If you had read the article, you would know it was someone else who named the law, and he was at Fairchild when he wrote the first paper. Carver, of the Carver, Meade team who wrote the most important textbook on VLSI design.
So I'll stay with my original point - this article used 10 pages to prove the mundane. Also,what most people will assume the article proved wasn't in the article at all.
While I agree with this, and just about every point in your comment, I think my assessment would be that it misses its own point, or perhaps it doesn't quite make one. The analysis is kind of interesting, if pointless. Although, it cheats by using processors instead of memory where the factors in play relate more closely to the ones in play when the original 1965 paper was written. With the regular interconnect requirements of memory, you aren't dealing with complexity issues along side technology either.
There is some value in a closer analysis of the data, and perhaps a serious point about how things can be distorted beyond recognition. Dispite the rhetoric about disproving the exponential, the data is full of them. It's just that it isn't one continuous curve, but segments. Does that remind you of anything? It looks suspiciously like organic growth to me; punctuated equilibrium, if you will.
Makes a lot of sense, when you think about it. Any exponential curve gets cut off by limits, but that doesn't mean there isn't a new growth phase in the future, perhaps at a different rate. I kind of skipped around much of the "extended" Moore's law stuff, because it looked technicaly pretty weak as well. While trying to make a point that it's hard to quantify increases in processing power, the paper doesn't bother to point out that over time the unit of processing power has gotten bigger, so it tends to deflate the actual MIPS or MFLOPS type numbers a bit (just as interconnect requirements make a lie of just counting transistors). It doesn't even touch on the advances in disks and such.
I did find a few interesting points buried in the conclusions section, but I can't help seeing them as a little backhanded. Technologists do often act like market demand will keep rising, and part of the boom/bust cycling is exactly because the evolution of current technology can quickly surpass the market's ability to absorb and apply it. On the other hand, the advances in price/performance eventually refuels demand because whole new classes of application come along. If it didn't, demand for computers would have dried up after the first dozen or so were built (I don't think I would have ever made it as low as 4).
Sorry, I don't really have references, its just that I remember all the same debates that we see today. There has always been a lot of resistance to the "viral" aspects of the licenses. This started right from the mid '80s when GPL was first used.
I'm pretty sure the term was started by opponents to the GPL philosophy, but since then some proponents embrace it as well. It is an intended effect of the license, and I think overall it is positive. I'm not particularly in favor of the strong stand against using LGPL more widely, but that's another debate.
No, the term viral dates from the earliest GPL debates, long before MS even knew it existed.
Re:The blame-the-victim ideology strikes again
on
Kevin Free
·
· Score: 2
I don't think this is blaming the victim, so much as making a point about being prudent. Further, in cases like this where the information is stolen from a third party, they have a fiduciary responsibility to protect the information. Now, that doesn't excuse the hacker either, but it's more like your doctor leaving the door unlocked so any passing person can invade your privacy. The doctor and his employees have a positive responsibility to secure that information, and the information equivalent to leaving the door unlocked doesn't cut it.
You wouldn't think it was funny or admirable if he hacked your credit cards or identity when he was on the run. I don't really recall what crimes he was being persued for in the first place, but what I have read about his flight makes him little different than a common criminal.
If you take the route of civil disobediance, you should be prepared to suffer the consequences. Running doesn't help your cause.
Re:Just too funny
on
Kevin Free
·
· Score: 1, Troll
Umm. When did Bush say "I hereby invade country X in the name of oil company Y"?
This part of the comment destroys any credibility that you may have had. You aren't seriously asserting that oil connections aren't a big part of current US foreign policy?
Not that the "bombing innocent people" gives the parent much cred either, but you know exactly what he meant, so drop the silly rhetoric.
A more important reason Bush gets no respect is that he stole the office, and ever since 9/11 he has been cynically using the issue to solidify his political base. I disgusts me that so many Americans are taken in by this. He isn't worthy of the job.
Kinds of Free
on
Kevin Free
·
· Score: 1, Offtopic
Aren't you mixing metaphors a bit here? I don't see much difference in the use of "free" in "free as a bird" vs "free Willy", accept that "free Willy" has a number of association beyond the movie (as pointed out in a nearby comment).
Of course, until we have a real AI personality, this send of free isn't appropriate to thinking about software, while the other two are. Are you incapable of making this distinction, or just being deliberately dense?
