Star Raiders for the Atari ST works under some of the ST emulators that I tried. Doesn't seem to work with STonX, which is the only one that runs under Unix:(, but I've played it under PacifiST. #define X(x,y) x##y
I think there is insight in the message for the liberal moderators it is targeted towards. Most of them are so dumb they didn't think of that, and needed to have it pointed out. #define X(x,y) x##y
> I was under the impression that there is no real difference between the collector and emitter > err, note that all we're talking about is bipolar junction transistors, not FET, so assume that's what I mean when I say "transistor". Well, qualitatively, the collector and emitter are equivalent. It's the junctions you have to worry about, and one junction is more heavily doped than the other. Also, IIRC, the geometry of some transisters (all, these days?) is such that the emitter is at the surface of the silicon, the base is a layer around that, and the collector is another layer around that. Therefore, the collector is much better at collecting electrons (or holes) emitted by the emitter than vice-versa. The net result here is that the gain is a lot higher if you put the transistor in the way it was meant to go. The junctions are doped differently, too, and you can tell which is the emitter and which is the collector by using a multimeter/diode tester across the junction. On this 2N3904 I've got here, the B-E junction measure 0.631V, and the B-C junction measures 0.655V, with a Micronta 22-181B DMM's diode tester function. (It measure bias voltage at some current like 1mA.) #define X(x,y) x##y
Sorry to be a tight-ass (in your opinion, probably), but if you've got a point, could you not use racial slurs to make it? I'm not Jewish, but that offends me. I'm not offended by words like "fuck", which don't actually insult anyone other than the person you use it against.
To answer your question, people who would try this are hardware hackers, and people with more skill than you have, luser. #define X(x,y) x##y
The guy's page said it was a pull-up resistor, IIRC. That means that some chip has a digital input that makes it do one thing if it gets a low input, and do another thing if it gets a high input. Basically, you flip the RAID-mode switch by moving the resistor. It's like a jumper that wasn't designed to be moved, but you can move it with some effort. #define X(x,y) x##y
When I was first getting into electronics, I was desoldering old boards that my high school electronics prof had lying around for parts. I just had a normal soldering iron, though. One time, I pressed too hard and the ceramic part that holds the metal tip broke, dropping a hot metal cylinder on my floor. I rolled it around until it cooled off, but there are still black marks on the sheet of masonite. hehe:) I later got a real desoldering iron, very nice:) Once I learned about programming and Unix, I mostly stopped hardware hacking, though. #define X(x,y) x##y
I think you want raid5. It uses one drive to hold an XOR of the other drives. (That means that one drive becomes the bottleneck, though.) So with 4 drives, it might work better to have divide your drives into two pairs, and make one pair the mirror of the other pair. Inside each pair, the drives are raid0 striped. All this is talked about in Linux docs about the kernel's software raid capabilities, so your hardware might not be so flexible. I think the raid10 described in the article said it was 2 pairs of drives striped and mirrored like I was saying, so do that. #define X(x,y) x##y
Maybe if you aren't familiar with using the program, then paper is almost as good. However, I'd agree with Omar that I was very surprised to see that comment. I use maple, and find it _much_ better than paper for working with big ugly equations. I also use it for checking my math homework, which I certainly don't want to do _again_ on paper:) There are some equations maple has to be coaxed to simplify the way you want, so maybe the q-chem guy has equations like this that maple doesn't deal with nicely. I've never seen a quantum-chemical equation, so don't ask me:) #define X(x,y) x##y
Your point about not wanting to get locked into a closed program is interesting. (Dal has a site license for maple, so I've got the full version for x86 linux. To save typing, I'll just talk about maple, and assume what I say applies to mathematica too.)
If maple dies, and stops being developed, there won't be any more bug fixes or ports to new architectures or OSes. Fortunately, it runs on at least one Free OS, x86 linux, so it is possible to get it to run on anything. Granted, this would not be efficient. However, by the time the x86 is uncommon (the sooner the better, as far as I'm concerned!), computers will be fast enough to emulate x86 linux. Since the OS is Free, we can make an efficient emulator and catch all the syscalls instead of catching hardware IO. If x86 is still around, but Linux dies (not likely, but _possible_), then we could wrap maple in a compatibility library to translate the linux syscalls.
