My understanding is that the Glenside Color Computer Club will hold an Eleventh Annual "Last" Chicago CoCoFest next year, probably once again in Elgin IL...so you can come be retro with the rest of us!
Just in case this isn't a troll: the most charitable interpretation of your post is that you haven't been following the x86oid world for a year or so. All reports that I have seen indicate that an Athlon will outrun a higher clock-rate P4 on all but a few benchmarks that munch large quantities of data in a highly predictable way (so that the RDRAM can keep the P4 fed and the P4's humongous pipeline is kept full and not derailed very much). (And the other possible advantage, SSE2, will be present in the AMD *Hammer CPUs.)
That said, I agree with you wholeheartedly; the x86 architecture is like the painting of Dorian Gray, and should have died long ago...but thanks to IBM's unfortunate choice in the early 1980s, the x86 has the advantages of economy of scale--enough people are buying them to make it worthwhile for several companies to flog the dead horse repeatedly. (Even they agree with us; the way they've come up with to keep it alive is to set up a Potemkin CPU, with a decent internal architecture that we, alas, can't get to.) Yes, we're geeks, and if I weren't in a situation in which I got more money for singing at Renaissance fairs than I did for stock options (true story!), I might go for an Alpha. But the hardware of the masses is inexpensive and improving steadily. (Did the Alpha's speed increase as quickly as that of x86oid CPUs?) If we geeks can take advantage of it, why shouldn't we?
Not sure where you get your value for how big a 32-bit CPU should be, but...a lot goes to cache, and a lot goes to all the stuff that interprets the "4004 + tons o' barnacles" instruction set into something decent that is what the processor really runs.
Tell me about it. My current dilemma is whether to spring now for a SiS 735-based mobo, or to wait to see how the motherboards with the nForce chipset go...since the high-end nForce will be set to suck down data from two DDR DIMMs at a time (remember when you had to buy SIMMs in pairs? similar thing here...), it should pretty well obliterate the only real advantage the P4 has, namely raw bandwidth, leaving the P4 choking on the dust of a (considerably lower clock rate) Athlon.
Everything you need for an operating system? OSs were written in assembly language for quite a while, Unix and Burroughs large systems being among the few exceptions until relatively recently. There's nothing magic about gcc; any compiler would have done for compiling Linux.
I don't see RMS insisting on the names GNU/Python, GNU/Nautilus, GNU/ls, GNU/xmms, etc. What's the difference between Linux and anything else that just happens to be compiled with gcc?
I'm not sure people will want adaptive systems replacing ads for them. How long until we see father and son researching something on the web together, and a pop-up ad appears based on the father's web usage..."Daddy, what is that woman doing?"
OK, suppose someone does that? Then Microsoft will simply play the same game it did with IBM with respect to win32s.dll, namely continually changing the interface to break compatibility to make everyone else waste time and resources catching up...or until they figure out a change that breaks some fundamental assumption in the other guy's system (e.g. the system call in win32s.dll version 1.30 that allocated memory past the 512Mbyte limit on OS/2 DOS sessions, which had no purpose whatsoever other than to make it impossible to run Windows apps using the DLL under OS/2).
Is there any laptop I can buy without paying the "MS tax"? The used alternative sounds good not only for cost reasons but because I can get one without paying the MS tax.
Bogus analogy. mozilla talkback doesn't make mozilla refuse to run certain plugins, or prevent you from running anything else. What Win XP evidently will have in it is not only a bug reporting facility, but something that will refuse to run software deemed "unsafe" or "incompatible." How long before MS uses that to cut off someone's air supply?
Come on...how much "mangling" does it take? Windows XP will have a facility in it that will let the OS refuse to run whatever software an MS-maintained database tells it to refuse to run, and you think that's not negative? How long will it be before it's switched from refusing to run software that shows evidence of crashing to requiring the MS seal of approval to be allowed to run? (And I wouldn't doubt that would have the potential for being an expensive process and one harder to get for anyone MS considers serious competition, or perhaps even requiring some kind of license agreement...)
Algorithmic information theory defines the amount of information in a string as the length of the (shortest, I would presume--you can always pad code) program that generates it. A random sequence is one that's uncompressible--the best you can do for a program to emit it is to have a copy of the sequence itself in initialized data and spit it out.
