My experience is exactly contrary to what you say. I've had Windows' PnP stuff fail miserably almost every time I try to change a machine's hardware configuration, but it works perfectly on Linux. In my opinion, those who say Linux's PnP support is inferior to Windows' are full of shit -- it's very much the other way around.
If it's a PCI modem, it should have automatically reconfigured itself to use a different port. Or perhaps a standard one, if all the standard COM1-4 ports were taken already.
--
A 2000 MHz clock rate could come in very handy for modern games. Provided the CPU actually does a significant amount of work per clock cycle (ie, AMD Athlon), you could make Quake really perty... Whether or not that's important to you is, of course, entirely another matter.
I wouldn't be so cynical. On my Linux machines, it's very rare for me to plug in a PCI or ISA PnP card and not have it autoconfigure itself. And since, unlike Windows, the PnP stuff is quite well tied into the kernel, the device drivers automatically detect all installed hardware that they support, and so installing hardware is generally a simple matter of plugging it in, powering it on, and running modprobe. And recompiling the kernel, if your kernel doesn't have the module compiled already, but I digress. My point is, you seem to underestimate PnP's reliability. PnP might not have worked too well a few years ago, but it seems to work like a clockwork now.
This is on Linux, mind you. On Windows, I could easily spend a few days trying to figure out what the hell is wrong and fix it...
Maybe you should read the comments in the file(1) magic files. I find many types that are commented out because they conflict with some other format.
The thing people seem to be forgetting is that file guesses the file type. It is not deterministic by any means. It can make mistakes. I don't like it when computers make mistakes. It's bad enough that human brains make mistakes; I don't need computers doing it too!
File is a crude, brutal, ugly hack, and the fact that it's necessary signifies a deficiency in the Unix system -- the lack of strong, deterministic file typing.
It never ceases to amaze me how the Mac is the only major system to ever implement this mode of file typing; all others depend on file magic to guess the file type based on its contents (stupid; prone to error) or file extensions (very stupid; a file's name and type should be distinctly separated from one another).
It also never ceases to amaze me how there is always an outcry against filesystem metadata, and how people cite the concept's inflexibility. This inflexibility is a design flaw on Apple's part, and when constructing a new system, there is no such constraint. Allow me to demonstrate:
$ chtype foo.html
text/plain
$ edit foo.html
[... your text editor starts... ]
$ chtype foo.html text/html
$ edit foo.html
[... your HTML editor starts... ]
Now that wasn't so hard, was it? In the above example, a chtype command is used to view/change the type of a file, and edit is used to edit the file. Of course, there would be configuration files (~/.ftyperc) for choosing which applications should handle which types, what functionality they provide (view, edit, print, etc), and perhaps some other information.
This is basically what Debian's mime-support package does, except that it uses file(1) instead of filesystem metadata. And it works very nicely.
$ chtype foo.html
text/plain
$ edit foo.html
[... your text editor starts... ]
$ chtype foo.html text/html
$ edit foo.html
[... your HTML editor starts... ]
And there you have it. File typing and filesystem metadata for the UNIX world. Elegant, no? Just because existing filesystem metadata implementations are absurdly complex doesn't mean they have to be. Note also the use of MIME types, to further simplify and standardize the file types.
Debian already does it sorta like this, only it uses file(1) magic to determine the MIME type, rather than using filesystem metadata. (Debian users: It's in the mime-support package.) Still, it provides view, edit, and other such commands, for performing various operations on files, regardless of their file type. It also can understand when (not) to run X applications to view/edit/whatever the file.
Yet another fine example of the reasons why Debian is the One True Distribution.
Re:Lessig, Litman, and Schneier
on
Taming the Web
·
· Score: 0
Your analogy with the war on drugs is flawed, for the simple reason that the war on drugs is a total failure. Despite major arrests, convictions, etc, drugs continue to flow into this country like a tsunami. Despite the best efforts of the government, drug use remains widespread. It's not quite as much of a failure as Prohibition, but it's damn close. It's basically the same idea as Prohibition, and, for the same reasons, it is destined to fail miserably, which is exactly what's happening.
Also, drug use is becoming increasingly accepted by general society. If you take a random survey now and 10 years from now, asking people if they think drug use should be outlawed, chances are, the percentage of "yes" votes will drop in the survey results 10 years from now.
