Embedded systems engineers are running into trouble for certain applications. On the one hand, StrongARM, MIPS, and even low-end microcontrollers like the 8051 are reliable and cool. But in the race for speed everyone has thrown power consumption to the wind. What if you need something more powerful than a StrongARM, but can't make use of Ultra SPARC and Intel processors because they simply run much too hot for embedded devices? So the embedded engineers are starting to resort to custom-designed processors tuned for specific purposes. But it would be much better if someone put effort into higher performance CPUs that didn't munch up 50 watts of power.
Power is cheap, in the end, and people want top-end speed and performance.
Rephrase that as "people like to think they need top end speed and performance" and I'll agree with you. It's the same reason people buy 4-wheel drive SUVs when they live in the midwest.
I can sympathize with the sentiment of the original poster, though not his particular situation. What you do in school is often much more fun that what you'll do on the job later. In school you can be introduced to some new idea, then get all enamored with it and have fun just learning for the sake of learning. On the job you mostly deal with a small set of known technologies and there's not much variety.
That aside, I've gotten disenchanted with computers after getting a B.S. and working professionally as a programmer for over a decade. The problem is that programming has been replaced by drowning in a sea of APIs and help files and systems that are loosely defined yet still require 1,000 pages of documentation. A certain GUI call doesn't do what you expect it to do, so you look through knowledge bases and do google searches and sift through example code. Then you find out that no one really knows the answer, so you twiddle around until it works, but then you're never sure why the fix works at all.
For example, try to understand the format of a Windows executable to the point where you could create your own (let's say you're writing a small compiler). Searching for an hour on the web turns up lots of specs, but they're all thin and full of holes and they don't agree with each other. Just *exactly* what does field X do? When do you need to set it? Why do some executables set it to $FF when the legal values are in the range of 0..3, and yet those executables still work? These kinds of issues are pervasive in any modern system, be it Windows, MacOS, or Linux+X+KDE. And they erode your soul.
Am I right in assuming that this is due to the equivalent of pre-compiled headers and overall better modularization than C(++) allows? Paraphrased: is it the compiler or the language that gives opportunity for these short compile times?
Sure, there are some things about Object Pascal that make it simpler to compile than C: no macro preprocessor, module info isn't contained in huge text files that need to be compiled over and over again, syntax is generally cleaner, and so on. But there are clones of Object Pascal and they don't get anywhere near the amazing compile times that Borland gets.
The big implementation wins, as best I can tell, are:
1. They removed the need for a traditional, general linker. Modules are written to disk in a very simple format, what seems to be a binary image that just needs a small bit of patching.
2. The compiler is simple recursive descent, without separate lexing and parsing phases.
3. Much of the core of the compiler is highly optimized, with much of the critical portion written in finely tuned assembly language.
Kdevelop and the recently released kdestudio 3.0 gold is playing hard.
You've never used Borland's Object Pascal compiler. For all intents and purposes, even going back to the early days of the Pentium, it compiles instantaneously. Give it a large project on a 333MHz PII and--bang--it is compiled and linked before you lift your finger off of the Build key. This is a huge, huge productivity boost.
Does the compiler do as much optimization as gcc? No. But it's still an optimizing compiler that gets within the "I don't care about the difference" range.
When I see people talking about needing dual Athlons to get their gcc compile times down to the single digit minutes, then I'm appalled. With Delphi you're at *zero*. That's liberating beyond belief.
It's always amusing to see the dichotomy between Slashdotters blindly hating all corporations (Borland is giving their C++ compiler away for free, but not the source code! They are evil; down with Borland!) and the love of various other corporations, like AMD, Paramount, Lucasfilm, FOX, and so on.
The majority of Slashdot readers are students without any notable software engineering experience. Sure, not everyone here fits this description, but it's certain that there will be lots of hearsay, what-my-professor-told-me responses, and misguided personal theories based on blind idealism.
Seriously, apart from a few more FPS while fragging how much of a difference do those few extra clock cycles actually make?
C'mon, we're talking about major speed boosts that increase the processor performance by up to 5%. Wait..five percent. Geez. That's a far cry from the days when we jumped from a 33MHz 486 to 66MHz. These twiddles aren't even worth discussing, especially as a bad driver could kill performance by 50% or more.
there something inherently open-source-dogma-friendly about the corporate philosophies about AMD and Transmeta (though I doubt it
It's not just Slashdot, it's the zealot personality in general. There's the perception that AMD is both the underdog and a grassroots organization run out of a cabin in Topeka, and more power to them for going up against Evil Intel (tm).
