, and saying "hey, you knew about this for 45 days; don't you think your customers should know?", you're encouraging software companies to get their act together.
If left to themselves, no, they won't fix bugs out of the goodness of their hearts. The only people this sort of thing affects are the consumers; big businesses have firewalls and probably use better, more expensive products internally, wherever possible.
I think it's pretty sad that it has to be this way, but that's the way it is. Taking a laissez-faire attitude to big software companies doesn't seem to work, because there is too much potential for abuse.
I also don't like the whole "unauthorized negative reviews aren't allowed" business. Who cares about freedom of speech, eh? We can have unauthorized biographies of Bill Gates, and not unauthorized reviews of Front Page. Whatever you say, guys. For example, I did a lot of benchmarking between DOS and Linux's DOSEmu; my findings at the time were that DOSEmu was about 3% slower in raw CPU than actual native DOS (testing using the DOS 32-bit BYTEMarks), but that defragging a native DOS partition was *much* faster under DOSEmu, due to Linux's cache subsystem. Those are the facts; why should they be censored just because someone else isn't happy about it?
---
pb Reply or e-mail; don't vaguely moderate.
Why, when you get to make a standard, would you make it *intentionally* incompatible everywhere? It's bad enough with TVs and cellphones, since the US and Europe consistently manage to pick different standards, but why do this when you *have* a choice?
I know this is about control and greed, and not about common sense. Therefore, why do we let these people make the standards? We need an independent group of experts deciding these things, a standards board releasing drafts, and companies implementing the drafts. Period. Also, there can be no patents held on the standards, or if there are patents, they can't be used to restrict an implementation of a draft or standard.
I don't care if this has to be a government regulated activity, or if it takes a little longer, (but not too much longer; that's why we need the drafts) but we can't let the corporations mess up important technology for their own personal gain. The web browser standards war was useless enough, and that didn't have nearly as much money riding on it.
---
pb Reply or e-mail; don't vaguely moderate.
Oh, indeed. There's nothing wrong with testing that, but the out of the box settings often suck for both of them, and anyone doing real work will have to tune something. For instance, you never get security out of the box, and you often don't get speed; the initial settings are pretty newbie-friendly.
So, yes, it would be good to use the out-of-the-box settings for some sort of ease-of-use test, but not for a real benchmarking test.
---
pb Reply or e-mail; don't vaguely moderate.
Heck, lets go back to flaming ZD-Net, and telling Jesse Berst "I bet you're getting paid by Microsoft to say that"... :)
Well, yeah; arla definitely needs a lot of work before that happens, but it makes an okay client now; I understand OpenLDAP is good, but I never had to implement it. However, the distributions have the resources to sink into these projects, easily. And back in the day, Caldera would have just paid for the real stuff, from Netscape or Transarc...
---
pb Reply or e-mail; don't vaguely moderate.
I seem to remember that, actually, but I've gotten too lazy on slashdot to post links lately.
However, there you have it folks, Linux rules on a real test--a mix of static and dynamic content--even on the evil mindcraft test platform.
Mad props to IntlHarvester for his MLP skillz; mod him up for bothering to find the link!
And since I'm also too lazy to reply to your other post: if I want a scalable networked filesystem with ACLs on Linux, I'd use AFS, or just use arla, and get it to work with AFS. If I wanted security, I'd use kerberos on top of that. And since I'm at NCSU, I currently do all that anyhow to integrate with the network.
The NT boxes on the network do the same, but the interface is lousy and not available to the students, the users have to type their passwords in twice and wait an ungodly amount of time for everything, and the software is more expensive and takes up more resources. Hopefully they'll switch those boxes back over, too, now that Linux is a supported platform here.
---
pb Reply or e-mail; don't vaguely moderate.
That page wasn't accurate *then*; I'm sure you can look up several refutations of it.
:)
However, I'd be interested in seeing another round of benchmarks, especially between Linux 2.4.0 (preferably whenever it's officially released) and Windows 2000. Also, a comparison between the many actual journaling filesystems that run under Linux and the one (not really) journaling filesystem that runs on NT/2000 would be interesting; heck, include BeOS in that one, too!
Ground rules: use identical hardware, use stable, recent software, tune it all as much as possible, and test with multiple hardware configurations, (i.e. not just 4 Processors and 4 NIC cards; that's not terribly realistic in the first place) to figure out what is the best overall solution.
Also, when determining web server performance, make sure there are some tasks that are CPU-bound or IO-bound as well as simply network-bound; you want to test everything.
That all having been said, I'm pretty sure that a lot of the Mindcraft-specific differences have been fixed in 2.4, and therefore Linux should perform significantly better on those sorts of tests than it did in the past, as well. I'd rather see some more well-rounded tests constructed to go along with that, though.
---
pb Reply or e-mail; don't vaguely moderate.