I haven't received a copy and I didn't receive a copy of Melissia et al either. Does that mean that nobody has me in their address book?
I haven't got one either. However, many of the people I corespond with on a regular basis use a Unix, BeOS, or MacOS [probably more techie/geeky college student types use those than the average]. And most of the Windows users I do know use either Eudora or Communicator. Though warnings are going around at my university: possibly it's being blocked on the main SMTP servers.
Grepping through my mail directory for X-Mailer shows mostly Elm, Mutt, and Mozilla (Pine doesn't set X-Mailer or it would also be well-represented); there are only a few Outlooks or Eudoras. And there's one guy I know who uses exmh.:)
None of the ciphers mentioned by the first poster have increased security under multiple encryption. Vigenere multiple times with different keys is just Vigenere with another key.
Yes, but if you use the ciphers in the order I suggested, they don't commute. Unless I'm doing something stupid (BTW, yes, I do know something about crypto, I'm just finishing up a 300 level crypto class and a 400 level crypto class this semester). Matrix multiplication, permuations, and xor should not commute with each other under most circumstances, however. Consider this:
Lets say we choose a set of keys for the ciphers and then encrypt the plaintext x. Specifically, we choose the keys for the Vignere ciphers to be the same (call then K1). Remember V_K(V_K(x)) == x (I'm using the xor version of Vignere here), and that the keys for the Hill and Permuation ciphers stay the same.
Then choose another set of Vignere keys (again equal to each other), and encrypt x again (I'm encrypting copies of x, not the first x again). Call these keys K2.
However: Each ciphertext letter of a Hill cipher depends on several letters of the plaintext. Since V_K1(x) != V_K2(x), and we don't change the keys for the Hill ciphers, H(V_K1(x)) != H(V_K2(x)). By extending this we see that H(P(H(V_K1(x)))) != H(P(H(V_K2(x)))).
This last step is not mathematically correct (in the sense that I can't prove it): it's just intuition. However, it seems highly unlikely that H(P(H(V_K2(x))) would change in such a way that V_K1(H(P(H(V_K1(x))))) is equal to V_K2(H(P(H(V_K2(x))))) [most of the time, anyway], due to the avalanch cause by changing the key in the first step (the first Vignere cipher), and the subsequent Hill and Permuation ciphers.
So [assuming you agree that my last step sounds reasonable], we see that V(H(P(H(V(x))))) is more secure than V(V(H(H(P(x))))) in at least this one special case (there are other I can think of too). I can't think of a general case proof right at the moment, and anyway I have a Calc III final in 2 hours that I really need to study for.
Yeah, but which is gonna raise a red flag at the NSA faster: A message encrypted with patterns similar to those used in the latest greatest encryption, or a letter from Bob about how his mommy didn't love him enough?
Well then you don't want crypto, you want stego. Different (though related).
criminals still have easy ways of communicating securely over the net without using government-restricted encryption techniques.
OK, say you're using this in some country like China. Do you really think the cops are not going to come knock on your door and take you away, just because you've been using a paper-and-pencil cipher and not 3DES? Yeah, right. If they can't break it, they'll come get it out of you personally. "Secure communication" and "not government restricted" are diametrcally opposite in such countries, no matter what method is used.
And if you're not in such a crappy country, you have easy access to SSH, PGP/GnuPG, OpenSSL (well except the US because of the damned RSA patent), S/MIME clients, etc, etc. Not to mention the dozens of crypto libraries floating around. So there is no reason to use such a slow and relativley insecure method.
A real test of decryption skills needs to assume that you don't know the algorithm.
Not at all true. You would be considered very skilled if you could, say, recover the plaintext of a message encrypted with 128-bit keyed Blowfish given only the ciphertext. You know the algorithm, but it's still a major challenge.
Of course it is not _required_ that one know the algorithm in order to test skillz. Especially attacking a cipher which can be implemented by hand. Though there are do-it-by-hand ciphers which would be quite hard to crack - for instance encrypting with Vigenere, Hill, Permutation, Hill, Vignere [all with different keys] would be a major challenge. Though also very slow.
if they want to be jerks about it, fsck 'em. The fact seems to be that they did not pay the the fee. Therefore the domain was fair game for whoever wanted it. I think that if they had called you up, asked nicely, and offered to pay you back the $70, the polite thing to would be to give it back to them (but they didn't).
