those bases were consolidated into nearby bases... completely different than flat-out closing the base.
Sort of. Some bases were flat out closed, but of course the materiel located on them will merge with other bases (e.g. a bomber wing giving its bombs to another wing, mechanics scattering to other bases, etc). My wing (ESC) is currently separated on three bases in Ohio, Alabama, and Massachussettes (sic). The BRAC report suggests merging them together, but I hardly consider AL and MA to be "nearby."
Anyway, several people in this thread seem to think the BRAC exists to save money so the government can spend money on militarizing space. I call bullshit. We are spending approximately four gajillion dollars per day over in Iraq (Operation Iraqi Liberation), which is a HUGE money sink. If anything, the BRAC is going to save money so Dubya can play general with his "plastic" army men and invade other countries. Because we all know that the reality of war is that it is Someone Else's Problem when people die -- not the politicians'.
most browser exploits are buffer overflows - giving wierd input. To combat that you have to add code (input checking), not change code.
Yes, but what happens when that input checking is not completely tested and needs a fix? What happens when that input checking code needs to accomodate multiple inputs? There is more to browser exploits than buffer overflows. Certain types of malformed input can cause this too, e.g. "/../../winnt/" in a URL in older versions of IIS.
My point is that input validation is more complex than "buffer overruns," and sometimes that code itself needs to be modified, resulting in the scenario in my earlier post.
It's not too bad a plan, athough I'd like to see more unmanned missions in the works.
Manned Shuttle missions are much more expensive than unmanned missions (e.g. using rockets instead of the Shuttle to launch ISS modules). Since somebody would rather fight a war on the other side of the planet, it looks like NASA has a lot of unmanned flight in its future. While I would rather privitize or flat out eliminate most of the federal government (yes, I am one of those crazy Libertarians), NASA is one of the areas where I think the government needs to increase spending and have more explorations and missions.
I don't see how basically a patch against what is most often just a few lines of code can open more holes, either. That's just dumb.
I see you have never worked on an enterprise-class application, otherwise you would know that just changing the boolean algebra inside an if() statement can have catastrophic consequences. Usually what happens is there is a bug. To fix this bug, the developer must modify this conditional (i.e. a transaction is not always processing because the if() skips it under weird circumstances). However, there is some obscure requirement that, despite being well-documented, is difficult to understand. That if() statement has conflicting requirements, and the logic needs to be expanded to accomodate both situations. However, desparate for a quick, one line fix, the developer changes a single line (or character, e.g. "!" not logic). This breaks a bunch of other stuff.
Some applications are like a house of cards -- precariously perched, even one small error can bring the whole structure down. Good configuration and requirements management can mitigate this risk, but the possibility of error is always there.
assuming you're a fellow, how about i come over to your house and "share" your wife / partner? for thousands of years women have been the common property of men in western society.
The difference is that music, stories, art, etc. can be copied. A man's wife cannot. The analogy breaks down with physical objects, which is not really the point here anyway -- we are talking about music, art, and other intangibles and their effects on society.
I think part of the issue is that as a group, Americans value entertainment more than education. Sure, copyrights might be part of the problem, but keep in mind that math textbooks are copyrighted even if the mathematics itself is not protected by IP law.
She looks like a 25 cent crackwhore. I don't have the link but there is a site that shows a woman who started out being hot, but a life of drugs and prostitution turned her into a beast that is an instant erection-stopper. O'Gara looks like that woman about halfway through her transformation.
Please, Maureen, if you are reading these comments, do us all a favor and stop taking cocaine. Then get some plastic surgery. Oh, and stay away from my children. You are scary!
All three are liberal-leaning sites. Instead of giving into a profit-first mentality, they give into the political-party-first mentality. Every place has a bias of some sort. That includes Slashdot, which is owned by OSTG, a Linux company.
Left, right, they're all the same: politcal zealots who cannot see past the ends of their own noses.