Anyway, my point is that even if we use a closed package like maple, we will always be able to use it in the form it's at right now. For something like maple, that's good enough for most people. Math is math is math. 1+1=2, and that doesn't change. If new theories are developed, stuff to work with them can be released by independent authors, like what's in the maple share library. I don't imagine it would need (or even need for good performance, since I'm almost sure maple is turing complete.:) new support in the maple kernel. (but then again, you never know somebody will think of, otherwise you'd think of it yourself!) #define X(x,y) x##y
Octave has a great UI, as far as I'm concerned. It runs on a tty, and uses GNU readline for input. This gives it line editting just like bash has, and tab-completion on everything, which is much nicer than the Unix matlab on the SPARCs at school.
BTW, octave uses gnuplot for all its plotting, so it doesn't include any X code at all, AFAIK. This keeps things small. (well, smaller, I guess. I don't know why there is an Octave function for all kinds of system calls and libc functions.)
One really great thing about Octave is that it is almost completely compatible with matlab, so I can hand in my linalg homework done with octave, and apply stuff our prof tells us about matlab. one function it doesn't have is rref, but I got around that by snagging rref.m from the school's commercial matlab copy:) This year they got around to teaching us that lu factorization essentially does the same thing, so I don't need rref anymore:) #define X(x,y) x##y
The/. community thanks you, English-police man. What's with pulling out the dictionary and everything? Are you trying to show that you didn't know what it meant until you looked it up? I'd agree with your sentiments, though. It would be nice if people were more careful with their words, especially when it isn't an obvious typo that they didn't bother fixing, but a gramatical error. #define X(x,y) x##y
Using/dev/random would take _forever_ to wipe anything of any reasonable size. Reads on it block until there is enough randomness in the entropy pool to give out cryptographically secure random numbers. Using/dev/urandom would drain the entropy pool almost completely (unless the kernel keeps some entropy in reserve or something), besides the fact that it is much slower. (The kernel uses some complicated code to deal with the random pool, and I'm not sure that it could keep up with normal disk speeds of 14MB/s.)
I don't think writing random data over a disk is much more secure than writing zeros, and probably not worth it. Maybe it makes it harder for people who are trying to detect the remaining weak magnetic moments which the zeroing didn't reverse, but you could probably make that a lot harder by writing a disk full of ones after zeroing the disk. If things are _that_ important, it might be better to open up the case and trash the mechanism alignment. Then nobody would stand a chance on a modern high density hard disk. (the alignment is really sensitive in those things.) #define X(x,y) x##y
It is if you run any kind of Unix on it. There is an ssh-capable terminal prog, so they aren't useless even if they're in a computer lab where you shouldn't really hack them:) #define X(x,y) x##y
The GPL applies to software, not to physical objects. If I give you the source to some software, that _doesn't_ mean I don't have it anymore. The GPL isn't about lending things, it's about sharing knowledge and information. #define X(x,y) x##y
>Wealth creation is the only human activity that >can be morally justified.
I think that life boils down to having fun. After all, would you rather be doing something that you consider fun, and live (for example) in an ideal communist society (obviously this can't happen, since ideal communism always breaks), or would you rather have to do a crappy job that makes lots of money, but you hate it. I know I'd take the fun life.
In this day and age, wealth can usually keep you happy, though.
Reading the rest of your post, I've come to the conclusion that you are either trolling for fun, or are incredible bigotted and ignorant, and quite probably very stupid. If you are trolling for _fun_, then note that you've just donated some evidence in support of my argument that humans do things to try to have more fun overall. (sometimes we do things we don't like, but we do them because they give results which we do like.)
Try taking physics at Dalhousie. Sheesh. 10 hour quantum mechanics assignments, labs which basically require you to come in outside the normal time to be able to finish...
On the plus side being a physicist landed me a summer job with JDS Uniphase:) #define X(x,y) x##y
Your point about how the protocol looks like FTP is meaningless. The important part is the data that is sent by the protocol, what sequence, and what encoding (encrypted or cleartext, etc.) is used. The other important part is what the hosts are supposed to do based on the protocol commands they receive.
The fact that something looks similar has no effect whatsoever on security. Security is in the details. (I was going to beat on micros~1 and how their stuff often looks like it's the same as the secure stuff, etc., but I'll leave that as an exercise for the reader:)
I've found linux to be more reliable than NT in my own experience, but the study is to me meaningless unless I see it for myself. I'm not surprised they found linux more reliable, but I couldn't use it as evidence for anything. That's not quite true, since they do give hard numbers on how many failures each system had. The problem is that they don't tell you what hardware they have, or what the machines were doing, or anything important like that.