If it's not trademarked, you are, AFAIK, well within your rights to register the domain, as long as you can show that you aren't squatting. Probably putting up some sort of webpage (just a quick thing describing what you're planning on doing with the site) would be a good defense against squatting claims.
I'll probably be accused of being a "zealot", but I really think that big-money competitions just aren't in the spirit of free software. There's nothing wrong with receiving some financial compensation for your work, but coding solely to earn money was exactly the concept that free software was supposed to eliminate.
So basically you're saying that free software is supposed to eliminate all commercial software (including free software written by a commercial interest). That is not true: the free software movement was started because (RMS felt that) all software should be free for people to use (and modify and copy and redistribute). And RMS is OK with the sale of free software; see http://www.gnu.org/philosophy/selling.html for more information about this.
Who cares if code is written for no other reason that to make money? As long as the source is available and freely modifiable, it's just as "good" (in the moral sense) as a piece of software written for love alone. It is "free".
Note that personally I think BSD/MIT code is more "free" in the sense of allowing freedoms to the end user. Just to keep people from responding to this with "GPL is not free", etc, etc.
Install Slackware. I've never used it, but I do know that it doesn't use RPM, so you can do a minimal install and be pretty much set. FreeBSD is also very cool (/usr/ports is nice as well).
If you want to keep using an RPM based distribution, some hints. First, don't install X or any of it's related stuff. Also don't install anything that you aren't going to need the second you turn on the box (like gcc or make or apache or ftpd or telnet or ssh or basically ANYTHING). All you want are the very basic packages (in RH stuff like basesystem, dev, kernel, bash, util-linux, lilo, etc). And make sure you have a very big/usr/local.:)
Personally I do clean installs every year or so just to keep this kind of problem from getting too much out of hand: I build a lot of stuff from source and also install a lot of third-party RPMs, so things get crufty fairly quickly. I keep stuff that I download in/usr/local/packages and it seems that I've installed or upgraded about 130 items just since after Christmas when I got a new disk.
The other alternative is to compile a kernel on one machine, make a boot disk, and start totally from scratch, downloading source and building it. Better be a fast machine, though...
Note that the problem is probably not so much a problem with RPM as badly made packages. I've had problems with Redhat stuff before that was caused by bad depenedencies. Though RPMs relocatable package stuff isn't as good as it could be (hmmm... there's another thing to go on the rainy day project list).
if you or one of the other people on your project have a some decent bandwidth (DSL, Cable, T1, whatever), just put a Unix box on the line and that's that. You seem to be asking for a private system anyway: I can't imagine any public service going along with your demands.
find hookers, dealers, whatever you need. open source alternative lifestyles, i say!!
I agree, though most hookers are pretty fscking ugly. In fact, I've seen some that were about as good looking as ESR (ick!). Just kidding Eric, I'm just really drunk (at 6:30 AM no less) and in the mood to troll. Hey, who needs karma, right?
Re:nVidia has incentive to remain closed-source...
on
GPL Violation - NVIDIA
·
· Score: 1
Ok, someone tell me why the drivers are the hottest thing about all these 3d cards?
Good drivers mean better performance. For instance, on Linux, 3dfx kicks nVidias ass, even though nVidia's hardware is better, because 3dfx's Linux drivers are far superior to nVidia's.
I thought it was the chip that was doing the real work, why do these companies know if I can write a better driver?
The chip is doing the work, but often drivers do quite a bit as well. For instance the Voodoo3 uses 16-bit color internally, but 32 bit interfaces on the drivers (this is actually probably wrong but it works as an example). So probably the drivers have to convert from 32 bit to 16 bit color. Stuff like that.
And, nothing personal, but if you're asking this you probably _can't write a better driver (neither could I). But people working on XFree86, Utah-GLX, and Mesa probably can (given register-level specs and/or a GPLed, unobfuscated driver).
Would that better driver be able to drive any 3d card?
No, that's what SVGA drivers are for: they run on virtually any video card, and they're quite slow. The reason we have different drivers for different cards is to take advantage of specific properties of the card.
If thats the case, why waste money on the hardware we're getting if its only the software that matters. No one has explained this to me yet..
Well, that's why I have a Voodoo3 and not a nVidia card... the hardware is important but so is the software (in this case the drivers). I was using the SVGA driver on an i740, then started using the SVGA driver on a Voodoo3. Software stayed the same, but the hardware got better. Then I got the 3dfx drivers and things got better still. Software and hardware always work together. Linux on a 10 Mhz 386 would suck, as would MS-DOS on an Athlon. They would suck for different reasons, but both would suck.