Now...if there's an algorithm to generate an arbitrary digit of pi, obviously you can use it to write a function to generate all of them (eventually, in the sense that for any fixed N, you'll only have to wait a finite amount of time for the Nth digit to come out). That seems pretty darned compressible to me, so how the heck can the digits of pi be random? Is my understanding totally off here, or do counterintuitive things happen for infinite strings?
If your "digital cable" is like what we have in Des Moines, it's lousy (save for the improved selection). Des Moines started out with a low-quality cable outfit called Heritage Cable that only allowed for about 35 channels (and some of those were time-shared, so that just when the program you wanted on the Discovery Channel was about to come on, bam! it's pre-empted by Des Moines City Council meeting reruns...), which was later bought by TCI, then AT&T (and now another outfit is buying it). Rather than bother to provide people with a reasonable selection via analog cable, they opted to sleaze out and save bandwidth and give people bad MPEG, plus all the bother people may recall from the days of separate cable converters. (Oh, you say you bought a special picture-in-picture TV? Too bad; if the digital cable box is on, it's worthless. Oh, you want to watch one digital cable channel while time-shifting another channel, digital cable or not? Sorry, Charlie...)
MS may not have succeeded totally--check out the questions about Nautilus, where some people got confused about whether they were in a web browser or a file manager. I thought that when MS did that with IE, it was a wonderful thing.
Come to think of it, what are the details of the greater flexibility for OEMs? Is that real, or will MS play games as with per-processor licensing? "Oh, you want to add Java? OK, that will be an extra $50e6 for Windows...."
Re:The better question is ...
on
Why not Ruby?
·
· Score: 2
Hey, none of this wimpy language stuff...Real Programmers write Turing Machine state transition tables!:-)
In the Church-Turing sense, we only need one programming language (or processor, for that matter), but practically speaking it would make programming hell. The question is when do we reach the point of diminishing returns in writing yet another scripting/prototyping language, and how do we know? I personally would just as soon people experiment, until such time as we can prove a language optimal for a class of problems.
Well...there are several reasons. Obviously, the easy thing to do is borrow German words for German things--if you wanted to come up with an "English" word to replace, say, Sauerbraten, what could you do other than recite the definition?
The other reason has to do with where the action is culturally. Germany was the place to be for chemists in the 19th century--vide Beilstein, the reference for organic chemists--and was a major place for philosophy as well. A bunch of those borrowings from German are philosophical terms, e.g.Zeitgeist, Weltanschauung. Somehow "spirit of the time" and "world view" just don't have the same ring to them (and the first one is longer than the German word, too--remember Zipf's Law?).
I think you're thinking of "go," not checkers. Look up the work of Arthur Samuels on checkers-playing software (which, BTW, was done so long ago that the data representation he used was given as an example in the IBM 360 assembly language text I used in college--Struble, sitting to my left as I type...and thank goodness I don't have to write for those any more!).
Well, yes, but...relativity knocked the socks out from under Newton's notions of absolute time and place, but people still use Newtonian mechanics to design car engines. Having to be consistent with what we already know means that even if the theoretical underpinnings change, the results will look like refinement, not revolution. (There's a fine Asimov essay on this, titled something like "The Rightness of Wrong.")
Somewhere around a year or so ago, Todd Rundgren came to town. Having listened to his music for a long time, I had to go see/hear the show. I got a seat at the back, which was just as well, because it was quite loud. Todd came on stage, and started to play guitar and sing...and a puzzled look crept over my face. "Surely," I thought, "he's not really singing 'I hate my frickin' ISP.' I'll find out later what the song was."
Todd was also pushing PatroNet, which I still think is a spiffy idea, and which I'm surprised isn't discussed more here on/. in view of the MP3 brouhaha, so the next day I headed for the web page. There, big as life, was "Click here to hear Todd's latest song, 'I Hate My ISP'!"
Weren't we just through this with the "brain shields"? The market will provide what people want, even if their desires arise from self-delusion. At least when you lust after the latest hard drive, we can all agree on how fast it turns and can read and write.