It would be cool if somebody who has a copy of the report umm, posted some of it somewhere;) #define X(x,y) x##y
CNN _does_ realize this. The Fox story he linked to says that CNN claim they were not hacked, etc. Fox is trying to say that they _were_ hacked, presumably to make them look bad. Remember, that's the same fox (AFAIK) who had (they've fixed it now. Ummm, yay, I'm so glad I can look at it now:) their web page (fox.com) not accessable to people who aren't running windoze of mac. #define X(x,y) x##y
There was no equation, just a series of definitions. r is definined in terms of P, p, and something else (don't feel like looking back to the original:). The post was perfectly clear to me, anyway. #define X(x,y) x##y
Maybe the author didn't take/proc or/dev into account, or/proc was different on the kernel he was using relative to 2.2.14 (which I assume most people are using.) I haven't run the program on my system, and I don't plan to. If there are command line options to control what dirs are scanned, then maybe someone should try limitting it to that. Maybe the program reads whole files into memory before checking them, so big files take massive amounts of RAM. For some people who have IDE disk drives but haven't used hdparm to tweak them, they will almost certainly find that the system is _much_ more responsive while doing massive I/O if they set multi-count (-m) as high as possible, and use -c 1 -u 1 -d 1. On my P200MMX w/ Quantum Fireball CR, quake remains playable while updatedb is running:)
If someone is running it now, use strace -o logfile -p pid to take a peek at what it's doing. See if it reads in the whole file or what. #define X(x,y) x##y
I was just going to ask whether you could copy an encrypted DVD image right off the disk, and burn that back onto another disk. You say that it's possible, and I imagine it is. Is anybody pointing this out to the judge or company lawyers? If not, then somebody needs to! #define X(x,y) x##y
Slashdot Mirror
Star Raiders for the Atari ST works under some of the ST emulators that I tried. Doesn't seem to work with STonX, which is the only one that runs under Unix :(, but I've played it under PacifiST.
#define X(x,y) x##y
I think there is insight in the message for the liberal moderators it is targeted towards. Most of them are so dumb they didn't think of that, and needed to have it pointed out.
#define X(x,y) x##y
> I was under the impression that there is no real difference between the collector and emitter
>
err, note that all we're talking about is bipolar junction transistors, not FET, so assume that's what I mean when I say "transistor".
Well, qualitatively, the collector and emitter are equivalent. It's the junctions you have to worry about, and one junction is more heavily doped than the other. Also, IIRC, the geometry of some transisters (all, these days?) is such that the emitter is at the surface of the silicon, the base is a layer around that, and the collector is another layer around that. Therefore, the collector is much better at collecting electrons (or holes) emitted by the emitter than vice-versa.
The net result here is that the gain is a lot higher if you put the transistor in the way it was meant to go. The junctions are doped differently, too, and you can tell which is the emitter and which is the collector by using a multimeter/diode tester across the junction. On this 2N3904 I've got here, the B-E junction measure 0.631V, and the B-C junction measures 0.655V, with a Micronta 22-181B DMM's diode tester function. (It measure bias voltage at some current like 1mA.)
#define X(x,y) x##y
Sorry to be a tight-ass (in your opinion, probably), but if you've got a point, could you not use racial slurs to make it? I'm not Jewish, but that offends me. I'm not offended by words like "fuck", which don't actually insult anyone other than the person you use it against.
To answer your question, people who would try this are hardware hackers, and people with more skill than you have, luser.
#define X(x,y) x##y
The guy's page said it was a pull-up resistor, IIRC. That means that some chip has a digital input that makes it do one thing if it gets a low input, and do another thing if it gets a high input. Basically, you flip the RAID-mode switch by moving the resistor. It's like a jumper that wasn't designed to be moved, but you can move it with some effort.
#define X(x,y) x##y
When I was first getting into electronics, I was desoldering old boards that my high school electronics prof had lying around for parts. I just had a normal soldering iron, though. One time, I pressed too hard and the ceramic part that holds the metal tip broke, dropping a hot metal cylinder on my floor. I rolled it around until it cooled off, but there are still black marks on the sheet of masonite. hehe :) I later got a real desoldering iron, very nice :) Once I learned about programming and Unix, I mostly stopped hardware hacking, though.