If the stuff that's left isn't just a old case and a few screws, mosty likely someone will still be able to get use out of it. MA (well Boston really) is supposed to be a geek-heavy area, right? Find a computer club or student ACM chapter and ask if they want some old hardware. We take stuff from the CS department all the time (SPARC ELCs [weird X terminals], I think we got a VAX from them, etc). A 486 or old Pentium would be quite welcome in most such places, especially if they aren't getting any money from the student activities funds and have to scavange hardware and hope for donations.
Unless you really don't have the room, start collecting old 486's/Pentiums/whatevers, and cluster the hell out those things! I know I'd like a beowulf in my closet! And all the chicks dig it too!
Careful about putting it in a closet: we've got an 8-node Pentium II Beowulf in a smallish room, and when all of them are doing things (like RC5, that's all it ever seems to be doing), the room gets _very_ warm. Nice in winter though (hey, use it to save on heating bills!).
Well, if you know one of them but not the other, it's probably a good idea to use the one you know. It's been a while since chemistry, but I don't think you use many math symbols in it (unless this is really advanced chemistry aka physics), so TEI's math abilities don't seem much of a bonus.
And if you do have a lot of math, LaTeX seems a hell of a lot nicer than the example W3C gives for formatting x**2 + 4x + 4 = 0. In MathML, [hopefully/. won't fuck this up] it's:
If you actually have to do serious math in this paper, I'd consider you nuts for doing it in anything but LaTeX. If not, well, personally I like DocBook but having never used TEI I won't recommend anything.
I've never seen GCC crash in all the time I've used it. If it's a mess then it sure is well working mess!
I actually did cause an internal consitency error on gcc 2.95 once. I should check and see if it's still in 2.95.2.. the problem is that the bug relies on a 14,000 line C++ library, so I don't think they'd be too interested in a bug report.
I suppose I'm a tad biased against GCC just because of the fairly slow code it generates [and annoying spurious warnings], but for stability it probably is about the best thing around.
The vast majority *is* crap. But the stuff that is important, libc, the Linux kernel, GCC etc. isn't.
From the quite limited time I've spent looking at kernel code, it seems pretty good (modular, clean, etc). I'm glad as I may end up writing device drivers sometime this summer.:)
I've never looked at glibc so I can't comment (I've heard libc5 was pretty bad though, which is one of the reasons for the switch to glibc). GCC, OTOH, is pretty horrible. I think the problem is that it was designed to handle K&R C, then hacked on to support ANSI C, then C++, ObjC, Fortran, and now who-knows-what-else (IIRC there are Pascal and Ada versions too, just not in the main tree yet). But for whatever reason GCC is not at all pretty. It's odd that nobody else (including the *BSD people) have done work on a free C/C++/ObjC compiler: the only other C compiler I could find on freshmeat is lcc, which is only for non-commercial use, and only does ANSI C.
I've got a very nice Logitech 3-button. I've been using it about 20 months now, and it's working just fine to this day (though I've got to occasionaly clean out the cruft that collects in the cracks between the buttons). I don't expect I'll have to replace it for at least another 6-12 months, either. Though I'm not using the mouse continiously when I'm using the computer, either (being a CLI dork). Are you using the mouse for _everything_ on your computer? Perhaps you're just putting a lot more strain on it than the average [clicking too hard?!?!?], in which case you'll just have to deal with it.
Just write up a Perl script that randomly chooses words and HTML tags. That's how slashdot works, you know.:P
Re:What about VA/MD residents?
on
Fighting UCITA
·
· Score: 1
Yup, JHU. Actually, the Maryland adoption of UCITA was a small factor in my consideration of schools. (CMU was my other primary grad-school consideration, but ultimately JHU won out).
Well, the Physics department here is pretty good, I think (it least it gets tons of money thrown at it, which is a start). It's well liked being Bloombergs pet department, unlike CS, the unwanted stepchild of JHU. (sigh)
I'm technically working there now (though I haven't been in in some weeks, this term has really sucked for me). I had to do inventory back in March, let me tell you there are some nice ass machines in that department (I once saw a flatscreen that was fscking _huge_, and there are ~4-5 quad machines). And also a few VAXen and NeXT machines. It's a funny place.:)
Re:What about VA/MD residents?
on
Fighting UCITA
·
· Score: 1
Will this law affect me writing freeware games/utilities even though I am not a resident of Maryland? I only go to school here.