My point here is that history has demonstrated, time and time again, that, to borrow a phrase from an all too well-known race of odd-looking, predominantly metallic creatures, resistance is futile. Just as Prohibition failed, and the war on drugs is failing, so Internet censorship shall fail. (By the way, I consider the DMCA to be a form of censorship, so I'll be using "censorship" in this post to refer to the DMCA and its effects, as well as the CDA, et al.)
Most Slashdot users seem to believe that the average American citizen trusts government. This is simply not the case. People are afraid of the government doing something shady at their expense. They fight the government even when there is no reason to do so, and they fight the government with even greater fervor when there is. If you don't believe me when I say this, just take a look at all the conspiracy theories that are on national television all the time. To say that people don't watch them is fallacy; if they were watched by only a tiny minority, they would not be profitable, and would be taken off the air. But they are profitable, because people do watch them, which indicates that people are paranoid about their government doing nasty things behind their backs. Whether they've been programmed this way by big media or by other means is irrelevant; the point is that people will fight government civil liberties violations, for the simple reason that civil liberties are involved. Trust is a foreign concept to the modern American -- especially trust of government. The concept of Big Brother existed long before the Internet.
People will fight. They will fight like battle-frenzied fanatics, because civil liberty is one of the things they cherish the most.
Another proof that people are anything but complacent about civil liberties is the continued existence of the ACLU. If people felt that their civil liberties were not threatened, the ACLU would have disbanded long ago, but it hasn't, for the simple reason that people still feel that the ACLU is necessary. Or, at least, the ACLU would have become a niche group, which couldn't be farther from reality.
The particularly intense media coverage of DeCSS is, by the way, a fine example of how such unsavory organizations as big media can be used to our advantage as well as that of our enemies. Media sensationalism, for example, can work for us, too. We must use advantages like these if we are to win this war.
We will prevail. It might cause us not inconsiderable discomfort in the process, but we will prevail. Every indication I can see points to the victory of civil liberties over forces which would try to inhibit them.
Yeah, and he'll probably have AIDS as a result of getting gang raped in prison. Not my idea of humane, especially since the prison officials seem to unofficially condone prison rape. It's not as bad as beatings, starvation, sleeping in a 3-foot-square box, in 120 degree weather, or going 3 days at a time without water, but the summary execution is arguably preferable to having AIDS for the rest of your life.
We are Slashbots of Borg. Your website's signal-to-noise ratio will adapt to service us. Meaningful posts are irrelevant. Resistance is futile.
Re:Homogeny isn't a bad thing.
on
Windows in 2020
·
· Score: 0
I talked to a guy recently who was writing a free (open source, I think) 3D modeller. He was complaining about getting Direct3D and MFC to work together, so I suggested that he use a cross-platform toolkit and OpenGL. That way, I said, his code would be portable. He told me of his personal distaste for Unix, and that he didn't think there was any value in porting his software to it.
Because more people will be able to use your software! Duh! What a self-centric idiot. It's people like that who arrived at the conclusion that the earth is the center of the universe; they ignored the facts pointing to this certainly not being the case, dismissing them as irrelevant or whatever excuse they came up with (or none at all). Sad.
I was shocked. I'm sure many of you were, too. But then, how many of you have written non-portable software for Linux? You probably figured Windows sucks, and there was no reason to support it. If so, you were no better than that guy.
Most of my code is written in Java, and is therefore, by definition, portable. However, I have written software to provide a Java interface to a less than particularly portable application, so my code is obviously limited in portability to that of the application it provides an interface to. My code will work, but it won't do anything useful.:)
Wonderful platforms like BeOS are suffering because people won't write portable code; there is a serious lack of good software for any OS other than Windows, Mac, and Linux (with a few Unix's managing to get easy ports of the Linux stuff).
The reason for this is ABIs. The nice thing about Linux is that there are few undocumented system calls, and those that are undocumented can be understood by looking at the kernel/library source. So Linux is a sort of meta-OS -- developers of other operating systems will find it relatively easy (compared to Windows, anyway) to provide an ABI on their operating system to run Linux software. Unless I'm mistaken, FreeBSD comes standard with a Linux ABI. You don't see any operating system coming standard with a Win32 ABI, do you? (And no, OS/2 only did Win16.)