I think the real issue is reporting facts, regardless of slant. I for one read several news sources and look for the common threads. Usually this leaves the facts, although sometimes even then the leftovers are awfully suspicious. However, there are the non-stories like I suspect O'Gara writes (I haven't read them but her reputation precedes her), the same sort of useless bullshit that the "E" channel broadcasts 24 hours a day. There are no facts, and if there are, they are useless facts, especially in a trade magazine.
I think it's worth mentioning here that the copyright system punishes and rewards in such a way that that promotes hype over substance. I think it's unfair to "blame society", while at the same time holding this system of punishment and reward in place. In a copyright society, it is always the information that turns the most heads that gets the most money, where in a non copyright world the information that has the most value is rewarded the most.
Could you give some examples? I don't think I am making the connection you are trying to explain, maybe that's just because I cannot comprehend life without copyrights. It's not like I am 400 years old;-)
TSMC and UMC are not a monopoly per se, but their owners do get to dictate the direction of the industry.
That makes them a "cartel," like the MPAA, RIAA, OPEC, et al. They are still as corrupt to the bone as a monopoly, but have a low chance that a cartel member will break ranks.
Strange, I would think the GOP would really want that power.
Why do they need it? It is not like the Democrats can get any legislation passed through Congress. The Republicans are all, for the most part, in bed with each other and scratch each others' backs even if a particular bill is not 100% in line with the GOP platform.
I assume from your post that you've never tried to call someone from a concert or a party (hey this is/. after all) where there's loud music. There really are times when you can't hear shit.
There are times when the ability to send a quick, silent message is important as well. For example, sometimes I am in requirements meetings and need to send a quick message to my wife (e.g. I'll grab Subway and be home for lunch in 30 minutes). Another example is the golf course. If my wife needs to tell me something urgent, she can TXT me and I will get the message without disturbing other golfers.
You are correct, speed and convenience are not the only factors. Sometimes the surrounding environment dictates which type of message we use. I used to think TXT was for 14 year old girls at the mall, but my wife and I use it every day and I find it gets the message across quickly and without interrupting my workflow. At $5 for 1,000 messages per month it is cheap, too.
Re:The performance of compiled code
on
A Review of GCC 4.0
·
· Score: 3, Interesting
I think that companies should re-evaluate their "need" for an extra 5% performance. Here's an idea -- if you need something 10 minutes faster, why not start the process 10 minutes sooner?
In any large organization, the process gets in the way. Some suit decides the product needs a new feature, or needs to ship sooner, or whatever, and this slowly trickles down to the developers who suddenly are put in crunch time where every minute counts. Schedules and deadlines may change daily. People's jobs may be at risk. Shit happens.
Nobody really likes it, but that is sometimes how we arrive at the point where we "need" an extra 5% performance, where we "need" the program to finish ten minutes sooner. Starting earlier is not always an option, usually because you don't know you even have to start *at all* until the last minute.
Doesn't let you run as non-admin Requires the disk
These are my two biggest beefs with Neverwinter Nights. For a game that took years to make and is as big as it is, common sense (yeah, I know) would say they would follow good design and integrate with the OS the way Win2K/XP intended it to. Nope. It stores my settings, characters, and save games in the "Program Files" directory instead of the "Documents and Settings" directory. This is like storing save games in/usr instead of/home. This forces either running as admin or creative use of directory permissions. Completely retarded. Why would a Windows version of Jade Empire be any different?
Running with the CD is a hassle because I have to switch between CDs. If I leave my computer for a bathroom break, I always get nervous that my son (18 months) will grab the CD and try to "play" with it. He already know how to open the drive. He loves to stand on top of toys, no matter how big or small. I guess he is jealous that I am three times his height:-)
Just go ahead and use the long names, that alone will reduce documentation/commenting
I think variable names should answer "what," not "how." As in, what does this variable represent? Comments describe how it is used. Using concise (not necessarily short) variable names make sure you can look at a variable and say "hey, that's a window handle, that other one is a user's login, etc." Variables may have more complex usage, but usually those will be classes and have their own documentation anyway.
