(save in the case of a superconductor, but no one's going to suggest superconducting computers)
You know, it is funny but for the last 15 years of my life I've been personally involved with designing just such a beast and I can claim that I do suggest building it pretty soon...;-) In any case, superconductor technology is way more mature than anything that photonics can offer right now.
Actually, the quote in GP Subj: was from my former adviser Prof. Kostya Likharev said in exactly this context: with Josephsonics you can get extremely sharp pulses (~1ps) propagating along superconductor transmission line with the speed of light (in medium) AND they also can easily interact with each other, unlike photons.
Well, I can continue this discussion if you find it entertaining...;-)
Paul Bunyk
P.S. Reminds me of an old probability theory joke about conditional probabilities. OK, the odds that I look out my window and the first pedestrian I see is a man is about 1/2, odds that I see two men in a row is 1/4, etc. So, two guys made a bet, one betting that they will see 10 men in a row (1/1024, right?), the other thought that he was stupid, of course. They walked to the window and saw some kind of a military parade marching by on the street.;-)
Now, what are the odds that you find a superconductor electronics guy on/. ?
As my former advisor used to say in a very relatex context. Yes, if you build any kind of a waveguide (e.g., stripline or coplanar waveguide, or a real one, with metal walls) and send an electric pulse down that, the resulting wave will propagate all the way to the end. And the wave can be viewed as a PHOTON (except that lower frequency than optical).
If your metal is resistive, you'll dissipate some energy in that, the same as if your leght-transmitting medium is slightly opaque. But no, you do not need to feed extra power to keep the wave propagating.
The power dissipated in modern semiconductor processors (which do NOT use this mode of wave propagation, BTW) is mostly in CV^2/2 charging/discharging gate and line capacitors to sufficient voltage to open/close the next transistor.
Pascal -> modula -> modula 2 -> Ada (fighter planes and other DoD S/W) -> VHDL (hardware)
As to education, the original intent (as far as I remember) was not only to teach "Introduction to Programming" in Pascal, but "Compilation and Interpretation of Languages" as well -- thus syntax intentionally suitable for writing a simple enough Pascal compiler in Pascal itself.
Paul B.
P.S. And, of course, it is quite proper for UCSD to host the event -- I have fond memories of good old UCSD Pascal (compiled to P-Code!) running on Bulgarian clones of Apple ][!;-)
This seems to be on the presentation/rendering level, not on the actual website content level.
Sounds like mouse gestures, decent electronic (and other) CAD programs have had those for years. As in, draw an 'L' with middle button pressed in Mentor and it will zoom into the area defined by L's bounding box. Or draw Z/z to zoom in/out, etc...
OTOH, in the finest/. tradition I did not RTFA, and I can agree with the viepoint that MS can do an especially screwed-up implementation of a simple and nice idea.
What, are you some kind of a weird kid too young for/. ??? In my time the radios mentioning "7 transistors inside" were made out of germanium transistors!
They make you watch "How to Steal a Million" in high school!!!;-)
No, seriously, back in late 70s-early 80s our school (Moscow #45, anyone else from there?) somehow got that movie and it was a required part of our (middle? or high? do not remember which grade) school English class. No, not only watching once, but we had its script printed out and had to memorize and recite scenes...
Not that I had any fun doing this back then, but it was lots of fun to watch it on TV a couple of years ago and shock my GF by reciting the dialogue before characters actually say it.;-)
Hey, I've beend designing and testing low-temperature superconducting chips for the last, what, like 15 years... No, there is not too much difference in the degree of coldness, if anything, LN2 can actually freeze you if you manage to stick your finger into it and keep it there long enough, LH woudl evaporate long before that!;-)
As to the cost -- yes, it is much cheaper to use LN2 for smallish systems (as demonstrated by the FA), but it does make maybe a factor of 10 inm electricity bills to cool a chip down to 4K rather than 77K. On top of that, right now you can get a $30K cryocooler off the shelf, if any market woudl emerge those would get down to probably a couple of thousands. Not something to put in your PC, but definitely something which can be put into a nice workstation.