I do like the idea of adding more kinds of free to the mix, as well as poking some fun at those who take themselves too seriously, but I think it needs a little work to actually be funny.
Certainly not useless, because without some actual programming projects, you wouldn't even find out who can and cannot program. We have all seen students who struggle with the most basic of programs; not everyone has a mind organized to do programming work.
On the other hand, beyond the intro courses, your teaching people who want to do this for a living so I would expect most projects small enough to include as part of a semester course to be too trivial to even be interesting. If they aren't, you should be thinking about a different major.
In my coursework at MIT, there was very little in terms of programming projects. The one exception was the programming lab course that I have referred to in other comments. Even in the lab, you wouldn't be able to do the assignments by just copying a program since you had to write up what you did, why, and such. In the other courses, you wouldn't pass the final if you didn't understand the programming projects. Well designed courses are not easy to cheat.
Actually, I was going to skip this course because it had been added as a degree requirement during my two years working, so it wasn't required for me. My advisor convinced me to take it, but in hindsight, I had learned most of this course while working, and I really wanted to take Doc Edgerton's strobe lab. Oh well.
I think you are right on about collaboration. It is central to doing good work in industry and on open source projects. The final project in my programming lab was a three or four person project which gets a the problem a little bit, but joining an Open Source project would be way better.
Your idea about introducing errors is pretty good. Did you hear the recent reports about Leonardo De Vinci's inventions? It seems he did exactly that. He put errors into the designs that would not be apparent until you actually built the device and it didn't work right. I guess this was on/., but I heard it on NPR as well.
I don't think the commercial projects for a fee is that practical because of communications problems. On the other hand, Open Source is perfect because transparency is always a positive value in doing the work.
This is actually something I'm very interested in doing if I can find the right situation, partners, funding and such. The concept is to set up a non-profit that can help establish open-ended project lab courses using the faculty and facilities of the institution. Industry connections would be good as well, but ideally it would be to commercially apply the Open Source project work. I'm thinking as much about Open Hardware, but any real project is always multi-disciplinary anyway. Hardware, docs, software, supporting web-sites, etc.
The biggest problem with school programming projects is, as you said, they are too trivial. I worked full time for two years before going back to finish my B.S., so when I went back, I had a much better apreciation for the problems of "programming in the large". Several hundred or even a thousand lines of code just don't give you enough complexity to see what the real problems are. Also, in the real world, you rarely are starting something from scratch, which is also a good feature of working on Open Source projects (large existing code base to modify/extend).
Yes, it's sort of the same. The young guy is fast enough to run down the perps, while the senior partner keeps him from making the big mistakes.
Let's see, '76 I took Fortran as a junior in HS, and the next summer I got a job programming in Basic (after perusing a Basic book for a few days). It only paid $2.50 an hour, but I probably would have been willing to do it for nothing just for the experience (ok, maybe I wasn't smart enough to see the value in that back then).
Back then I could devour whole language books like reading a novel, and retain most of it. Now, some of the bits fall out and I have to keep the docs handy until I've used the constructs a few times. Of course, now I don't always have to read everything because I know what will be there and I can go right to the specifics I need once I know the basic structure.
The best people are learning and teaching all the time, whatever the level of experience. I've also worked with people who are threatened by anyone who know more than they do, and those people are to be avoided at all costs.
So, he is asking about how to follow an ethics/conduct code and you are telling him to lie? Don't be stupid.
Like the previous comment, you wrote it, you own it, I don't see how a reasonable policy can stop you, and if it's not reasonable, it probably isn't legal or enforcable. I wouldn't publish the problem that the code solves without permission because that isn't yours.
The prof shouldn't be using the same problems term after term anyway, but it would be the student that uses your work for his/her assignment who would be being dishonest. In any case, if half of next terms class is trying to hand in a solution off this web site, how hard is it going to be for the professor to spot that?
As a software engineer with over 20 years, I would say this is the same in the tech industry. You need a mix of experience and younger talent on any project. Unfortunately, there are a lot of engineering managers out there who don't know this, and do a lot of damage for the organizations that employ them.
Experience usually shows in saving resources by not doing unproductive things, as well as being able to see the big picture. Often this effect is hard to measure because you only see it if your team doesn't have the experience to avoid major pitfalls.