Of course the simple truth is that AMD is also a hulking corporation, just like Intel, so Slashdotters should hate them just as they hate any other company trying to make a buck.
I honestly feel that in the near future. Maybe even the recent past, we hit the point where home users will NEVER need more Mhz power.
For home users we hit that point at 200MHz. Now we're at the point where developers don't even care. I have an 866MHz Pentium III and I do hardcore commercial software development (read "games"). No one where I work is running out and buying the latest 1.4+ GHz machines. It just doesn't matter to us; we don't see a noticible difference. Most of the games that run slow on so-called midrange machines are just because of sloppy coding, not because they really need 900MHz, and I think consumers are starting to pick up on this. Game X may require a 1GHz processor. Game Y may require a 500MHz processor and have noticibly more sophisticated visuals. Hmmm.
More cycles will NOT make Word run faster.
Again, this has been true since the original 200MHz Pentium.
Are you currently searching EBay for that "steal" on a 486-DX4-100?
At one time the difference between, say, a 386-33 and 486-66 were astounding, in terms of *feel*. But a few years ago I used NT on the job running on a 200MHz Pentium. Today I use an 866MHz Pentium III and it feels about the same. Compiles are faster, sure, games run better, yes, but it's not astounding. If I upgraded to a 1.4 GHz processor I doubt it would even matter to me.
So I can profit from silly people who think that "1.4 GHz is slow" and constantly have to upgrade to whatever comes along next. The rest of us, the people doing actual work, have given up caring about CPU speed.
OFten it seems to me as though when people talk about "making money" in Open Source (or in other contexts) they seem to be talking about "making shitloads of money", often with the idea of not doing much work.
No, not really. Even a company of 25 people may need more than US$100,000 a _month_ just to pay their employees, cover rent and electricity, and so on. So to keep such a business going, you've got to at least pull in 1.2 million a year, after taxes.
Considering that many dot-coms weren't pulling in *any* income, this is very significant. You really need a good, solid, and somewhat sizable income just to keep afloat. Most of the responses in this thread are from college students thinking "I only need $500 a month to get by," and not thinking in terms of what's involved in a real business. And the question then becomes "How can I use Open Source to pull in 1.2 million a year, just so I can stay in business? That's a tough question.
Perl 6 is important. Please don't let the little details you may don't like make you forget about the fact, that Perl definitetly needed a rewrite
But when you balance the two factors:
(1) Perl 6 is most certainly an incremental improvement over Perl 5, not something completely new.
(2) Completely rewriting a huge and previously stable language.
Then it doesn't make sense. That a complete rewrite is somehow better is a standard myth among inexperienced programmers.
Bashing languages with the usual "But you can also do that in language X!" comments and general defensive put-downs is misguided. Of course you can do the same thing in any language; that's a fundamental principle of computer science. The reason we have a variety of languages is because some languages make things easier than others.
In Perl, munging through text files is a snap. The syntax is succint, and regular expressions are supported at the language level. But that doesn't mean Perl is good for everything, though. Regular expressions don't scale up to what you'd need to write a full BNF parser in Perl. And, sure, you can hook to http and ftp libraries, but they aren't integrated into the language in the same way that the "file exists?" operator is.
REBOL has several strengths. One is that its parsing features effectively *are* BNF, so you can write complex parsers for mini-languages with great ease, and without resorting to lex, yacc, and such. The other advantage is that having language-level support for internet protocols is very convenient. Sure, you can get at them through a Perl module, but if you argue that then you have to ask why regular expressions shouldn't be a separate module as well.
All this blind bashing of languages is tiring. It's exactly like the kiddies who bash whatever game console they didn't get for Christmas.
MS-DOS doesn't deserve a fond remembrance
on
MS DOS: A Eulogy
·
· Score: 4, Informative
This is a classic example of how nostalgia can be stronger than history. MS-DOS was terrible, so terrible, in many ways. It has nothing to do with the 16-bitness of it, or even driver memory crunch hell, but simply that the command prompt side of it was an embarrassment from day one.