Hard to say, especially as it appears the legislature hacked on UCITA a bit before passing it. However, there was a seminar in the CS department this week about licenses, and UCITA came up. Apperantly, as long as you make sure to disclaim all warranty, you're basically OK. However, if you don't put any kind of license in with the code and then distribute it, you can be sued if your stuff messes up and does something bad.
BTW, if you're still going to be at JHU next year, might want to stop by some of the ACM or JHLUG meetings, a good time is always had by all [we just had our last meeting of the year yesterday].
Re:What about VA/MD residents?
on
Fighting UCITA
·
· Score: 1
Emigrate to a state that isn't completely run by big software interests.
Trust me, Maryland has very little in the way of software at all. Yeah, a few companies here and there (same as anywhere else), but the big software states are California and Texas (with probably Washington and Oregon trailing them, after that I'm not sure (Illinois maybe?)).
Re:What about VA/MD residents?
on
Fighting UCITA
·
· Score: 1
I'll be starting grad school for physics in Baltimore, Maryland next year.
OK, it's fairly well established in the US that you are not supposed to profit from your crimes (ie selling a book detailing the events, etc). However, how far does this go? For instance, Mitnik probably learned quite a bit about computer security while cracking; can he later get a job doing computer security stuff? Or how about this: he found holes in such-and-such a product while cracking, but the company could never figure out how he did it. Could he sell his services to the company and help them fix the holes he found while cracking their stuff? I'm really curious how far these rules extend.
Is this supposed to stop him from passing on techniques to others
Maybe that's what they claim, though in actuality probably most anything he did can be found on Bugtraq or whatever. Though personally I don't follow security issues farther than making sure the machines I admin are secure.
but doesn't this conflict with his constitutional rights to free speech?
I think possibly the fact that he's on parole or probation or whatever may mean that he has to give up that right in some respects.
Slashdot Mirror
I haven't received a copy and I didn't receive a copy of Melissia et al either. Does that mean that nobody has me in their address book?
:)
I haven't got one either. However, many of the people I corespond with on a regular basis use a Unix, BeOS, or MacOS [probably more techie/geeky college student types use those than the average]. And most of the Windows users I do know use either Eudora or Communicator. Though warnings are going around at my university: possibly it's being blocked on the main SMTP servers.
Grepping through my mail directory for X-Mailer shows mostly Elm, Mutt, and Mozilla (Pine doesn't set X-Mailer or it would also be well-represented); there are only a few Outlooks or Eudoras. And there's one guy I know who uses exmh.
None of the ciphers mentioned by the first poster have increased security under multiple encryption. Vigenere multiple times with different keys is just Vigenere with another key.
Yes, but if you use the ciphers in the order I suggested, they don't commute. Unless I'm doing something stupid (BTW, yes, I do know something about crypto, I'm just finishing up a 300 level crypto class and a 400 level crypto class this semester). Matrix multiplication, permuations, and xor should not commute with each other under most circumstances, however. Consider this:
Lets say we choose a set of keys for the ciphers and then encrypt the plaintext x. Specifically, we choose the keys for the Vignere ciphers to be the same (call then K1). Remember V_K(V_K(x)) == x (I'm using the xor version of Vignere here), and that the keys for the Hill and Permuation ciphers stay the same.
Then choose another set of Vignere keys (again equal to each other), and encrypt x again (I'm encrypting copies of x, not the first x again). Call these keys K2.
However: Each ciphertext letter of a Hill cipher depends on several letters of the plaintext. Since V_K1(x) != V_K2(x), and we don't change the keys for the Hill ciphers, H(V_K1(x)) != H(V_K2(x)). By extending this we see that H(P(H(V_K1(x)))) != H(P(H(V_K2(x)))).
This last step is not mathematically correct (in the sense that I can't prove it): it's just intuition. However, it seems highly unlikely that H(P(H(V_K2(x))) would change in such a way that
V_K1(H(P(H(V_K1(x))))) is equal to V_K2(H(P(H(V_K2(x))))) [most of the time, anyway], due to the avalanch cause by changing the key in the first step (the first Vignere cipher), and the subsequent Hill and Permuation ciphers.