No, make the fax machine look like a shredder
on
Modding A Paper Shredder
·
· Score: 3, Insightful
That was my first thought when I read this item. Just making a faster shredder isn't that interesting, but either sending someone a copy of what is being shredded, and/or getting a picture of the person doing the shredding seems a lot more interesting.
Ok, so there are legitimate uses of shredders that would make it bad to just send out copies, but legitimate users wouldn't mind having their picture taken (much). Maybe the mods should enable themselves after hours.
Illinois Gov. George Ryan would never have become Governor if it hadn't taken so long before we learned about the shredding party at the Sec. of State offices before he was elected.
I just typed that phrase into Google and got lots of links back. I followed the first one (which said "Unix Midi Plugin has moved" with another link). I'm not interested in this software so I didn't verify this all the way, but it sure looks like it is there.
My point was that if an sizable number of people are using it, someone is going to still have the source. It may take some searching, but you are going to find it. It might not even be in a public location any more, but if someone has it, you probably can get it, and find someone else to re-host it. If not, then it must not have been the good or useful.
Makes ya wonder if intel is saying this to discourage AMD from further active research of thier own
AMD would have to be pretty dumb to base their tech strategy on this type of thing. Intel has been pushing the technology envelope for a while now just to keep up with more efficient RISC architectures, so they are probably living closer to the edge anyway. The Itanic is so complex that they need to push the technology just to get it to work. It does give a hint about why these chips eat something like 180 watts, though.
AMD is probably still chasing Intel at the process level, so they have to keep going just to catch up, or at least not fall further behind.
Ultimately, that is the final limit in the laws of physics. At that point, we really will have to turn to design, architecture and software optimization to make any more progess. On the other hand, we are way far from these limits, so the likelyhood is we will have a shift in basic implementation technology.
On has to wonder to what extent Moore's Law is a self-fulfilling prophecy. If you're in the chip business, and you don't keep striving to keep up with Moore's law, you're going to be worried about being trampled by the competition. So all the engineers and scientists are looking for breakthroughs in any areas just to keep pace with the competition. It's an arms race and it doesn't stop until you really hit the wall.
Of course, they should Open Source whatever they can. I would expect that IBM and their lawyers would have done their best to establish and maintain full control over their products, so it is unclear just how much MS would be able to say about what they do with it now. Unless you have access to the contracts which are unlikely to be publicly available.
This points to the tragedy of the source code to abandoned projects and products ending up in the bit bucket rather than being released for people to study and investigate, if not actually reuse. At least if the code to an open source project goes missing, you can be sure that nobody found it interesting or useful.
The reason for looking at it this way should be obvious. A file that is well-protected today may be only trivially protected in 10 years, thanks to advances in technology.
The real point is that it is illegal to copy a copyrighted work without proper permission (beyond fair use, that is). Just like the stupid example of the head in the freezer, the issue is that just looking at what is in plain site is not a search, but opening things up, or even just moving other objects legally constitutes a search, and requires a warrent in most cases.
Putting up a legal framework to make it illegal to circumvent even trivial copy protection is just plain wrong-headed. Will it prevent pirating for profit? No! Can circumvention techniques be used to make legal copies? Yes!
One thing I'm a little unclear on. Does the DMCA make the tools themselves illegal, or just using them to make illegal copies? I can't really see how the former could be constitutional. Probably we need some good test cases that actually get to the heart of this issue. This one could be good since it has been demonstrated that this software isn't being used widely to pirate, which implies that it is being used legally or not at all.
Don't count on the court getting this right, particularly without an appeal since the lower courts tend to assume Congress has a right to legislate, and the law is considered valid by default.
With this new realization, is there any doubt that there exists life on other planets?
Yes, there is still plenty of doubt. Nothing about this suggests an extraterestrial origin of any life that has been found. We just don't know if there was some unique accident that started it all, or if the earth was infected from an outside source.
It is an interesting data point, and it certainly is suggestive, particularly if we don't find any variety of simple life forms in any of the "extreme" environments in the solar system. Logically, the emergence of life is a pretty amazing thing, and I wouldn't believe it was even possible if we, ourselves, were not an existence proof.