And, BTW, you can get a liter of liquid helium now for about ten bucks, if you care (it is heavily subsidized by the govt.)
Reading your post is like one of those "How Many Mistakes Are In This Picture" things you see in Highlights.
In these quarters we call 'em "Trolls" and do not really bother responding. If you ever think that "Noone would be THAT stupid!", it is a good sign for a troll...
Otherwise, thanks for taking time to provide a great clarification to ones here who are actually curious about those things!
Reading that old Morris & Thompson paper was great fun though!;-)
... not warm and fuzzy feelings towards an average user!;-)
If software would verify user's password immediately without actually spending the time to decrypt the data it will increase the chances of successful dictionary attack on the password. If you were trying to login into a remote system it can disable the account you are trying to use after small number of unsuccessful attempts, but here the software runs not on the drive itself, but on the computer which is completely controlled by the attacker and he can try as many times as he wishes.
Thus, it makes sense to perform full "decryption" on every attempt and return just garbage, or, to make it a bit more user-friendly, run 'file' command on the decrypted file afterwards and verify that it has meaningful signature.
Because of this, hashing is irreversable, and therefor only an idiot would use it for encryption. It's proper purpose is for checksuming.
MD5 *does* have something to do with cryptography (why else would Schneier devote the whole 14th chapter of Applied Cryptography to "One-way hash functions"), and the reason is simple: it is used to encrypt your *password*, not your data (Lexar was claiming that they use 256-bit AES encryption for the data itself).
For authentication you do not store the password in plaintext, only its MD5 hash, when user enters the password, MD5 of that is computed and compared to the stored MD5 string, if they match -- your user is authenticated. Of course XOR with a "magic number" could be used for the same purposes, but it would be much weaker. Thus, I think that the GP was not a troll and made a valid point: use MD5 to hash your passwords, and preferrable add some salt value to prevent against dictionary attack.
The other questiuon is why did Lexar had to store passwords on the drive at all, one does not need to authenticate users in their scenario (the drive itself is not a self-cointained computer to which a user needs to gain access) -- they could've just asked for the password, convert it to the key used in AES algorithm, decode the data and give the result: if password is incorrect, the decoded data is garbage.
Exactly the point I wanted to make, and waited for someone to make it...;-) Not to mention the fact that this is the way I actually deal with day-to-day hassles of not having the most up-to-date, say, gnuplot, available on our company's servers, which does not require a call to sysadm types, but also, as a friend of mine said long time ago "You never learn much of UNIX until you've spent your time on a system on which you had no root..."
Excellent post, man! And sorry to hear about your situation, but of all the people around here you have an honest chance to drive it in...
On the other hand, you were intelligent enough to work for Argonne, so you will find a couple of people agreeing with you over here (I hope!), but not in the general populance...;-(
Do you mean this Fatburger, my favorite burger place?;-) Of course some people say that In-n-Out is better, but I totally disagree! And Fatburger would NEVER offer ham & swiss sandwich, do not troll us, true believers!;-)
Paul B.
P.S. Yes, I do think that there is place in this world for many burger joints, even obscure ones; as well as for many "experimental" distros. From this article someone will learn about that "prelinking" thing and it might make his day. It's/., after all!
You did not take this thought to the end...
on
Linux Clustering
·
· Score: 1
The codes that run well on blue-gene are MORE like the codes run on clusters than those on a traditional vector super.
And if you code your application for MPI you can debug/test/optimize it on a cheap cluster. THEN when you start running into communication latency and problems too large to be solved on commodity hardware you can recompile your code on big(ger) iron, like Blue Gene/L.
Really, in my post I was talking mostly about the moral side, not about the legal side and yours is a great addition to the discussion.... as anyone who's taken a concealed handgun course will tell you.