So [assuming you agree that my last step sounds reasonable], we see that V(H(P(H(V(x))))) is more secure than V(V(H(H(P(x))))) in at least this one special case (there are other I can think of too). I can't think of a general case proof right at the moment, and anyway I have a Calc III final in 2 hours that I really need to study for.
Yeah, but which is gonna raise a red flag at the NSA faster: A message encrypted with patterns similar to those used in the latest greatest encryption, or a letter from Bob about how his mommy didn't love him enough?
Well then you don't want crypto, you want stego. Different (though related).
criminals still have easy ways of communicating securely over the net without using government-restricted encryption techniques.
OK, say you're using this in some country like China. Do you really think the cops are not going to come knock on your door and take you away, just because you've been using a paper-and-pencil cipher and not 3DES? Yeah, right. If they can't break it, they'll come get it out of you personally. "Secure communication" and "not government restricted" are diametrcally opposite in such countries, no matter what method is used.
And if you're not in such a crappy country, you have easy access to SSH, PGP/GnuPG, OpenSSL (well except the US because of the damned RSA patent), S/MIME clients, etc, etc. Not to mention the dozens of crypto libraries floating around. So there is no reason to use such a slow and relativley insecure method.
A real test of decryption skills needs to assume that you don't know the algorithm.
Not at all true. You would be considered very skilled if you could, say, recover the plaintext of a message encrypted with 128-bit keyed Blowfish given only the ciphertext. You know the algorithm, but it's still a major challenge.
Of course it is not _required_ that one know the algorithm in order to test skillz. Especially attacking a cipher which can be implemented by hand. Though there are do-it-by-hand ciphers which would be quite hard to crack - for instance encrypting with Vigenere, Hill, Permutation, Hill, Vignere [all with different keys] would be a major challenge. Though also very slow.
if they want to be jerks about it, fsck 'em. The fact seems to be that they did not pay the the fee. Therefore the domain was fair game for whoever wanted it. I think that if they had called you up, asked nicely, and offered to pay you back the $70, the polite thing to would be to give it back to them (but they didn't).
If it's not trademarked, you are, AFAIK, well within your rights to register the domain, as long as you can show that you aren't squatting. Probably putting up some sort of webpage (just a quick thing describing what you're planning on doing with the site) would be a good defense against squatting claims.
I'll probably be accused of being a "zealot", but I really think that big-money competitions just aren't in the spirit of free software. There's nothing wrong with receiving some financial compensation for your work, but coding solely to earn money was exactly the concept that free software was supposed to eliminate.
So basically you're saying that free software is supposed to eliminate all commercial software (including free software written by a commercial interest). That is not true: the free software movement was started because (RMS felt that) all software should be free for people to use (and modify and copy and redistribute). And RMS is OK with the sale of free software; see http://www.gnu.org/philosophy/selling.html for more information about this.
Who cares if code is written for no other reason that to make money? As long as the source is available and freely modifiable, it's just as "good" (in the moral sense) as a piece of software written for love alone. It is "free".
Note that personally I think BSD/MIT code is more "free" in the sense of allowing freedoms to the end user. Just to keep people from responding to this with "GPL is not free", etc, etc.
Install Slackware. I've never used it, but I do know that it doesn't use RPM, so you can do a minimal install and be pretty much set. FreeBSD is also very cool (/usr/ports is nice as well).
/usr/local. :)
/usr/local/packages and it seems that I've installed or upgraded about 130 items just since after Christmas when I got a new disk.
If you want to keep using an RPM based distribution, some hints. First, don't install X or any of it's related stuff. Also don't install anything that you aren't going to need the second you turn on the box (like gcc or make or apache or ftpd or telnet or ssh or basically ANYTHING). All you want are the very basic packages (in RH stuff like basesystem, dev, kernel, bash, util-linux, lilo, etc). And make sure you have a very big
Personally I do clean installs every year or so just to keep this kind of problem from getting too much out of hand: I build a lot of stuff from source and also install a lot of third-party RPMs, so things get crufty fairly quickly. I keep stuff that I download in
The other alternative is to compile a kernel on one machine, make a boot disk, and start totally from scratch, downloading source and building it. Better be a fast machine, though...