To answer your question: government shouldn't distribute free hypodermics because that's not a legitimate function of government. (Admittedly, neither is the vast majority of what the US government does.)
By the way, that is not what happened. The coffee did not fall; the woman placed the cup between her legs and then removed the lid--the main thing reinforcing the cup against the inward pressure of her legs!--so she could add cream and sugar. (The intent to add condiments is from memory, but the rest is documented at this page of the 'Lectric Law Library.) The car was driven by her grandson--which makes me wonder a bit about natural selection.
Slashdot Mirror
The RAM is worth checking out: the K6-3 system I set my sister up with refused to install RH. I replaced the RAM and all was well.
Surely this isn't what was meant in the famous X windows "virus alert" by the line "X Windows. You'll envy the dead."...
My understanding is that the Glenside Color Computer Club will hold an Eleventh Annual "Last" Chicago CoCoFest next year, probably once again in Elgin IL...so you can come be retro with the rest of us!
That said, I agree with you wholeheartedly; the x86 architecture is like the painting of Dorian Gray, and should have died long ago...but thanks to IBM's unfortunate choice in the early 1980s, the x86 has the advantages of economy of scale--enough people are buying them to make it worthwhile for several companies to flog the dead horse repeatedly. (Even they agree with us; the way they've come up with to keep it alive is to set up a Potemkin CPU, with a decent internal architecture that we, alas, can't get to.) Yes, we're geeks, and if I weren't in a situation in which I got more money for singing at Renaissance fairs than I did for stock options (true story!), I might go for an Alpha. But the hardware of the masses is inexpensive and improving steadily. (Did the Alpha's speed increase as quickly as that of x86oid CPUs?) If we geeks can take advantage of it, why shouldn't we?
Not sure where you get your value for how big a 32-bit CPU should be, but...a lot goes to cache, and a lot goes to all the stuff that interprets the "4004 + tons o' barnacles" instruction set into something decent that is what the processor really runs.
Tell me about it. My current dilemma is whether to spring now for a SiS 735-based mobo, or to wait to see how the motherboards with the nForce chipset go...since the high-end nForce will be set to suck down data from two DDR DIMMs at a time (remember when you had to buy SIMMs in pairs? similar thing here...), it should pretty well obliterate the only real advantage the P4 has, namely raw bandwidth, leaving the P4 choking on the dust of a (considerably lower clock rate) Athlon.
I don't see RMS insisting on the names GNU/Python, GNU/Nautilus, GNU/ls, GNU/xmms, etc. What's the difference between Linux and anything else that just happens to be compiled with gcc?
I'm not sure people will want adaptive systems replacing ads for them. How long until we see father and son researching something on the web together, and a pop-up ad appears based on the father's web usage..."Daddy, what is that woman doing?"
OK, suppose someone does that? Then Microsoft will simply play the same game it did with IBM with respect to win32s.dll, namely continually changing the interface to break compatibility to make everyone else waste time and resources catching up...or until they figure out a change that breaks some fundamental assumption in the other guy's system (e.g. the system call in win32s.dll version 1.30 that allocated memory past the 512Mbyte limit on OS/2 DOS sessions, which had no purpose whatsoever other than to make it impossible to run Windows apps using the DLL under OS/2).
Is there any laptop I can buy without paying the "MS tax"? The used alternative sounds good not only for cost reasons but because I can get one without paying the MS tax.
When that kind of attitude goes clear down to the subatomic level, that's pretty bad.
Bogus analogy. mozilla talkback doesn't make mozilla refuse to run certain plugins, or prevent you from running anything else. What Win XP evidently will have in it is not only a bug reporting facility, but something that will refuse to run software deemed "unsafe" or "incompatible." How long before MS uses that to cut off someone's air supply?
Come on...how much "mangling" does it take? Windows XP will have a facility in it that will let the OS refuse to run whatever software an MS-maintained database tells it to refuse to run, and you think that's not negative? How long will it be before it's switched from refusing to run software that shows evidence of crashing to requiring the MS seal of approval to be allowed to run? (And I wouldn't doubt that would have the potential for being an expensive process and one harder to get for anyone MS considers serious competition, or perhaps even requiring some kind of license agreement...)