Exactly -- the upper limits on Windows x64 seem high now, but in a few years some of us will actually hit those limits.
I would be more interested in comparing Win64 to 64-bit Linux, BSD, Solaris, etc. How does performance compare? Hardware limits? I am off to see the wizard, the wonderful wizard of google...
While the digital media maybe flimsy, there is no reasonable reason why the information therein should _not_ survive for more than 40 years.
Actually, there are some reasons. First, the data density in books is far less than on a CD or DVD. One word in a book may be measured in millimeters, while the same word encoded in binary on a digital medium would be measured in nanometers. Smudge a book and I guarantee you can read enough of the text and figure out the smudged parts to get 95% or better data retention. Smudge the bits on a CD, and you might lose the whole thing, which could be a hundred books worth of data.
Another reason is the quality of the inks and dyes. Older books last because they used thicker paper and tough dyes using "old" technology that lasts, sort of like old lead paint on cars. I had a 1986 Ford where the paint had all worn off, leaving it a dull grayish color. I had a tough time damaging the "paint" that was left over. I nicked the edge of the bed on my truck with a piece of plastic, leaving a small gouge in the paint. Modern books do not use high quality inks, partly because of costs, partly because of environmental concerns. The same is true for CDs. If you burn your digital archives to a true CD, they will last a while, at least 20 years if stored in a climate controlled vault (we are talking libraries/archives here, not a CD sitting on your desk at home). My brother has CDs from the 1980s, some of the first CDs ever released, and they play fine even though they were stored in adverse conditions most of their lives. However, CD-Rs operate by marking dyes on the aluminum platter, which is also exposed to air because it sits on top the plastic, not inside it. Don't trust CD-Rs for long-term storage.
I backup my data every week or two, and destroy old CDs every few months. I know I cannot depend on CD-Rs to last more than a year or two, so I don't. This also gives me an opportunity to keep my data in modern formats, as I convert them every few years when some great new format becomes widespread.
Yes, but keep in mind that back in high school she played on the boys' football team (and was a star player -- I like my women mean:-) ), and she reserves her private software interfaces for me -- everyone else gets to use the java.lang.fist interface.
I *never* partition my drives. System performance is severely limited by hard drive latency and you're insuring many loooong trips across the drive face. Even if you have ample RAM, there is always *some* swapping going on.
Partitioning makes sense if you can separate static and dynamic data. For example, my Linux box has a/usr (including/usr/local) partition, and one each for/var and/tmp. There are a few other partitions but these three illustrate my point. Obviously,/usr does not change much, so fragmentation is low./var and/tmp change often, so fragmentation is high. Separate partitions ensure the fragmentation stays localized, and in the event of a catastrophic filesystem error (not hardware error), I only have to fix a small portion of my system.
Windows is a bit different. Microsoft "borrowed" the logical view of the Unix filesystem structure with the whole "My Computer" concept, but it is still a drive-oriented physical structure under the covers. However, by moving Program Files to its own partition (and hacking the registry to make it work correctly), Documents and Settings, and the swap file, you get essentially the same thing as in Unix: you can separate static and dynamic data.
Of course, in Linux or Windows you can just create one huge partition that contains everything (well, Linux does need a dedicated swap partition), and it will work. I think the whole idea of partitioning is just for performance tweaking more than anything.
There are ways to allow a program to degrade gracefully depending on which version of Windows it is currently running on. If you detect support for NTFS, for example, you know it is NT-based with all the user logins, home directories, and ACLs. Otherwise, use the old ugly hacks and encourage your users to upgrade to NT-based Windows or Linux.