Note that the problem is probably not so much a problem with RPM as badly made packages. I've had problems with Redhat stuff before that was caused by bad depenedencies. Though RPMs relocatable package stuff isn't as good as it could be (hmmm... there's another thing to go on the rainy day project list).
if you or one of the other people on your project have a some decent bandwidth (DSL, Cable, T1, whatever), just put a Unix box on the line and that's that. You seem to be asking for a private system anyway: I can't imagine any public service going along with your demands.
find hookers, dealers, whatever you need. open source alternative lifestyles, i say!!
I agree, though most hookers are pretty fscking ugly. In fact, I've seen some that were about as good looking as ESR (ick!). Just kidding Eric, I'm just really drunk (at 6:30 AM no less) and in the mood to troll. Hey, who needs karma, right?
Ok, someone tell me why the drivers are the hottest thing about all these 3d cards?
Good drivers mean better performance. For instance, on Linux, 3dfx kicks nVidias ass, even though nVidia's hardware is better, because 3dfx's Linux drivers are far superior to nVidia's.
I thought it was the chip that was doing the real work, why do these companies know if I can write a better driver?
The chip is doing the work, but often drivers do quite a bit as well. For instance the Voodoo3 uses 16-bit color internally, but 32 bit interfaces on the drivers (this is actually probably wrong but it works as an example). So probably the drivers have to convert from 32 bit to 16 bit color. Stuff like that.
And, nothing personal, but if you're asking this you probably _can't write a better driver (neither could I). But people working on XFree86, Utah-GLX, and Mesa probably can (given register-level specs and/or a GPLed, unobfuscated driver).
Would that better driver be able to drive any 3d card?
No, that's what SVGA drivers are for: they run on virtually any video card, and they're quite slow. The reason we have different drivers for different cards is to take advantage of specific properties of the card.
If thats the case, why waste money on the hardware we're getting if its only the software that matters. No one has explained this to me yet..
Well, that's why I have a Voodoo3 and not a nVidia card... the hardware is important but so is the software (in this case the drivers). I was using the SVGA driver on an i740, then started using the SVGA driver on a Voodoo3. Software stayed the same, but the hardware got better. Then I got the 3dfx drivers and things got better still. Software and hardware always work together. Linux on a 10 Mhz 386 would suck, as would MS-DOS on an Athlon. They would suck for different reasons, but both would suck.
If the stuff that's left isn't just a old case and a few screws, mosty likely someone will still be able to get use out of it. MA (well Boston really) is supposed to be a geek-heavy area, right? Find a computer club or student ACM chapter and ask if they want some old hardware. We take stuff from the CS department all the time (SPARC ELCs [weird X terminals], I think we got a VAX from them, etc). A 486 or old Pentium would be quite welcome in most such places, especially if they aren't getting any money from the student activities funds and have to scavange hardware and hope for donations.
Unless you really don't have the room, start collecting old 486's/Pentiums/whatevers, and cluster the hell out those things! I know I'd like a beowulf in my closet! And all the chicks dig it too!
Careful about putting it in a closet: we've got an 8-node Pentium II Beowulf in a smallish room, and when all of them are doing things (like RC5, that's all it ever seems to be doing), the room gets _very_ warm. Nice in winter though (hey, use it to save on heating bills!).
Well, if you know one of them but not the other, it's probably a good idea to use the one you know. It's been a while since chemistry, but I don't think you use many math symbols in it (unless this is really advanced chemistry aka physics), so TEI's math abilities don't seem much of a bonus.
/. won't fuck this up] it's:
And if you do have a lot of math, LaTeX seems a hell of a lot nicer than the example W3C gives for formatting x**2 + 4x + 4 = 0. In MathML, [hopefully
<mrow>
<mrow>
<msup> <mi>x</mi> <mn>2</mn> </msup> <mo>+</mo>
<mrow>
<mn>4</mn>
<mo>&invisibletimes;</mo>
<mi>x</mi>
</mrow>
<mo>+</mo>
<mn>4</mn>
</mrow>
<mo>=</mo>
<mn>0</mn>
</mrow>
In LaTeX, its:
$x^2 + 4x + 4 = 0$
If you actually have to do serious math in this paper, I'd consider you nuts for doing it in anything but LaTeX. If not, well, personally I like DocBook but having never used TEI I won't recommend anything.
Well, Mandrake is at 7.1 now, and I think slack is above 7, too...
Well he couldn't possibly mean Slackware, as it went from 4 to 7.