Algorithmic information theory defines the amount of information in a string as the length of the (shortest, I would presume--you can always pad code) program that generates it. A random sequence is one that's uncompressible--the best you can do for a program to emit it is to have a copy of the sequence itself in initialized data and spit it out.
Now...if there's an algorithm to generate an arbitrary digit of pi, obviously you can use it to write a function to generate all of them (eventually, in the sense that for any fixed N, you'll only have to wait a finite amount of time for the Nth digit to come out). That seems pretty darned compressible to me, so how the heck can the digits of pi be random? Is my understanding totally off here, or do counterintuitive things happen for infinite strings?
If your "digital cable" is like what we have in Des Moines, it's lousy (save for the improved selection). Des Moines started out with a low-quality cable outfit called Heritage Cable that only allowed for about 35 channels (and some of those were time-shared, so that just when the program you wanted on the Discovery Channel was about to come on, bam! it's pre-empted by Des Moines City Council meeting reruns...), which was later bought by TCI, then AT&T (and now another outfit is buying it). Rather than bother to provide people with a reasonable selection via analog cable, they opted to sleaze out and save bandwidth and give people bad MPEG, plus all the bother people may recall from the days of separate cable converters. (Oh, you say you bought a special picture-in-picture TV? Too bad; if the digital cable box is on, it's worthless. Oh, you want to watch one digital cable channel while time-shifting another channel, digital cable or not? Sorry, Charlie...)
MS may not have succeeded totally--check out the questions about Nautilus, where some people got confused about whether they were in a web browser or a file manager. I thought that when MS did that with IE, it was a wonderful thing.
Come to think of it, what are the details of the greater flexibility for OEMs? Is that real, or will MS play games as with per-processor licensing? "Oh, you want to add Java? OK, that will be an extra $50e6 for Windows...."
In the Church-Turing sense, we only need one programming language (or processor, for that matter), but practically speaking it would make programming hell. The question is when do we reach the point of diminishing returns in writing yet another scripting/prototyping language, and how do we know? I personally would just as soon people experiment, until such time as we can prove a language optimal for a class of problems.
The other reason has to do with where the action is culturally. Germany was the place to be for chemists in the 19th century--vide Beilstein, the reference for organic chemists--and was a major place for philosophy as well. A bunch of those borrowings from German are philosophical terms, e.g. Zeitgeist, Weltanschauung. Somehow "spirit of the time" and "world view" just don't have the same ring to them (and the first one is longer than the German word, too--remember Zipf's Law?).
I think you're thinking of "go," not checkers. Look up the work of Arthur Samuels on checkers-playing software (which, BTW, was done so long ago that the data representation he used was given as an example in the IBM 360 assembly language text I used in college--Struble, sitting to my left as I type...and thank goodness I don't have to write for those any more!).
Well, yes, but...relativity knocked the socks out from under Newton's notions of absolute time and place, but people still use Newtonian mechanics to design car engines. Having to be consistent with what we already know means that even if the theoretical underpinnings change, the results will look like refinement, not revolution. (There's a fine Asimov essay on this, titled something like "The Rightness of Wrong.")
Todd was also pushing PatroNet, which I still think is a spiffy idea, and which I'm surprised isn't discussed more here on /. in view of the MP3 brouhaha, so the next day I headed for the web page. There, big as life, was "Click here to hear Todd's latest song, 'I Hate My ISP'!"
So, friends, the apocalypse started a while back.
Weren't we just through this with the "brain shields"? The market will provide what people want, even if their desires arise from self-delusion. At least when you lust after the latest hard drive, we can all agree on how fast it turns and can read and write.
To answer your question: government shouldn't distribute free hypodermics because that's not a legitimate function of government. (Admittedly, neither is the vast majority of what the US government does.)
By the way, that is not what happened. The coffee did not fall; the woman placed the cup between her legs and then removed the lid--the main thing reinforcing the cup against the inward pressure of her legs!--so she could add cream and sugar. (The intent to add condiments is from memory, but the rest is documented at this page of the 'Lectric Law Library.) The car was driven by her grandson--which makes me wonder a bit about natural selection.