Now, being open source does NOT automatically make your software portable! If you use POSIX system calls all over your code (and I'd hate to see your code if you do), porting the thing to Windows would probably be harder than simply re-writing the damned thing from scratch. You must consider portability from the beginning!
I take it you haven't heard of Cygwin, then. Its sole purpose (and one which it achieves admirably) is to provide a POSIX layer on top of Win32, so running portable Unix applications on Windows is a simple matter of recompiling them. Cygwin then makes Win32 API calls in proxy for POSIX library/system calls. It also provides the usual GNU toolchain (bash, gcc, et al), so if you can't./configure;make, you've probably found a bug in Cygwin. Then there's Winelib, which is intended to provide the reverse -- Win32 API calls are mapped to POSIX and Xlib calls.
I'm not saying that you should personally port your software to every known OS -- that would be impossible
A lot of modern interpreted and semi-interpreted languages are designed to be portable to just about every reasonably modern OS known to man. A short list of such languages includes:
Java (well, duh; Sun has been hyping Java's portability for a while now)
Tcl
Python
Perl
I think that's all the most popular such languages, so you're all set, pretty much. All are transparently portable (to a degree -- I dunno about Python and Perl, and all provide some interfaces to platform-specific functionality like chmod, but portably written code should work on any platform), so basically, by writing code in these languages, you do port your software to every known OS -- or, at least, those to which a VM/interpreter has been ported. (And someone will port/write a VM/interpreter for their favorite OS if they feel the need to do so -- that is, so that they can run your application on it.)
And just think, Red Hat Linux is basically the Microsoft Windows of the Linux world. It is the slowest, least secure, least stable, and least functional Linux distribution I have ever had the displeasure of using.
Note that I refer to Red Hat Linux as the Microsoft Windows of the Linux world strictly on the basis of quality of the product, and not shady business practices. Quite the contrary; Red Hat has made so many contributions to the community I couldn't begin to enumerate them.
If you had contributed to Tux Racer, you could and should sue their sorry asses. There was a big fiasco between id Software and some renegade company or another trying to sell modified Quake source! Closed! Their tactics were basically to stall indefinitely -- saying they would open it (just give us a few more weeks), and they never did. I dunno what happened there, but it looked pretty close to court to me...
The tactics being employed by SunSpire are more or less the same -- they say they'll release the code some time after the release. Small problem: BREACH OF CONTRACT.
3. You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three
years, to give any third party, for a charge no more than your
cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer
to distribute corresponding source code. (This alternative is
allowed only for noncommercial distribution and only if you
received the program in object code or executable form with such
an offer, in accord with Subsection b above.)
Sorry, SunSpire, but the GPL says NOW, not when you feel like it. And since you don't employ each and every contributor (nor can you), you can't legally relicense it. Oops! There's probably other violations that I haven't paid attention to as well.
P.S., sorry, that GPL excerpt doesn't look so great. If Slashdot let me use fucking <PRE> I wouldn't have this problem. Oh well. Not my fault...
Guess what? That's exactly what the GPL was intended to solve in the first place -- this kind of blatant back-stabbing by these Jive Software people. They're basically stealing code and selling it. Hmm, I wonder how many developers will think twice about contributing to projects like this, now that even
open source developers are committing what amounts to treason of the open source
community!
Debian comes with a program that converts between
the package formats listed in the subject line.
Obviously, some info may be lost/generated in the
translation, but it's better than nothing by a
long shot. As an example, I've installed Sun's
Java 2 SDK 1.3 on my Debian machine, by aliening
their RPM to a Debian package.
Theo is behaving the same damned way I would if I were in his shoes. I have a big problem with software companies of any kind (save maybe game companies), and I have absolutely no qualms with reaching out and crushing them at every available opportunity.
Besides, it's called OpenSSH, not simply SSH. Like GNUscape Navigator instead of Netscape Navigator. Big difference. And as mentioned in an earlier post, Microsoft's lawyers set the appropriate precedent in court already.
The bit about pre-emptively cancelling OpenSSH's domain was really low, by the way.
"Technology people never gave their stuff away," Schroeder says. "But now folks are saying, 'You mean the New England Journal of Medicine is charging people?'"