Also, %HOMEDIR% is just one method. Ideally you would use API calls to grab the directories you need. For example, there is one that returns the "Application Data" directory in a user's home directory. All you have to do is strcat() your application's name you are set. Then alter your message loop (or slots in Qt, or whatever MFC uses) to detect user-login messages. If you detect a user switches, you can save off the user's data first then set a flag or something so when the user switches back you can reload the data.
If you want more information send me an email (my address is unobfuscated in the message header) and I can dredge my MSDN links for specific information.
The article says it's not optimal to put your swap on a 2nd partition of your master drive.
Okay, maybe I was not too explicit, but my Windows box has two physical drives on two separate P-IDE channels (until I can afford a new SATA drive). Swap is the first partition on my secondary drive.
Even if it wasn't, having my swap file in a single contiguous chunk on a partition in the middle of my hard drive is okay performance, even by what the article says.
Keep in mind that I believe in the/usr and/home idea of Unix, so I keep programs and data separate in Windows. Sure, I am not the typical user of Windows that brings his computer home from Worst Buy or CompUSA, but I am a nerd, and I do believe at least a little bit in the spirit of Unix:-)
Set the registry key to ignore "out of space" warnings for that drive....
What registry key is that?
Annoyances.org has a page about this, but I know there is a way to specify this on a per-drive basis. I don't know for sure, but TweakUI might help. I did search my registry for my "s: (swap) drive" just to be 100% sure, but came up with a bunch of results for "Neverwinter Nights: subtopic" so I just gave up after about 20 results. I know I have it set up on my system just for that one drive. Sorry.
Yes, it can be a lot of work running as non-admin. But there's a certainly feeling of secureness knowing that half the worms and trojans out there can't get very far without admin rights too.
I run all my computers behind a cable modem NAT firewall. Though simple, it stops 99% of malware at the gate. Windows firewall and paranoia stops the other 1% (I run Firefox, OpenOffice, etc. The only non-OSS I run is Windows itself). Even if this box is compromised, I turn 90 degrees to my right and use my Linux box: instant security:-)
Slashdot Mirror
those bases were consolidated into nearby bases... completely different than flat-out closing the base.
Sort of. Some bases were flat out closed, but of course the materiel located on them will merge with other bases (e.g. a bomber wing giving its bombs to another wing, mechanics scattering to other bases, etc). My wing (ESC) is currently separated on three bases in Ohio, Alabama, and Massachussettes (sic). The BRAC report suggests merging them together, but I hardly consider AL and MA to be "nearby."
Anyway, several people in this thread seem to think the BRAC exists to save money so the government can spend money on militarizing space. I call bullshit. We are spending approximately four gajillion dollars per day over in Iraq (Operation Iraqi Liberation), which is a HUGE money sink. If anything, the BRAC is going to save money so Dubya can play general with his "plastic" army men and invade other countries. Because we all know that the reality of war is that it is Someone Else's Problem when people die -- not the politicians'.
most browser exploits are buffer overflows - giving wierd input. To combat that you have to add code (input checking), not change code.
Yes, but what happens when that input checking is not completely tested and needs a fix? What happens when that input checking code needs to accomodate multiple inputs? There is more to browser exploits than buffer overflows. Certain types of malformed input can cause this too, e.g. "/../../winnt/" in a URL in older versions of IIS.
My point is that input validation is more complex than "buffer overruns," and sometimes that code itself needs to be modified, resulting in the scenario in my earlier post.
It's not too bad a plan, athough I'd like to see more unmanned missions in the works.
Manned Shuttle missions are much more expensive than unmanned missions (e.g. using rockets instead of the Shuttle to launch ISS modules). Since somebody would rather fight a war on the other side of the planet, it looks like NASA has a lot of unmanned flight in its future. While I would rather privitize or flat out eliminate most of the federal government (yes, I am one of those crazy Libertarians), NASA is one of the areas where I think the government needs to increase spending and have more explorations and missions.
I don't see how basically a patch against what is most often just a few lines of code can open more holes, either. That's just dumb.