As you can tell, I have not taken that particular course and I do not actually own a gun (moreover, however well I shot BB gun when I was 9 did not help me at all to shoot AK45 and handgun when I had a chance to do that, must be my declinign eyes...;-) ), and if I would I'd make sure to learn the laws of the state I live in.
Repeat after me. NEVER, EVER, fire a warning shot. Either you are justified in using deadly force or you are not. Period. The only time you discharge a weapon in a self-defense situation is with the intent of hitting the person/thing causing the deadly threat.
I can see your point, but take my comment in the context of the post I was replying to, to quotre below: Killing someone because of trespassing? Someone that's almost surely unarmed!? That's the last solution I would've chosen, if at all. Flame me all you want, but deep down you know it's wrong. Guess Michael Moore was into something in the movie Bowling For Columbine.
My point was that one does not have to "kill for trespassing", but to use one's advantage (i.e., having a gun) and one's brains to the best. Which might include a warning shot to make one's point without actually physically harming anyone.
If you fire a warning shot, you become the aggressor. The attacker is now justified in killing you because they are (rightfully) in fear for their life.
How is it, if a) the attacker is an agressor to be on my property without my premission (I am not talking about bar brawl here, in that situation I agree it is best NOT to make any shots whatsoever unless you seriously intend to kill!) and b) if the shot is clearly a warning and they are free to leave/run/escape, it is better for them to do just that. As to being in fear for their life -- is not it the whole purpose?
Ah, my friend, you have not studied your history...
Ah, my friend, I did study my history and my Penal Code too, and in the parts of the world I am from it was outright illegal for a private citizen to own a weapon for self-defence, any weapon (not only a gun, as far as I remember carrying a knife longer than palm width was illegal in Soviet Russia as well, and I personally spend a night in jail for wearing a heavy-metal style bracelet with metal spikes;-) ). There were exception, for hunters, of course (but still assault style weapons were not considered to be for hunting, at even shotguns were heavily restricted), and the most "funny" one was for daggers wore as part of traditional national dress by people from Souther Russia/Caucasis. What is going on down there NOW is a good example of "...only outlaws will have guns", unfortunately...;-(
Hitler had disarmed the Jews long before they were being carted off for slaughter. He was praised at the time by many world leaders for bringing about peace.
Still, if several armed men come into your house to take you away and kill, one would expect at least some would try to grab a gun from one to shoot the others. And do not tell me that Jews were all those harmless scholars and merchants -- at least now Israeli soldiers are quite tough, are not they? Maybe it took them 50 years, who knows...
As to "many world leaders" -- did not know that, have any references? I woudl not be surprized at all. Can not trust those "world leaders" at all, can we?
The U.S. founding fathers had it right, though I prefer C. Heston's version: "... from my cold,... dead,... hands!"
But the pork-barreling in social programs is nothing compared to the Pentagon/"Defense"/"Homeland Security" sector.
Look it up! It is roughly half of what the _Federal_ govt spends just on "public education", a role in which they should not even be, according to the US Constitution, leaving that the the States and local govt/charities/businesses.
Slashdot Mirror
... have feelings too, the proper way to refer to something unknown is he/she/it, to be abbrevaiated as s/h/it! ;-)
Paul B.
No, really, it did not sound messed up at all. And thanks for providing an insightful point of view which I happen to agree with... ;-)
Paul B.
But this one caught my eye:
;-) In any case, superconductor technology is way more mature than anything that photonics can offer right now.
;-)
;-)
/. ?
(save in the case of a superconductor, but no one's going to suggest superconducting computers)
You know, it is funny but for the last 15 years of my life I've been personally involved with designing just such a beast and I can claim that I do suggest building it pretty soon...
Actually, the quote in GP Subj: was from my former adviser Prof. Kostya Likharev said in exactly this context: with Josephsonics you can get extremely sharp pulses (~1ps) propagating along superconductor transmission line with the speed of light (in medium) AND they also can easily interact with each other, unlike photons.