I've never seen GCC crash in all the time I've used it. If it's a mess then it sure is well working mess!
I actually did cause an internal consitency error on gcc 2.95 once. I should check and see if it's still in 2.95.2.. the problem is that the bug relies on a 14,000 line C++ library, so I don't think they'd be too interested in a bug report.
I suppose I'm a tad biased against GCC just because of the fairly slow code it generates [and annoying spurious warnings], but for stability it probably is about the best thing around.
The vast majority *is* crap. But the stuff that is important, libc, the Linux kernel, GCC etc. isn't.
:)
From the quite limited time I've spent looking at kernel code, it seems pretty good (modular, clean, etc). I'm glad as I may end up writing device drivers sometime this summer.
I've never looked at glibc so I can't comment (I've heard libc5 was pretty bad though, which is one of the reasons for the switch to glibc). GCC, OTOH, is pretty horrible. I think the problem is that it was designed to handle K&R C, then hacked on to support ANSI C, then C++, ObjC, Fortran, and now who-knows-what-else (IIRC there are Pascal and Ada versions too, just not in the main tree yet). But for whatever reason GCC is not at all pretty. It's odd that nobody else (including the *BSD people) have done work on a free C/C++/ObjC compiler: the only other C compiler I could find on freshmeat is lcc, which is only for non-commercial use, and only does ANSI C.
I've got a very nice Logitech 3-button. I've been using it about 20 months now, and it's working just fine to this day (though I've got to occasionaly clean out the cruft that collects in the cracks between the buttons). I don't expect I'll have to replace it for at least another 6-12 months, either. Though I'm not using the mouse continiously when I'm using the computer, either (being a CLI dork). Are you using the mouse for _everything_ on your computer? Perhaps you're just putting a lot more strain on it than the average [clicking too hard?!?!?], in which case you'll just have to deal with it.
Just write up a Perl script that randomly chooses words and HTML tags. That's how slashdot works, you know. :P
Yup, JHU. Actually, the Maryland adoption of UCITA was a small factor in my consideration of schools. (CMU was my other primary grad-school consideration, but ultimately JHU won out).
:)
Well, the Physics department here is pretty good, I think (it least it gets tons of money thrown at it, which is a start). It's well liked being Bloombergs pet department, unlike CS, the unwanted stepchild of JHU. (sigh)
I'm technically working there now (though I haven't been in in some weeks, this term has really sucked for me). I had to do inventory back in March, let me tell you there are some nice ass machines in that department (I once saw a flatscreen that was fscking _huge_, and there are ~4-5 quad machines). And also a few VAXen and NeXT machines. It's a funny place.
Will this law affect me writing freeware games/utilities even though I am not a resident of Maryland? I only go to school here.
Hard to say, especially as it appears the legislature hacked on UCITA a bit before passing it. However, there was a seminar in the CS department this week about licenses, and UCITA came up. Apperantly, as long as you make sure to disclaim all warranty, you're basically OK. However, if you don't put any kind of license in with the code and then distribute it, you can be sued if your stuff messes up and does something bad.
BTW, if you're still going to be at JHU next year, might want to stop by some of the ACM or JHLUG meetings, a good time is always had by all [we just had our last meeting of the year yesterday].
Emigrate to a state that isn't completely run by big software interests.
Trust me, Maryland has very little in the way of software at all. Yeah, a few companies here and there (same as anywhere else), but the big software states are California and Texas (with probably Washington and Oregon trailing them, after that I'm not sure (Illinois maybe?)).
I'll be starting grad school for physics in Baltimore, Maryland next year.
Just curious: JHU?
OK, it's fairly well established in the US that you are not supposed to profit from your crimes (ie selling a book detailing the events, etc). However, how far does this go? For instance, Mitnik probably learned quite a bit about computer security while cracking; can he later get a job doing computer security stuff? Or how about this: he found holes in such-and-such a product while cracking, but the company could never figure out how he did it. Could he sell his services to the company and help them fix the holes he found while cracking their stuff? I'm really curious how far these rules extend.
Is this supposed to stop him from passing on techniques to others
Maybe that's what they claim, though in actuality probably most anything he did can be found on Bugtraq or whatever. Though personally I don't follow security issues farther than making sure the machines I admin are secure.
but doesn't this conflict with his constitutional rights to free speech?
I think possibly the fact that he's on parole or probation or whatever may mean that he has to give up that right in some respects.