I see you have never worked on an enterprise-class application, otherwise you would know that just changing the boolean algebra inside an if() statement can have catastrophic consequences. Usually what happens is there is a bug. To fix this bug, the developer must modify this conditional (i.e. a transaction is not always processing because the if() skips it under weird circumstances). However, there is some obscure requirement that, despite being well-documented, is difficult to understand. That if() statement has conflicting requirements, and the logic needs to be expanded to accomodate both situations. However, desparate for a quick, one line fix, the developer changes a single line (or character, e.g. "!" not logic). This breaks a bunch of other stuff.
Some applications are like a house of cards -- precariously perched, even one small error can bring the whole structure down. Good configuration and requirements management can mitigate this risk, but the possibility of error is always there.
assuming you're a fellow, how about i come over to your house and "share" your wife / partner? for thousands of years women have been the common property of men in western society.
The difference is that music, stories, art, etc. can be copied. A man's wife cannot. The analogy breaks down with physical objects, which is not really the point here anyway -- we are talking about music, art, and other intangibles and their effects on society.
I think part of the issue is that as a group, Americans value entertainment more than education. Sure, copyrights might be part of the problem, but keep in mind that math textbooks are copyrighted even if the mathematics itself is not protected by IP law.
You haven't seen Maureen O'Gara have you...
She looks like a 25 cent crackwhore. I don't have the link but there is a site that shows a woman who started out being hot, but a life of drugs and prostitution turned her into a beast that is an instant erection-stopper. O'Gara looks like that woman about halfway through her transformation.
Please, Maureen, if you are reading these comments, do us all a favor and stop taking cocaine. Then get some plastic surgery. Oh, and stay away from my children. You are scary!
All three are liberal-leaning sites. Instead of giving into a profit-first mentality, they give into the political-party-first mentality. Every place has a bias of some sort. That includes Slashdot, which is owned by OSTG, a Linux company.
Left, right, they're all the same: politcal zealots who cannot see past the ends of their own noses.
I think the real issue is reporting facts, regardless of slant. I for one read several news sources and look for the common threads. Usually this leaves the facts, although sometimes even then the leftovers are awfully suspicious. However, there are the non-stories like I suspect O'Gara writes (I haven't read them but her reputation precedes her), the same sort of useless bullshit that the "E" channel broadcasts 24 hours a day. There are no facts, and if there are, they are useless facts, especially in a trade magazine.
I think it's worth mentioning here that the copyright system punishes and rewards in such a way that that promotes hype over substance. I think it's unfair to "blame society", while at the same time holding this system of punishment and reward in place. In a copyright society, it is always the information that turns the most heads that gets the most money, where in a non copyright world the information that has the most value is rewarded the most.
Could you give some examples? I don't think I am making the connection you are trying to explain, maybe that's just because I cannot comprehend life without copyrights. It's not like I am 400 years old ;-)
TSMC and UMC are not a monopoly per se, but their owners do get to dictate the direction of the industry.
That makes them a "cartel," like the MPAA, RIAA, OPEC, et al. They are still as corrupt to the bone as a monopoly, but have a low chance that a cartel member will break ranks.
Strange, I would think the GOP would really want that power.
Why do they need it? It is not like the Democrats can get any legislation passed through Congress. The Republicans are all, for the most part, in bed with each other and scratch each others' backs even if a particular bill is not 100% in line with the GOP platform.
I assume from your post that you've never tried to call someone from a concert or a party (hey this is /. after all) where there's loud music. There really are times when you can't hear shit.
There are times when the ability to send a quick, silent message is important as well. For example, sometimes I am in requirements meetings and need to send a quick message to my wife (e.g. I'll grab Subway and be home for lunch in 30 minutes). Another example is the golf course. If my wife needs to tell me something urgent, she can TXT me and I will get the message without disturbing other golfers.