Well, I can continue this discussion if you find it entertaining...
Paul Bunyk
P.S. Reminds me of an old probability theory joke about conditional probabilities. OK, the odds that I look out my window and the first pedestrian I see is a man is about 1/2, odds that I see two men in a row is 1/4, etc. So, two guys made a bet, one betting that they will see 10 men in a row (1/1024, right?), the other thought that he was stupid, of course. They walked to the window and saw some kind of a military parade marching by on the street.
Now, what are the odds that you find a superconductor electronics guy on
n/t
As my former advisor used to say in a very relatex context. Yes, if you build any kind of a waveguide (e.g., stripline or coplanar waveguide, or a real one, with metal walls) and send an electric pulse down that, the resulting wave will propagate all the way to the end. And the wave can be viewed as a PHOTON (except that lower frequency than optical).
If your metal is resistive, you'll dissipate some energy in that, the same as if your leght-transmitting medium is slightly opaque. But no, you do not need to feed extra power to keep the wave propagating.
The power dissipated in modern semiconductor processors (which do NOT use this mode of wave propagation, BTW) is mostly in CV^2/2 charging/discharging gate and line capacitors to sufficient voltage to open/close the next transistor.
Paul B.
Seems like my .sig was finally ack'ed on /. ... ;-)
Paul B.
... is programmed in Pascal derivative! ;-)
;-)
(and hardware is designed another one as well)
Pascal -> modula -> modula 2 -> Ada (fighter planes and other DoD S/W) -> VHDL (hardware)
As to education, the original intent (as far as I remember) was not only to teach "Introduction to Programming" in Pascal, but "Compilation and Interpretation of Languages" as well -- thus syntax intentionally suitable for writing a simple enough Pascal compiler in Pascal itself.
Paul B.
P.S. And, of course, it is quite proper for UCSD to host the event -- I have fond memories of good old UCSD Pascal (compiled to P-Code!) running on Bulgarian clones of Apple ][!
Paul B.
This seems to be on the presentation/rendering level, not on the actual website content level.
/. tradition I did not RTFA, and I can agree with the viepoint that MS can do an especially screwed-up implementation of a simple and nice idea.
Sounds like mouse gestures, decent electronic (and other) CAD programs have had those for years. As in, draw an 'L' with middle button pressed in Mentor and it will zoom into the area defined by L's bounding box. Or draw Z/z to zoom in/out, etc...
OTOH, in the finest
Paul B.
What, are you some kind of a weird kid too young for /. ??? In my time the radios mentioning "7 transistors inside" were made out of germanium transistors!
Paul B.
It was more of a physicist's joke thing, come on!
If we are going to do some serious nit-picking, think: you see the light (E/M waves/photons), you hear the sound (sound waves).
And, by the way, one can "see" individual photons, in a room dark enough, of course.
Paul B.
Scientists believe this machine, due to come online in 2007, will enable them finally to understand why all the things we can see and touch have mass.
;-)
No, the only thing that we can see (a photon of certain wavelength) does not actually have mass!!! BBC got it wrong...
Paul B.
They make you watch "How to Steal a Million" in high school!!! ;-)
;-)
No, seriously, back in late 70s-early 80s our school (Moscow #45, anyone else from there?) somehow got that movie and it was a required part of our (middle? or high? do not remember which grade) school English class. No, not only watching once, but we had its script printed out and had to memorize and recite scenes...
Not that I had any fun doing this back then, but it was lots of fun to watch it on TV a couple of years ago and shock my GF by reciting the dialogue before characters actually say it.
And yes, it is actually a great movie!
Paul B.