And as usual, she's so wrong it's almost funny. The one thing about this statement that offends me more than Schroeder's obvious ignorance is her obvious arrogance. She says things with perfect confidence that she is right, yet she is obviously and blatantly wrong.
Her obvious "money, money, money" attitude is also quite offensive, and I can't believe this fool has the gall to call herself a Democrat.
There are those who have a fscking clue, and then there is drougie. Sigh.
The difference between defacing a web page and DDoSing Undernet is NOT the fact that it's the 'innocent Undernet', but that DDoS is trivial, next to unstoppable, and doesn't expose anything that hasn't been known for a good decade. Defacing the US government's web sites exposes security holes in their httpd or whatever.
The encryption work is still done by the machine
you're sitting in front of, and you're passing
encrypted data through that proxy. Otherwise, it
would kind of defeat the purpose of SSL...
Ummm, I hate to tell you all this, but unless Microshaft completes its plans of world domination, I would really like to see China pull that off. Remember, the Internet (then the ARPANET) was designed to operate during a war without failing. So far, it's still essentially as attack-resistant, except that now it's mainly a civilian playground.
The Internet was designed especially not to have a central command point for just this reason: so that enemies cannot attack or subvert it.
My conclusion is that either the Washington Times is grossly unreliable, or China's idea of information warfare is totally harmless. Worry not, fellow Slashdotters. Stupidity is safety, when it's your enemy that's stupid.
Slashdot Mirror
My experience is exactly contrary to what you say. I've had Windows' PnP stuff fail miserably almost every time I try to change a machine's hardware configuration, but it works perfectly on Linux. In my opinion, those who say Linux's PnP support is inferior to Windows' are full of shit -- it's very much the other way around.
If it's a PCI modem, it should have automatically reconfigured itself to use a different port. Or perhaps a standard one, if all the standard COM1-4 ports were taken already.
--
A 2000 MHz clock rate could come in very handy for modern games. Provided the CPU actually does a significant amount of work per clock cycle (ie, AMD Athlon), you could make Quake really perty... Whether or not that's important to you is, of course, entirely another matter.
I wouldn't be so cynical. On my Linux machines, it's very rare for me to plug in a PCI or ISA PnP card and not have it autoconfigure itself. And since, unlike Windows, the PnP stuff is quite well tied into the kernel, the device drivers automatically detect all installed hardware that they support, and so installing hardware is generally a simple matter of plugging it in, powering it on, and running modprobe. And recompiling the kernel, if your kernel doesn't have the module compiled already, but I digress. My point is, you seem to underestimate PnP's reliability. PnP might not have worked too well a few years ago, but it seems to work like a clockwork now.
This is on Linux, mind you. On Windows, I could easily spend a few days trying to figure out what the hell is wrong and fix it...
Maybe you should read the comments in the file(1) magic files. I find many types that are commented out because they conflict with some other format.
The thing people seem to be forgetting is that file guesses the file type. It is not deterministic by any means. It can make mistakes. I don't like it when computers make mistakes. It's bad enough that human brains make mistakes; I don't need computers doing it too!
File is a crude, brutal, ugly hack, and the fact that it's necessary signifies a deficiency in the Unix system -- the lack of strong, deterministic file typing.
It never ceases to amaze me how the Mac is the only major system to ever implement this mode of file typing; all others depend on file magic to guess the file type based on its contents (stupid; prone to error) or file extensions (very stupid; a file's name and type should be distinctly separated from one another).
It also never ceases to amaze me how there is always an outcry against filesystem metadata, and how people cite the concept's inflexibility. This inflexibility is a design flaw on Apple's part, and when constructing a new system, there is no such constraint. Allow me to demonstrate:
Now that wasn't so hard, was it? In the above example, a chtype command is used to view/change the type of a file, and edit is used to edit the file. Of course, there would be configuration files (~/.ftyperc) for choosing which applications should handle which types, what functionality they provide (view, edit, print, etc), and perhaps some other information.
This is basically what Debian's mime-support package does, except that it uses file(1) instead of filesystem metadata. And it works very nicely.