You are correct, speed and convenience are not the only factors. Sometimes the surrounding environment dictates which type of message we use. I used to think TXT was for 14 year old girls at the mall, but my wife and I use it every day and I find it gets the message across quickly and without interrupting my workflow. At $5 for 1,000 messages per month it is cheap, too.
I think that companies should re-evaluate their "need" for an extra 5% performance. Here's an idea -- if you need something 10 minutes faster, why not start the process 10 minutes sooner?
In any large organization, the process gets in the way. Some suit decides the product needs a new feature, or needs to ship sooner, or whatever, and this slowly trickles down to the developers who suddenly are put in crunch time where every minute counts. Schedules and deadlines may change daily. People's jobs may be at risk. Shit happens.
Nobody really likes it, but that is sometimes how we arrive at the point where we "need" an extra 5% performance, where we "need" the program to finish ten minutes sooner. Starting earlier is not always an option, usually because you don't know you even have to start *at all* until the last minute.
I just came across this post during meta-moderation, and it is one of the funniest human/computer analogies I have seen in a long time. Good job.
Doesn't let you run as non-admin
Requires the disk
These are my two biggest beefs with Neverwinter Nights. For a game that took years to make and is as big as it is, common sense (yeah, I know) would say they would follow good design and integrate with the OS the way Win2K/XP intended it to. Nope. It stores my settings, characters, and save games in the "Program Files" directory instead of the "Documents and Settings" directory. This is like storing save games in /usr instead of /home. This forces either running as admin or creative use of directory permissions. Completely retarded. Why would a Windows version of Jade Empire be any different?
Running with the CD is a hassle because I have to switch between CDs. If I leave my computer for a bathroom break, I always get nervous that my son (18 months) will grab the CD and try to "play" with it. He already know how to open the drive. He loves to stand on top of toys, no matter how big or small. I guess he is jealous that I am three times his height :-)
Just go ahead and use the long names, that alone will reduce documentation/commenting
I think variable names should answer "what," not "how." As in, what does this variable represent? Comments describe how it is used. Using concise (not necessarily short) variable names make sure you can look at a variable and say "hey, that's a window handle, that other one is a user's login, etc." Variables may have more complex usage, but usually those will be classes and have their own documentation anyway.
Exactly -- the upper limits on Windows x64 seem high now, but in a few years some of us will actually hit those limits.
I would be more interested in comparing Win64 to 64-bit Linux, BSD, Solaris, etc. How does performance compare? Hardware limits? I am off to see the wizard, the wonderful wizard of google...
In other words, Star Trek's special effects get an upgrade.
While the digital media maybe flimsy, there is no reasonable reason why the information therein should _not_ survive for more than 40 years.
Actually, there are some reasons. First, the data density in books is far less than on a CD or DVD. One word in a book may be measured in millimeters, while the same word encoded in binary on a digital medium would be measured in nanometers. Smudge a book and I guarantee you can read enough of the text and figure out the smudged parts to get 95% or better data retention. Smudge the bits on a CD, and you might lose the whole thing, which could be a hundred books worth of data.
Another reason is the quality of the inks and dyes. Older books last because they used thicker paper and tough dyes using "old" technology that lasts, sort of like old lead paint on cars. I had a 1986 Ford where the paint had all worn off, leaving it a dull grayish color. I had a tough time damaging the "paint" that was left over. I nicked the edge of the bed on my truck with a piece of plastic, leaving a small gouge in the paint. Modern books do not use high quality inks, partly because of costs, partly because of environmental concerns. The same is true for CDs. If you burn your digital archives to a true CD, they will last a while, at least 20 years if stored in a climate controlled vault (we are talking libraries/archives here, not a CD sitting on your desk at home). My brother has CDs from the 1980s, some of the first CDs ever released, and they play fine even though they were stored in adverse conditions most of their lives. However, CD-Rs operate by marking dyes on the aluminum platter, which is also exposed to air because it sits on top the plastic, not inside it. Don't trust CD-Rs for long-term storage.