Hey, I've beend designing and testing low-temperature superconducting chips for the last, what, like 15 years... No, there is not too much difference in the degree of coldness, if anything, LN2 can actually freeze you if you manage to stick your finger into it and keep it there long enough, LH woudl evaporate long before that! ;-)
As to the cost -- yes, it is much cheaper to use LN2 for smallish systems (as demonstrated by the FA), but it does make maybe a factor of 10 inm electricity bills to cool a chip down to 4K rather than 77K. On top of that, right now you can get a $30K cryocooler off the shelf, if any market woudl emerge those would get down to probably a couple of thousands. Not something to put in your PC, but definitely something which can be put into a nice workstation.
And, BTW, you can get a liter of liquid helium now for about ten bucks, if you care (it is heavily subsidized by the govt.)
Paul B.
How much do you think your friends at Google gave of their IPO profits?
;-)
Hey, some free web searches for all the school kids, maybe? Or (wait!) for teens, and adults, and seniors as well? As in YOU and ME?..
And somehow they managed to do that even BEFORE their IPO! Must be sure that $$ are coming their way...
Paul B.
Reading your post is like one of those "How Many Mistakes Are In This Picture" things you see in Highlights.
;-)
In these quarters we call 'em "Trolls" and do not really bother responding. If you ever think that "Noone would be THAT stupid!", it is a good sign for a troll...
Otherwise, thanks for taking time to provide a great clarification to ones here who are actually curious about those things!
Reading that old Morris & Thompson paper was great fun though!
Regards,
Paul B.
... not warm and fuzzy feelings towards an average user! ;-)
If software would verify user's password immediately without actually spending the time to decrypt the data it will increase the chances of successful dictionary attack on the password. If you were trying to login into a remote system it can disable the account you are trying to use after small number of unsuccessful attempts, but here the software runs not on the drive itself, but on the computer which is completely controlled by the attacker and he can try as many times as he wishes.
Thus, it makes sense to perform full "decryption" on every attempt and return just garbage, or, to make it a bit more user-friendly, run 'file' command on the decrypted file afterwards and verify that it has meaningful signature.
Paul B.
Because of this, hashing is irreversable, and therefor only an idiot would use it for encryption. It's proper purpose is for checksuming.
MD5 *does* have something to do with cryptography (why else would Schneier devote the whole 14th chapter of Applied Cryptography to "One-way hash functions"), and the reason is simple: it is used to encrypt your *password*, not your data (Lexar was claiming that they use 256-bit AES encryption for the data itself).
For authentication you do not store the password in plaintext, only its MD5 hash, when user enters the password, MD5 of that is computed and compared to the stored MD5 string, if they match -- your user is authenticated. Of course XOR with a "magic number" could be used for the same purposes, but it would be much weaker. Thus, I think that the GP was not a troll and made a valid point: use MD5 to hash your passwords, and preferrable add some salt value to prevent against dictionary attack.
The other questiuon is why did Lexar had to store passwords on the drive at all, one does not need to authenticate users in their scenario (the drive itself is not a self-cointained computer to which a user needs to gain access) -- they could've just asked for the password, convert it to the key used in AES algorithm, decode the data and give the result: if password is incorrect, the decoded data is garbage.
Paul B.
Exactly the point I wanted to make, and waited for someone to make it... ;-) Not to mention the fact that this is the way I actually deal with day-to-day hassles of not having the most up-to-date, say, gnuplot, available on our company's servers, which does not require a call to sysadm types, but also, as a friend of mine said long time ago "You never learn much of UNIX until you've spent your time on a system on which you had no root..."
Paul B.
Excellent post, man! And sorry to hear about your situation, but of all the people around here you have an honest chance to drive it in...
;-(
On the other hand, you were intelligent enough to work for Argonne, so you will find a couple of people agreeing with you over here (I hope!), but not in the general populance...
Good luck, and let's hope things will change....
Paul B.
Do you mean this Fatburger, my favorite burger place? ;-) Of course some people say that In-n-Out is better, but I totally disagree! And Fatburger would NEVER offer ham & swiss sandwich, do not troll us, true believers! ;-)
/., after all!