#define X(x,y) x##y
I think you want raid5. It uses one drive to hold an XOR of the other drives. (That means that one drive becomes the bottleneck, though.)
So with 4 drives, it might work better to have divide your drives into two pairs, and make one pair the mirror of the other pair. Inside each pair, the drives are raid0 striped. All this is talked about in Linux docs about the kernel's software raid capabilities, so your hardware might not be so flexible. I think the raid10 described in the article said it was 2 pairs of drives striped and mirrored like I was saying, so do that.
#define X(x,y) x##y
Maybe if you aren't familiar with using the program, then paper is almost as good. However, I'd agree with Omar that I was very surprised to see that comment. I use maple, and find it _much_ better than paper for working with big ugly equations. I also use it for checking my math homework, which I certainly don't want to do _again_ on paper :) There are some equations maple has to be coaxed to simplify the way you want, so maybe the q-chem guy has equations like this that maple doesn't deal with nicely. I've never seen a quantum-chemical equation, so don't ask me :)
#define X(x,y) x##y
Your point about not wanting to get locked into a closed program is interesting. (Dal has a site license for maple, so I've got the full version for x86 linux. To save typing, I'll just talk about maple, and assume what I say applies to mathematica too.)
:) new support in the maple kernel. (but then again, you never know somebody will think of, otherwise you'd think of it yourself!)
If maple dies, and stops being developed, there won't be any more bug fixes or ports to new architectures or OSes. Fortunately, it runs on at least one Free OS, x86 linux, so it is possible to get it to run on anything. Granted, this would not be efficient. However, by the time the x86 is uncommon (the sooner the better, as far as I'm concerned!), computers will be fast enough to emulate x86 linux. Since the OS is Free, we can make an efficient emulator and catch all the syscalls instead of catching hardware IO. If x86 is still around, but Linux dies (not likely, but _possible_), then we could wrap maple in a compatibility library to translate the linux syscalls.
Anyway, my point is that even if we use a closed package like maple, we will always be able to use it in the form it's at right now. For something like maple, that's good enough for most people. Math is math is math. 1+1=2, and that doesn't change. If new theories are developed, stuff to work with them can be released by independent authors, like what's in the maple share library. I don't imagine it would need (or even need for good performance, since I'm almost sure maple is turing complete.
#define X(x,y) x##y
don't follow his link, he's probably making money from the refresh to a CGI script, the bastard.
#define X(x,y) x##y
Octave has a great UI, as far as I'm concerned. It runs on a tty, and uses GNU readline for input. This gives it line editting just like bash has, and tab-completion on everything, which is much nicer than the Unix matlab on the SPARCs at school.
:) This year they got around to teaching us that lu factorization essentially does the same thing, so I don't need rref anymore :)
BTW, octave uses gnuplot for all its plotting, so it doesn't include any X code at all, AFAIK. This keeps things small. (well, smaller, I guess. I don't know why there is an Octave function for all kinds of system calls and libc functions.)
One really great thing about Octave is that it is almost completely compatible with matlab, so I can hand in my linalg homework done with octave, and apply stuff our prof tells us about matlab. one function it doesn't have is rref, but I got around that by snagging rref.m from the school's commercial matlab copy
#define X(x,y) x##y
The /. community thanks you, English-police man. What's with pulling out the dictionary and everything? Are you trying to show that you didn't know what it meant until you looked it up?
I'd agree with your sentiments, though. It would be nice if people were more careful with their words, especially when it isn't an obvious typo that they didn't bother fixing, but a gramatical error.
#define X(x,y) x##y
Using /dev/random would take _forever_ to wipe anything of any reasonable size. Reads on it block until there is enough randomness in the entropy pool to give out cryptographically secure random numbers. Using /dev/urandom would drain the entropy pool almost completely (unless the kernel keeps some entropy in reserve or something), besides the fact that it is much slower. (The kernel uses some complicated code to deal with the random pool, and I'm not sure that it could keep up with normal disk speeds of 14MB/s.)
I don't think writing random data over a disk is much more secure than writing zeros, and probably not worth it. Maybe it makes it harder for people who are trying to detect the remaining weak magnetic moments which the zeroing didn't reverse, but you could probably make that a lot harder by writing a disk full of ones after zeroing the disk. If things are _that_ important, it might be better to open up the case and trash the mechanism alignment. Then nobody would stand a chance on a modern high density hard disk. (the alignment is really sensitive in those things.)