I backup my data every week or two, and destroy old CDs every few months. I know I cannot depend on CD-Rs to last more than a year or two, so I don't. This also gives me an opportunity to keep my data in modern formats, as I convert them every few years when some great new format becomes widespread.
Yes, but keep in mind that back in high school she played on the boys' football team (and was a star player -- I like my women mean :-) ), and she reserves her private software interfaces for me -- everyone else gets to use the java.lang.fist interface.
I *never* partition my drives. System performance is severely limited by hard drive latency and you're insuring many loooong trips across the drive face. Even if you have ample RAM, there is always *some* swapping going on.
Partitioning makes sense if you can separate static and dynamic data. For example, my Linux box has a /usr (including /usr/local) partition, and one each for /var and /tmp. There are a few other partitions but these three illustrate my point. Obviously, /usr does not change much, so fragmentation is low. /var and /tmp change often, so fragmentation is high. Separate partitions ensure the fragmentation stays localized, and in the event of a catastrophic filesystem error (not hardware error), I only have to fix a small portion of my system.
Windows is a bit different. Microsoft "borrowed" the logical view of the Unix filesystem structure with the whole "My Computer" concept, but it is still a drive-oriented physical structure under the covers. However, by moving Program Files to its own partition (and hacking the registry to make it work correctly), Documents and Settings, and the swap file, you get essentially the same thing as in Unix: you can separate static and dynamic data.
Of course, in Linux or Windows you can just create one huge partition that contains everything (well, Linux does need a dedicated swap partition), and it will work. I think the whole idea of partitioning is just for performance tweaking more than anything.
There are ways to allow a program to degrade gracefully depending on which version of Windows it is currently running on. If you detect support for NTFS, for example, you know it is NT-based with all the user logins, home directories, and ACLs. Otherwise, use the old ugly hacks and encourage your users to upgrade to NT-based Windows or Linux.
Also, %HOMEDIR% is just one method. Ideally you would use API calls to grab the directories you need. For example, there is one that returns the "Application Data" directory in a user's home directory. All you have to do is strcat() your application's name you are set. Then alter your message loop (or slots in Qt, or whatever MFC uses) to detect user-login messages. If you detect a user switches, you can save off the user's data first then set a flag or something so when the user switches back you can reload the data.
If you want more information send me an email (my address is unobfuscated in the message header) and I can dredge my MSDN links for specific information.
The article says it's not optimal to put your swap on a 2nd partition of your master drive.
Okay, maybe I was not too explicit, but my Windows box has two physical drives on two separate P-IDE channels (until I can afford a new SATA drive). Swap is the first partition on my secondary drive.
Even if it wasn't, having my swap file in a single contiguous chunk on a partition in the middle of my hard drive is okay performance, even by what the article says.
Keep in mind that I believe in the /usr and /home idea of Unix, so I keep programs and data separate in Windows. Sure, I am not the typical user of Windows that brings his computer home from Worst Buy or CompUSA, but I am a nerd, and I do believe at least a little bit in the spirit of Unix :-)
Annoyances.org has a page about this, but I know there is a way to specify this on a per-drive basis. I don't know for sure, but TweakUI might help. I did search my registry for my "s: (swap) drive" just to be 100% sure, but came up with a bunch of results for "Neverwinter Nights: subtopic" so I just gave up after about 20 results. I know I have it set up on my system just for that one drive. Sorry.
Yes, it can be a lot of work running as non-admin. But there's a certainly feeling of secureness knowing that half the worms and trojans out there can't get very far without admin rights too.
I run all my computers behind a cable modem NAT firewall. Though simple, it stops 99% of malware at the gate. Windows firewall and paranoia stops the other 1% (I run Firefox, OpenOffice, etc. The only non-OSS I run is Windows itself). Even if this box is compromised, I turn 90 degrees to my right and use my Linux box: instant security :-)