It took over ten years before there was any kind of command history (with doskey, you could finally hit the up arrow to recall previous commands). There wasn't a real alias mechanism until doskey either. And heck--and everyone forgets this--you couldn't even properly edit the command line until doskey came along. File completion was never standard. The batch file commands were braindead and severely limited.
Sure, some third parties walked in with their own top notch command processors--most notably JP Software with 4DOS, which is still better than every UNIX shell I've ever used--but even with over a decade to work on it, the largest PC software company in the world couldn't manage to write decent command processor given years to do so. And the worst part is that it was so easy it could have been a high school project. Dr. Dobb's Journal even published the source code for a bash-like shell that replaced command.com.
I think the likely answer here is that Microsoft could have written something better, but they spent a decade trying to beat down MS-DOS and replace it with something else. Remember, Windows 1.0 shipped in 1985. So for all that time, MS-DOS users were stuck with an intentionally inferior product. It's difficult to forget the pain of those days.
Having seen the Star Wars movies as a kid, the TPM trailer gave me chills. "wow, just like the previous" movies. It was so perfect, so slick, I couldn't get over it. Then I saw the movie and _yikes_ was it horrible. Did I just grow up or was TPM just plain embarrassing? Sadly I think that you needed to be nine years old to appreciate it.
You don't go publicly badmouthing a company and burning your bridges because of private business dealings gone sour. That's what you'd expect from a 14 year old who hasn't been around.
Geez, another kiddy-kiddy game box from the kiddy-kiddy game company.
Comments like this have been amusing to watch for the last ten years. The pattern is kids love Nintendo, and make their parents buy games like crazy. Then, when they think they've outgrown kiddie stuff, they start badmouthing Nintendo as being crap for the kiddies. Gotta love it.
The new look of Windows XP to me could be described as bubbly. With the default theme, it's Ficher Price and bubbly.
And that's exactly what I thought when I first saw everyone going ga-ga over Enlightenment. Never underestimate the power of an obnoxious and goofy user interface scheme.
You and all the other idiot whiners who think they have "too much power" need to stop and think about what you are saying.
Sigh. I am a software developer. I write applications in Lisp. I make heavy use of graphic arts tools like Corel Draw. I also use 3D modelling packages. What machine do I do all this on? A 333MHz Pentium II that I bought new in 1998.
Do I have _any_ speed complaints at all? None. It is a zippy system. I can recompile the Lisp system I use--which is written in Lisp--in twenty seconds. I also do a lot of work in Delphi and I've never had a perceptible compile time yet (read "for all intents and purposes, compile time is instantaneous"). Corel Draw just zips along. The 3D modeller is more dependent on the video card than anything, so I put in a GeForce 2 and haven't had any--and I mean *any*--issues with speed.
People who talk of using the power of their 1.4 GHz processor don't have a clue. They like to think that they are a power user of some sort, and in all honesty they don't want to hear otherwise.
On the one hand you have Intel, who is trying to move into *completely* new territory, at least as far as breaking with the x86 past. Scary? Very. When Apple transitioned from the 68K to the PowerPC it was rough going for a long time. The PowerPC was much better for native applications, but those took their time in showing up. And it was much, much slower than a real 68K machine when it was emulating older code.
AMD is taking the incremental improvement route, which makes a lot of sense. But can the non-standard x86 extensions--practically a whole new processor in itself--ever be more than a niche? The 3DNow! extensions were more a novelty than anything else. Some drivers used them, most programs didn't. It's difficult as it is to support all the different computers running similar chips without getting into extensions that only work on a certain percentage of them. Is it worth shipping 64-bit Hammer code just for one market segment? It's not just a recompile; it's an entirely separate QA cycle. Thinking about hobbyists: Will they have both Itaniums and Pentiums around for testing?
And then there's the nagging doubt that we're talking about chips that are already so fast that no one cares--except a certain fanboy crowd--so now we're talking about the difference between 10x more speed than I know what to do with and 20x more speed than I know what to do with. Sure, games and some crazy high-end airflow simulation, but this begs the question of "Is it worth overturning the entire PC market just for those two minorities?"
Desktop computers that come already put together with software installed are only for sale to the masses. If you know more about computers than the average person there is no reason for you not to build your own computer.
Sure there is: It's very time consuming to track down components, learn to troubleshoot the hardware, and generally fiddle around until you get it all working. The last time I did this it took me days of web surfing, half a dozen runs to local stores, and about two weeks of total time. I have also ordered a PC from Dell in the past and it was a snap to order and set-up.