On the level of pure speculation, it seems awefully strange for the origin of life to be a unique event in the universe, so either we are not alone, or there is some sort of multi-worlds thing going on and we are in one of the lucky worlds where life got started.
Of course, the other problem in trying to meet the neighbors is that they might be so out of scale with us that we wouldn't know they exist even of we overlapped in physical range.
Hmmmm, the included story at theregister.co.uk seems to be word for word the same as a comment I read on/. yesterday. It doesn't surprise me that MS would send a "SWAT" team to head off Open Source installations, and offer to give away licenses to compete, but seeing the same story in two places makes me wonder if it is being copied around and such. Shouldn't there be many different stories with about the same parameters?
Of course, it would be just like MS to give away licenses to Win2K when it will be unsupported in a couple of years. I doubt anyone will be getting any discounts when they are forced to upgrade at that point. With Their differential pricing, they'll probably try to make up the revenue lost giving away the initial licenses.
I would be very concerned if a bug in mv caused a kernel panic. Probably its something in the filesystem code.
The raw system calls for thinks like link, unlink, and mknod expect the utilities that use them to maintain certain invariants (e.g. directories must be singly linked except for . and.. which must also be correct). Bad arguments to system calls can sometimes cause a panic, and it would probably be just inefficient to always check everything.
In some sense you are correct, it is in the filesystem, but the standard filesystem semantics expect user programs to follow the basic rules.
I'll admit to a little initial confusion about the example case, but still, string compare of the args doesn't get you anything. For 'mv' the usage case is moving a directory to an existing directory, which is a reduced case of 'mv [...] ' (moving a list of files and/or dirs to a dir). If the source is a file, you replace a file of the same name, but if there is a directory in the target dir matching the source name (just the basename part), then it should be an error whether the source is a file or directory.
Now, the issue WRT this now too long thread about a bad joke, is that the test is for existence of a dir in the target dir, not string comparison, or inodes (that's another case that may or may not work correctly).
Slashdot Mirror
If you had read the article, you would know it was someone else who named the law, and he was at Fairchild when he wrote the first paper. Carver, of the Carver, Meade team who wrote the most important textbook on VLSI design.
So I'll stay with my original point - this article used 10 pages to prove the mundane. Also,what most people will assume the article proved wasn't in the article at all.
While I agree with this, and just about every point in your comment, I think my assessment would be that it misses its own point, or perhaps it doesn't quite make one. The analysis is kind of interesting, if pointless. Although, it cheats by using processors instead of memory where the factors in play relate more closely to the ones in play when the original 1965 paper was written. With the regular interconnect requirements of memory, you aren't dealing with complexity issues along side technology either.
There is some value in a closer analysis of the data, and perhaps a serious point about how things can be distorted beyond recognition. Dispite the rhetoric about disproving the exponential, the data is full of them. It's just that it isn't one continuous curve, but segments. Does that remind you of anything? It looks suspiciously like organic growth to me; punctuated equilibrium, if you will.
Makes a lot of sense, when you think about it. Any exponential curve gets cut off by limits, but that doesn't mean there isn't a new growth phase in the future, perhaps at a different rate. I kind of skipped around much of the "extended" Moore's law stuff, because it looked technicaly pretty weak as well. While trying to make a point that it's hard to quantify increases in processing power, the paper doesn't bother to point out that over time the unit of processing power has gotten bigger, so it tends to deflate the actual MIPS or MFLOPS type numbers a bit (just as interconnect requirements make a lie of just counting transistors). It doesn't even touch on the advances in disks and such.
I did find a few interesting points buried in the conclusions section, but I can't help seeing them as a little backhanded. Technologists do often act like market demand will keep rising, and part of the boom/bust cycling is exactly because the evolution of current technology can quickly surpass the market's ability to absorb and apply it. On the other hand, the advances in price/performance eventually refuels demand because whole new classes of application come along. If it didn't, demand for computers would have dried up after the first dozen or so were built (I don't think I would have ever made it as low as 4).
I'm pretty sure the term was started by opponents to the GPL philosophy, but since then some proponents embrace it as well. It is an intended effect of the license, and I think overall it is positive. I'm not particularly in favor of the strong stand against using LGPL more widely, but that's another debate.
No, the term viral dates from the earliest GPL debates, long before MS even knew it existed.