#define X(x,y) x##y
It is if you run any kind of Unix on it. There is an ssh-capable terminal prog, so they aren't useless even if they're in a computer lab where you shouldn't really hack them :)
#define X(x,y) x##y
The GPL applies to software, not to physical objects. If I give you the source to some software, that _doesn't_ mean I don't have it anymore. The GPL isn't about lending things, it's about sharing knowledge and information.
#define X(x,y) x##y
so stop it already, lowlife scum.
#define X(x,y) x##y
That pretty much sums up the lower half of the /. denizens... :)
#define X(x,y) x##y
>Wealth creation is the only human activity that >can be morally justified.
I think that life boils down to having fun. After all, would you rather be doing something that you consider fun, and live (for example) in an ideal communist society (obviously this can't happen, since ideal communism always breaks), or would you rather have to do a crappy job that makes lots of money, but you hate it. I know I'd take the fun life.
In this day and age, wealth can usually keep you happy, though.
Reading the rest of your post, I've come to the conclusion that you are either trolling for fun, or are incredible bigotted and ignorant, and quite probably very stupid. If you are trolling for _fun_, then note that you've just donated some evidence in support of my argument that humans do things to try to have more fun overall. (sometimes we do things we don't like, but we do them because they give results which we do like.)
#define X(x,y) x##y
Try taking physics at Dalhousie. Sheesh. 10 hour quantum mechanics assignments, labs which basically require you to come in outside the normal time to be able to finish...
:)
On the plus side being a physicist landed me a summer job with JDS Uniphase
#define X(x,y) x##y
Your point about how the protocol looks like FTP is meaningless. The important part is the data that is sent by the protocol, what sequence, and what encoding (encrypted or cleartext, etc.) is used. The other important part is what the hosts are supposed to do based on the protocol commands they receive.
:)
The fact that something looks similar has no effect whatsoever on security. Security is in the details. (I was going to beat on micros~1 and how their stuff often looks like it's the same as the secure stuff, etc., but I'll leave that as an exercise for the reader
#define X(x,y) x##y
I've found linux to be more reliable than NT in my own experience, but the study is to me meaningless unless I see it for myself. I'm not surprised they found linux more reliable, but I couldn't use it as evidence for anything. That's not quite true, since they do give hard numbers on how many failures each system had. The problem is that they don't tell you what hardware they have, or what the machines were doing, or anything important like that.
;)
It would be cool if somebody who has a copy of the report umm, posted some of it somewhere
#define X(x,y) x##y
CNN _does_ realize this. The Fox story he linked to says that CNN claim they were not hacked, etc. Fox is trying to say that they _were_ hacked, presumably to make them look bad. Remember, that's the same fox (AFAIK) who had (they've fixed it now. Ummm, yay, I'm so glad I can look at it now:) their web page (fox.com) not accessable to people who aren't running windoze of mac.
#define X(x,y) x##y
There was no equation, just a series of definitions. r is definined in terms of P, p, and something else (don't feel like looking back to the original :). The post was perfectly clear to me, anyway.
#define X(x,y) x##y
Maybe the author didn't take /proc or /dev into account, or /proc was different on the kernel he was using relative to 2.2.14 (which I assume most people are using.) I haven't run the program on my system, and I don't plan to. :)
If there are command line options to control what dirs are scanned, then maybe someone should try limitting it to that. Maybe the program reads whole files into memory before checking them, so big files take massive amounts of RAM.
For some people who have IDE disk drives but haven't used hdparm to tweak them, they will almost certainly find that the system is _much_ more responsive while doing massive I/O if they set multi-count (-m) as high as possible, and use -c 1 -u 1 -d 1. On my P200MMX w/ Quantum Fireball CR, quake remains playable while updatedb is running
If someone is running it now, use strace -o logfile -p pid to take a peek at what it's doing. See if it reads in the whole file or what.
#define X(x,y) x##y
I was just going to ask whether you could copy an encrypted DVD image right off the disk, and burn that back onto another disk. You say that it's possible, and I imagine it is. Is anybody pointing this out to the judge or company lawyers? If not, then somebody needs to!
#define X(x,y) x##y