I'm always annoyed when I see the term "masses" used on Slashdot. Do you build your own car, or do you drive one of those built for "the masses." Do you drink Coke, watch The Simpsons, and buy clothes from a store? So do "the masses."
Slashdot Mirror
Embedded systems engineers are running into trouble for certain applications. On the one hand, StrongARM, MIPS, and even low-end microcontrollers like the 8051 are reliable and cool. But in the race for speed everyone has thrown power consumption to the wind. What if you need something more powerful than a StrongARM, but can't make use of Ultra SPARC and Intel processors because they simply run much too hot for embedded devices? So the embedded engineers are starting to resort to custom-designed processors tuned for specific purposes. But it would be much better if someone put effort into higher performance CPUs that didn't munch up 50 watts of power.
Power is cheap, in the end, and people want top-end speed and performance.
Rephrase that as "people like to think they need top end speed and performance" and I'll agree with you. It's the same reason people buy 4-wheel drive SUVs when they live in the midwest.
I can sympathize with the sentiment of the original poster, though not his particular situation. What you do in school is often much more fun that what you'll do on the job later. In school you can be introduced to some new idea, then get all enamored with it and have fun just learning for the sake of learning. On the job you mostly deal with a small set of known technologies and there's not much variety.
That aside, I've gotten disenchanted with computers after getting a B.S. and working professionally as a programmer for over a decade. The problem is that programming has been replaced by drowning in a sea of APIs and help files and systems that are loosely defined yet still require 1,000 pages of documentation. A certain GUI call doesn't do what you expect it to do, so you look through knowledge bases and do google searches and sift through example code. Then you find out that no one really knows the answer, so you twiddle around until it works, but then you're never sure why the fix works at all.
For example, try to understand the format of a Windows executable to the point where you could create your own (let's say you're writing a small compiler). Searching for an hour on the web turns up lots of specs, but they're all thin and full of holes and they don't agree with each other. Just *exactly* what does field X do? When do you need to set it? Why do some executables set it to $FF when the legal values are in the range of 0..3, and yet those executables still work? These kinds of issues are pervasive in any modern system, be it Windows, MacOS, or Linux+X+KDE. And they erode your soul.
Am I right in assuming that this is due to the equivalent of pre-compiled headers and overall better modularization than C(++) allows? Paraphrased: is it the compiler or the language that gives opportunity for these short compile times?
Sure, there are some things about Object Pascal that make it simpler to compile than C: no macro preprocessor, module info isn't contained in huge text files that need to be compiled over and over again, syntax is generally cleaner, and so on. But there are clones of Object Pascal and they don't get anywhere near the amazing compile times that Borland gets.
The big implementation wins, as best I can tell, are:
1. They removed the need for a traditional, general linker. Modules are written to disk in a very simple format, what seems to be a binary image that just needs a small bit of patching.
2. The compiler is simple recursive descent, without separate lexing and parsing phases.
3. Much of the core of the compiler is highly optimized, with much of the critical portion written in finely tuned assembly language.
Kdevelop and the recently released kdestudio 3.0 gold is playing hard.
You've never used Borland's Object Pascal compiler. For all intents and purposes, even going back to the early days of the Pentium, it compiles instantaneously. Give it a large project on a 333MHz PII and--bang--it is compiled and linked before you lift your finger off of the Build key. This is a huge, huge productivity boost.
Does the compiler do as much optimization as gcc? No. But it's still an optimizing compiler that gets within the "I don't care about the difference" range.
When I see people talking about needing dual Athlons to get their gcc compile times down to the single digit minutes, then I'm appalled. With Delphi you're at *zero*. That's liberating beyond belief.
So, you either have to love all corporations, or you have to hate all corporations? I guess no shades of grey are allowed.
If you are a zealot that rails against the evils of corporate control, then yes, you do.
It's always amusing to see the dichotomy between Slashdotters blindly hating all corporations (Borland is giving their C++ compiler away for free, but not the source code! They are evil; down with Borland!) and the love of various other corporations, like AMD, Paramount, Lucasfilm, FOX, and so on.
The majority of Slashdot readers are students without any notable software engineering experience. Sure, not everyone here fits this description, but it's certain that there will be lots of hearsay, what-my-professor-told-me responses, and misguided personal theories based on blind idealism.