Paul B.
P.S. Yes, I do think that there is place in this world for many burger joints, even obscure ones; as well as for many "experimental" distros. From this article someone will learn about that "prelinking" thing and it might make his day. It's
The codes that run well on blue-gene are MORE like the codes run on clusters than those on a traditional vector super.
And if you code your application for MPI you can debug/test/optimize it on a cheap cluster. THEN when you start running into communication latency and problems too large to be solved on commodity hardware you can recompile your code on big(ger) iron, like Blue Gene/L.
Paul B.
Really, in my post I was talking mostly about the moral side, not about the legal side and yours is a great addition to the discussion. ... as anyone who's taken a concealed handgun course will tell you.
;-) ), and if I would I'd make sure to learn the laws of the state I live in.
As you can tell, I have not taken that particular course and I do not actually own a gun (moreover, however well I shot BB gun when I was 9 did not help me at all to shoot AK45 and handgun when I had a chance to do that, must be my declinign eyes...
Thanks again!
Paul B.
Repeat after me. NEVER, EVER, fire a warning shot. Either you are justified in using deadly force or you are not. Period. The only time you discharge a weapon in a self-defense situation is with the intent of hitting the person/thing causing the deadly threat.
I can see your point, but take my comment in the context of the post I was replying to, to quotre below: Killing someone because of trespassing? Someone that's almost surely unarmed!? That's the last solution I would've chosen, if at all.
Flame me all you want, but deep down you know it's wrong. Guess Michael Moore was into something in the movie Bowling For Columbine.
My point was that one does not have to "kill for trespassing", but to use one's advantage (i.e., having a gun) and one's brains to the best. Which might include a warning shot to make one's point without actually physically harming anyone.
If you fire a warning shot, you become the aggressor. The attacker is now justified in killing you because they are (rightfully) in fear for their life.
How is it, if a) the attacker is an agressor to be on my property without my premission (I am not talking about bar brawl here, in that situation I agree it is best NOT to make any shots whatsoever unless you seriously intend to kill!) and b) if the shot is clearly a warning and they are free to leave/run/escape, it is better for them to do just that. As to being in fear for their life -- is not it the whole purpose?
Paul B.
Ah, my friend, you have not studied your history...
;-) ). There were exception, for hunters, of course (but still assault style weapons were not considered to be for hunting, at even shotguns were heavily restricted), and the most "funny" one was for daggers wore as part of traditional national dress by people from Souther Russia/Caucasis. What is going on down there NOW is a good example of "...only outlaws will have guns", unfortunately... ;-(
... dead, ... hands!"
Ah, my friend, I did study my history and my Penal Code too, and in the parts of the world I am from it was outright illegal for a private citizen to own a weapon for self-defence, any weapon (not only a gun, as far as I remember carrying a knife longer than palm width was illegal in Soviet Russia as well, and I personally spend a night in jail for wearing a heavy-metal style bracelet with metal spikes
Hitler had disarmed the Jews long before they were being carted off for slaughter. He was praised at the time by many world leaders for bringing about peace.
Still, if several armed men come into your house to take you away and kill, one would expect at least some would try to grab a gun from one to shoot the others. And do not tell me that Jews were all those harmless scholars and merchants -- at least now Israeli soldiers are quite tough, are not they? Maybe it took them 50 years, who knows...
As to "many world leaders" -- did not know that, have any references? I woudl not be surprized at all. Can not trust those "world leaders" at all, can we?
The U.S. founding fathers had it right, though I prefer C. Heston's version: "... from my cold,
I guess we are on the same page here, my friend!
Paul B.
But the pork-barreling in social programs is nothing compared to the Pentagon/"Defense"/"Homeland Security" sector.
Look it up! It is roughly half of what the _Federal_ govt spends just on "public education", a role in which they should not even be, according to the US Constitution, leaving that the the States and local govt/charities/businesses.
Paul B.