I don't think this is blaming the victim, so much as making a point about being prudent. Further, in cases like this where the information is stolen from a third party, they have a fiduciary responsibility to protect the information. Now, that doesn't excuse the hacker either, but it's more like your doctor leaving the door unlocked so any passing person can invade your privacy. The doctor and his employees have a positive responsibility to secure that information, and the information equivalent to leaving the door unlocked doesn't cut it.
If you take the route of civil disobediance, you should be prepared to suffer the consequences. Running doesn't help your cause.
This part of the comment destroys any credibility that you may have had. You aren't seriously asserting that oil connections aren't a big part of current US foreign policy?
Not that the "bombing innocent people" gives the parent much cred either, but you know exactly what he meant, so drop the silly rhetoric.
A more important reason Bush gets no respect is that he stole the office, and ever since 9/11 he has been cynically using the issue to solidify his political base. I disgusts me that so many Americans are taken in by this. He isn't worthy of the job.
Of course, until we have a real AI personality, this send of free isn't appropriate to thinking about software, while the other two are. Are you incapable of making this distinction, or just being deliberately dense?
I do like the idea of adding more kinds of free to the mix, as well as poking some fun at those who take themselves too seriously, but I think it needs a little work to actually be funny.
On the other hand, beyond the intro courses, your teaching people who want to do this for a living so I would expect most projects small enough to include as part of a semester course to be too trivial to even be interesting. If they aren't, you should be thinking about a different major.
In my coursework at MIT, there was very little in terms of programming projects. The one exception was the programming lab course that I have referred to in other comments. Even in the lab, you wouldn't be able to do the assignments by just copying a program since you had to write up what you did, why, and such. In the other courses, you wouldn't pass the final if you didn't understand the programming projects. Well designed courses are not easy to cheat.
Actually, I was going to skip this course because it had been added as a degree requirement during my two years working, so it wasn't required for me. My advisor convinced me to take it, but in hindsight, I had learned most of this course while working, and I really wanted to take Doc Edgerton's strobe lab. Oh well.
Your idea about introducing errors is pretty good. Did you hear the recent reports about Leonardo De Vinci's inventions? It seems he did exactly that. He put errors into the designs that would not be apparent until you actually built the device and it didn't work right. I guess this was on /., but I heard it on NPR as well.
This is actually something I'm very interested in doing if I can find the right situation, partners, funding and such. The concept is to set up a non-profit that can help establish open-ended project lab courses using the faculty and facilities of the institution. Industry connections would be good as well, but ideally it would be to commercially apply the Open Source project work. I'm thinking as much about Open Hardware, but any real project is always multi-disciplinary anyway. Hardware, docs, software, supporting web-sites, etc.
The biggest problem with school programming projects is, as you said, they are too trivial. I worked full time for two years before going back to finish my B.S., so when I went back, I had a much better apreciation for the problems of "programming in the large". Several hundred or even a thousand lines of code just don't give you enough complexity to see what the real problems are. Also, in the real world, you rarely are starting something from scratch, which is also a good feature of working on Open Source projects (large existing code base to modify/extend).
Let's see, '76 I took Fortran as a junior in HS, and the next summer I got a job programming in Basic (after perusing a Basic book for a few days). It only paid $2.50 an hour, but I probably would have been willing to do it for nothing just for the experience (ok, maybe I wasn't smart enough to see the value in that back then).
Back then I could devour whole language books like reading a novel, and retain most of it. Now, some of the bits fall out and I have to keep the docs handy until I've used the constructs a few times. Of course, now I don't always have to read everything because I know what will be there and I can go right to the specifics I need once I know the basic structure.
The best people are learning and teaching all the time, whatever the level of experience. I've also worked with people who are threatened by anyone who know more than they do, and those people are to be avoided at all costs.
Like the previous comment, you wrote it, you own it, I don't see how a reasonable policy can stop you, and if it's not reasonable, it probably isn't legal or enforcable. I wouldn't publish the problem that the code solves without permission because that isn't yours.
The prof shouldn't be using the same problems term after term anyway, but it would be the student that uses your work for his/her assignment who would be being dishonest. In any case, if half of next terms class is trying to hand in a solution off this web site, how hard is it going to be for the professor to spot that?