Seriously, apart from a few more FPS while fragging how much of a difference do those few extra clock cycles actually make?
C'mon, we're talking about major speed boosts that increase the processor performance by up to 5%. Wait..five percent. Geez. That's a far cry from the days when we jumped from a 33MHz 486 to 66MHz. These twiddles aren't even worth discussing, especially as a bad driver could kill performance by 50% or more.
there something inherently open-source-dogma-friendly about the corporate philosophies about AMD and Transmeta (though I doubt it
It's not just Slashdot, it's the zealot personality in general. There's the perception that AMD is both the underdog and a grassroots organization run out of a cabin in Topeka, and more power to them for going up against Evil Intel (tm).
Of course the simple truth is that AMD is also a hulking corporation, just like Intel, so Slashdotters should hate them just as they hate any other company trying to make a buck.
Hell, even Linux distributions are starting to follow this trend.
Starting? That's how its been for years. I don't mean that in a trolling way either.
I honestly feel that in the near future. Maybe even the recent past, we hit the point where home users will NEVER need more Mhz power.
For home users we hit that point at 200MHz. Now we're at the point where developers don't even care. I have an 866MHz Pentium III and I do hardcore commercial software development (read "games"). No one where I work is running out and buying the latest 1.4+ GHz machines. It just doesn't matter to us; we don't see a noticible difference. Most of the games that run slow on so-called midrange machines are just because of sloppy coding, not because they really need 900MHz, and I think consumers are starting to pick up on this. Game X may require a 1GHz processor. Game Y may require a 500MHz processor and have noticibly more sophisticated visuals. Hmmm.
More cycles will NOT make Word run faster.
Again, this has been true since the original 200MHz Pentium.
Are you currently searching EBay for that "steal" on a 486-DX4-100?
At one time the difference between, say, a 386-33 and 486-66 were astounding, in terms of *feel*. But a few years ago I used NT on the job running on a 200MHz Pentium. Today I use an 866MHz Pentium III and it feels about the same. Compiles are faster, sure, games run better, yes, but it's not astounding. If I upgraded to a 1.4 GHz processor I doubt it would even matter to me.
So I can profit from silly people who think that "1.4 GHz is slow" and constantly have to upgrade to whatever comes along next. The rest of us, the people doing actual work, have given up caring about CPU speed.
OFten it seems to me as though when people talk about "making money" in Open Source (or in other contexts) they seem to be talking about "making shitloads of money", often with the idea of not doing much work.
No, not really. Even a company of 25 people may need more than US$100,000 a _month_ just to pay their employees, cover rent and electricity, and so on. So to keep such a business going, you've got to at least pull in 1.2 million a year, after taxes.
Considering that many dot-coms weren't pulling in *any* income, this is very significant. You really need a good, solid, and somewhat sizable income just to keep afloat. Most of the responses in this thread are from college students thinking "I only need $500 a month to get by," and not thinking in terms of what's involved in a real business. And the question then becomes "How can I use Open Source to pull in 1.2 million a year, just so I can stay in business? That's a tough question.
Perl 6 is important. Please don't let the little details you may don't like make you forget about the fact, that Perl definitetly needed a rewrite
But when you balance the two factors:
(1) Perl 6 is most certainly an incremental improvement over Perl 5, not something completely new.
(2) Completely rewriting a huge and previously stable language.
Then it doesn't make sense. That a complete rewrite is somehow better is a standard myth among inexperienced programmers.
Bashing languages with the usual "But you can also do that in language X!" comments and general defensive put-downs is misguided. Of course you can do the same thing in any language; that's a fundamental principle of computer science. The reason we have a variety of languages is because some languages make things easier than others.
In Perl, munging through text files is a snap. The syntax is succint, and regular expressions are supported at the language level. But that doesn't mean Perl is good for everything, though. Regular expressions don't scale up to what you'd need to write a full BNF parser in Perl. And, sure, you can hook to http and ftp libraries, but they aren't integrated into the language in the same way that the "file exists?" operator is.
REBOL has several strengths. One is that its parsing features effectively *are* BNF, so you can write complex parsers for mini-languages with great ease, and without resorting to lex, yacc, and such. The other advantage is that having language-level support for internet protocols is very convenient. Sure, you can get at them through a Perl module, but if you argue that then you have to ask why regular expressions shouldn't be a separate module as well.