And there you have it. File typing and filesystem metadata for the UNIX world. Elegant, no? Just because existing filesystem metadata implementations are absurdly complex doesn't mean they have to be. Note also the use of MIME types, to further simplify and standardize the file types.
Debian already does it sorta like this, only it uses file(1) magic to determine the MIME type, rather than using filesystem metadata. (Debian users: It's in the mime-support package.) Still, it provides view, edit, and other such commands, for performing various operations on files, regardless of their file type. It also can understand when (not) to run X applications to view/edit/whatever the file.
Yet another fine example of the reasons why Debian is the One True Distribution.
Your analogy with the war on drugs is flawed, for the simple reason that the war on drugs is a total failure. Despite major arrests, convictions, etc, drugs continue to flow into this country like a tsunami. Despite the best efforts of the government, drug use remains widespread. It's not quite as much of a failure as Prohibition, but it's damn close. It's basically the same idea as Prohibition, and, for the same reasons, it is destined to fail miserably, which is exactly what's happening.
Also, drug use is becoming increasingly accepted by general society. If you take a random survey now and 10 years from now, asking people if they think drug use should be outlawed, chances are, the percentage of "yes" votes will drop in the survey results 10 years from now.
My point here is that history has demonstrated, time and time again, that, to borrow a phrase from an all too well-known race of odd-looking, predominantly metallic creatures, resistance is futile. Just as Prohibition failed, and the war on drugs is failing, so Internet censorship shall fail. (By the way, I consider the DMCA to be a form of censorship, so I'll be using "censorship" in this post to refer to the DMCA and its effects, as well as the CDA, et al.)
Most Slashdot users seem to believe that the average American citizen trusts government. This is simply not the case. People are afraid of the government doing something shady at their expense. They fight the government even when there is no reason to do so, and they fight the government with even greater fervor when there is. If you don't believe me when I say this, just take a look at all the conspiracy theories that are on national television all the time. To say that people don't watch them is fallacy; if they were watched by only a tiny minority, they would not be profitable, and would be taken off the air. But they are profitable, because people do watch them, which indicates that people are paranoid about their government doing nasty things behind their backs. Whether they've been programmed this way by big media or by other means is irrelevant; the point is that people will fight government civil liberties violations, for the simple reason that civil liberties are involved. Trust is a foreign concept to the modern American -- especially trust of government. The concept of Big Brother existed long before the Internet.
People will fight. They will fight like battle-frenzied fanatics, because civil liberty is one of the things they cherish the most.
Another proof that people are anything but complacent about civil liberties is the continued existence of the ACLU. If people felt that their civil liberties were not threatened, the ACLU would have disbanded long ago, but it hasn't, for the simple reason that people still feel that the ACLU is necessary. Or, at least, the ACLU would have become a niche group, which couldn't be farther from reality.
The particularly intense media coverage of DeCSS is, by the way, a fine example of how such unsavory organizations as big media can be used to our advantage as well as that of our enemies. Media sensationalism, for example, can work for us, too. We must use advantages like these if we are to win this war.
We will prevail. It might cause us not inconsiderable discomfort in the process, but we will prevail. Every indication I can see points to the victory of civil liberties over forces which would try to inhibit them.
Yeah, and he'll probably have AIDS as a result of getting gang raped in prison. Not my idea of humane, especially since the prison officials seem to unofficially condone prison rape. It's not as bad as beatings, starvation, sleeping in a 3-foot-square box, in 120 degree weather, or going 3 days at a time without water, but the summary execution is arguably preferable to having AIDS for the rest of your life.
We are Slashbots of Borg. Your website's signal-to-noise ratio will adapt to service us. Meaningful posts are irrelevant. Resistance is futile.
- Java (well, duh; Sun has been hyping Java's portability for a while now)
- Tcl
- Python
- Perl
I think that's all the most popular such languages, so you're all set, pretty much. All are transparently portable (to a degree -- I dunno about Python and Perl, and all provide some interfaces to platform-specific functionality like chmod, but portably written code should work on any platform), so basically, by writing code in these languages, you do port your software to every known OS -- or, at least, those to which a VM/interpreter has been ported. (And someone will port/write a VM/interpreter for their favorite OS if they feel the need to do so -- that is, so that they can run your application on it.)And just think, Red Hat Linux is basically the Microsoft Windows of the Linux world. It is the slowest, least secure, least stable, and least functional Linux distribution I have ever had the displeasure of using.