Experience usually shows in saving resources by not doing unproductive things, as well as being able to see the big picture. Often this effect is hard to measure because you only see it if your team doesn't have the experience to avoid major pitfalls.
Ok, so there are legitimate uses of shredders that would make it bad to just send out copies, but legitimate users wouldn't mind having their picture taken (much). Maybe the mods should enable themselves after hours.
Illinois Gov. George Ryan would never have become Governor if it hadn't taken so long before we learned about the shredding party at the Sec. of State offices before he was elected.
My point was that if an sizable number of people are using it, someone is going to still have the source. It may take some searching, but you are going to find it. It might not even be in a public location any more, but if someone has it, you probably can get it, and find someone else to re-host it. If not, then it must not have been the good or useful.
AMD would have to be pretty dumb to base their tech strategy on this type of thing. Intel has been pushing the technology envelope for a while now just to keep up with more efficient RISC architectures, so they are probably living closer to the edge anyway. The Itanic is so complex that they need to push the technology just to get it to work. It does give a hint about why these chips eat something like 180 watts, though.
AMD is probably still chasing Intel at the process level, so they have to keep going just to catch up, or at least not fall further behind.
On has to wonder to what extent Moore's Law is a self-fulfilling prophecy. If you're in the chip business, and you don't keep striving to keep up with Moore's law, you're going to be worried about being trampled by the competition. So all the engineers and scientists are looking for breakthroughs in any areas just to keep pace with the competition. It's an arms race and it doesn't stop until you really hit the wall.
This points to the tragedy of the source code to abandoned projects and products ending up in the bit bucket rather than being released for people to study and investigate, if not actually reuse. At least if the code to an open source project goes missing, you can be sure that nobody found it interesting or useful.
The real point is that it is illegal to copy a copyrighted work without proper permission (beyond fair use, that is). Just like the stupid example of the head in the freezer, the issue is that just looking at what is in plain site is not a search, but opening things up, or even just moving other objects legally constitutes a search, and requires a warrent in most cases.
Putting up a legal framework to make it illegal to circumvent even trivial copy protection is just plain wrong-headed. Will it prevent pirating for profit? No! Can circumvention techniques be used to make legal copies? Yes!
One thing I'm a little unclear on. Does the DMCA make the tools themselves illegal, or just using them to make illegal copies? I can't really see how the former could be constitutional. Probably we need some good test cases that actually get to the heart of this issue. This one could be good since it has been demonstrated that this software isn't being used widely to pirate, which implies that it is being used legally or not at all.
Don't count on the court getting this right, particularly without an appeal since the lower courts tend to assume Congress has a right to legislate, and the law is considered valid by default.
A friend sent me a link to this a couple years back. We had this posted along with the story about MS patenting ones and zeros. Pretty funny stuff.
Yes, there is still plenty of doubt. Nothing about this suggests an extraterestrial origin of any life that has been found. We just don't know if there was some unique accident that started it all, or if the earth was infected from an outside source.
It is an interesting data point, and it certainly is suggestive, particularly if we don't find any variety of simple life forms in any of the "extreme" environments in the solar system. Logically, the emergence of life is a pretty amazing thing, and I wouldn't believe it was even possible if we, ourselves, were not an existence proof.
On the level of pure speculation, it seems awefully strange for the origin of life to be a unique event in the universe, so either we are not alone, or there is some sort of multi-worlds thing going on and we are in one of the lucky worlds where life got started.
Of course, the other problem in trying to meet the neighbors is that they might be so out of scale with us that we wouldn't know they exist even of we overlapped in physical range.
Of course, it would be just like MS to give away licenses to Win2K when it will be unsupported in a couple of years. I doubt anyone will be getting any discounts when they are forced to upgrade at that point. With Their differential pricing, they'll probably try to make up the revenue lost giving away the initial licenses.
The raw system calls for thinks like link, unlink, and mknod expect the utilities that use them to maintain certain invariants (e.g. directories must be singly linked except for . and .. which must also be correct). Bad arguments to system calls can sometimes cause a panic, and it would probably be just inefficient to always check everything.
In some sense you are correct, it is in the filesystem, but the standard filesystem semantics expect user programs to follow the basic rules.
Now, the issue WRT this now too long thread about a bad joke, is that the test is for existence of a dir in the target dir, not string comparison, or inodes (that's another case that may or may not work correctly).