Note that I refer to Red Hat Linux as the Microsoft Windows of the Linux world strictly on the basis of quality of the product, and not shady business practices. Quite the contrary; Red Hat has made so many contributions to the community I couldn't begin to enumerate them.
If you had contributed to Tux Racer, you could and should sue their sorry asses. There was a big fiasco between id Software and some renegade company or another trying to sell modified Quake source! Closed! Their tactics were basically to stall indefinitely -- saying they would open it (just give us a few more weeks), and they never did. I dunno what happened there, but it looked pretty close to court to me...
The tactics being employed by SunSpire are more or less the same -- they say they'll release the code some time after the release. Small problem: BREACH OF CONTRACT.
Sorry, SunSpire, but the GPL says NOW, not when you feel like it. And since you don't employ each and every contributor (nor can you), you can't legally relicense it. Oops! There's probably other violations that I haven't paid attention to as well.
P.S., sorry, that GPL excerpt doesn't look so great. If Slashdot let me use fucking <PRE> I wouldn't have this problem. Oh well. Not my fault...
Guess what? That's exactly what the GPL was intended to solve in the first place -- this kind of blatant back-stabbing by these Jive Software people. They're basically stealing code and selling it. Hmm, I wonder how many developers will think twice about contributing to projects like this, now that even open source developers are committing what amounts to treason of the open source community!
And, erm, yes, I'm pissed.
color="32"m al\V$bold)$normal\304$bold($normal\`date +\"%a %B %-d %l:%M%P\"\`$bold)$normal\304$bold($normal\u$bold@$ normal\h$bold)$normal\304\304\n\300\304$bold[$norm al\w$bold]$off "
normal="\033[0;${color}m"
bold="\033[1;${color}m"
off="\033[0m"
export PS1="$normal\332\304\304$bold($normal\s$bold/$nor
There you go. I just got inspired by this article and decided to make up a new one based on user #1268's 'leet prompt.
This one's nice and themeable -- just change the `color' variable to change the color scheme. I like it green like that, but...
Debian comes with a program that converts between the package formats listed in the subject line. Obviously, some info may be lost/generated in the translation, but it's better than nothing by a long shot. As an example, I've installed Sun's Java 2 SDK 1.3 on my Debian machine, by aliening their RPM to a Debian package.
Besides, it's called OpenSSH, not simply SSH. Like GNUscape Navigator instead of Netscape Navigator. Big difference. And as mentioned in an earlier post, Microsoft's lawyers set the appropriate precedent in court already.
The bit about pre-emptively cancelling OpenSSH's domain was really low, by the way.
Get 'em, Theo!
And as usual, she's so wrong it's almost funny. The one thing about this statement that offends me more than Schroeder's obvious ignorance is her obvious arrogance. She says things with perfect confidence that she is right, yet she is obviously and blatantly wrong.
Her obvious "money, money, money" attitude is also quite offensive, and I can't believe this fool has the gall to call herself a Democrat.
The difference between defacing a web page and DDoSing Undernet is NOT the fact that it's the 'innocent Undernet', but that DDoS is trivial, next to unstoppable, and doesn't expose anything that hasn't been known for a good decade. Defacing the US government's web sites exposes security holes in their httpd or whatever.
Wrong. https is done via the HTTP `CONNECT'
method, something like this:
CONNECT host.dom.ain:443 HTTP/1.1
Host: some.proxy.server
The encryption work is still done by the machine
you're sitting in front of, and you're passing
encrypted data through that proxy. Otherwise, it
would kind of defeat the purpose of SSL...
Ummm, I hate to tell you all this, but unless Microshaft completes its plans of world domination, I would really like to see China pull that off. Remember, the Internet (then the ARPANET) was designed to operate during a war without failing. So far, it's still essentially as attack-resistant, except that now it's mainly a civilian playground.
The Internet was designed especially not to have a central command point for just this reason: so that enemies cannot attack or subvert it.
My conclusion is that either the Washington Times is grossly unreliable, or China's idea of information warfare is totally harmless. Worry not, fellow Slashdotters. Stupidity is safety, when it's your enemy that's stupid.