All this blind bashing of languages is tiring. It's exactly like the kiddies who bash whatever game console they didn't get for Christmas.
This is a classic example of how nostalgia can be stronger than history. MS-DOS was terrible, so terrible, in many ways. It has nothing to do with the 16-bitness of it, or even driver memory crunch hell, but simply that the command prompt side of it was an embarrassment from day one.
It took over ten years before there was any kind of command history (with doskey, you could finally hit the up arrow to recall previous commands). There wasn't a real alias mechanism until doskey either. And heck--and everyone forgets this--you couldn't even properly edit the command line until doskey came along. File completion was never standard. The batch file commands were braindead and severely limited.
Sure, some third parties walked in with their own top notch command processors--most notably JP Software with 4DOS, which is still better than every UNIX shell I've ever used--but even with over a decade to work on it, the largest PC software company in the world couldn't manage to write decent command processor given years to do so. And the worst part is that it was so easy it could have been a high school project. Dr. Dobb's Journal even published the source code for a bash-like shell that replaced command.com.
I think the likely answer here is that Microsoft could have written something better, but they spent a decade trying to beat down MS-DOS and replace it with something else. Remember, Windows 1.0 shipped in 1985. So for all that time, MS-DOS users were stuck with an intentionally inferior product. It's difficult to forget the pain of those days.
Having seen the Star Wars movies as a kid, the TPM trailer gave me chills. "wow, just like the previous" movies. It was so perfect, so slick, I couldn't get over it. Then I saw the movie and _yikes_ was it horrible. Did I just grow up or was TPM just plain embarrassing? Sadly I think that you needed to be nine years old to appreciate it.
You don't go publicly badmouthing a company and burning your bridges because of private business dealings gone sour. That's what you'd expect from a 14 year old who hasn't been around.
Geez, another kiddy-kiddy game box from the kiddy-kiddy game company.
Comments like this have been amusing to watch for the last ten years. The pattern is kids love Nintendo, and make their parents buy games like crazy. Then, when they think they've outgrown kiddie stuff, they start badmouthing Nintendo as being crap for the kiddies. Gotta love it.
The new look of Windows XP to me could be described as bubbly. With the default theme, it's Ficher Price and bubbly.
And that's exactly what I thought when I first saw everyone going ga-ga over Enlightenment. Never underestimate the power of an obnoxious and goofy user interface scheme.
You and all the other idiot whiners who think they have "too much power" need to stop and think about what you are saying.
Sigh. I am a software developer. I write applications in Lisp. I make heavy use of graphic arts tools like Corel Draw. I also use 3D modelling packages. What machine do I do all this on? A 333MHz Pentium II that I bought new in 1998.
Do I have _any_ speed complaints at all? None. It is a zippy system. I can recompile the Lisp system I use--which is written in Lisp--in twenty seconds. I also do a lot of work in Delphi and I've never had a perceptible compile time yet (read "for all intents and purposes, compile time is instantaneous"). Corel Draw just zips along. The 3D modeller is more dependent on the video card than anything, so I put in a GeForce 2 and haven't had any--and I mean *any*--issues with speed.
People who talk of using the power of their 1.4 GHz processor don't have a clue. They like to think that they are a power user of some sort, and in all honesty they don't want to hear otherwise.
I design custom machines for our enterprise rollouts, and it does not take three days to research or replace any part.
It does if you are someone who does not design custom machines for enterprise rollouts.
On the one hand you have Intel, who is trying to move into *completely* new territory, at least as far as breaking with the x86 past. Scary? Very. When Apple transitioned from the 68K to the PowerPC it was rough going for a long time. The PowerPC was much better for native applications, but those took their time in showing up. And it was much, much slower than a real 68K machine when it was emulating older code.
AMD is taking the incremental improvement route, which makes a lot of sense. But can the non-standard x86 extensions--practically a whole new processor in itself--ever be more than a niche? The 3DNow! extensions were more a novelty than anything else. Some drivers used them, most programs didn't. It's difficult as it is to support all the different computers running similar chips without getting into extensions that only work on a certain percentage of them. Is it worth shipping 64-bit Hammer code just for one market segment? It's not just a recompile; it's an entirely separate QA cycle. Thinking about hobbyists: Will they have both Itaniums and Pentiums around for testing?
And then there's the nagging doubt that we're talking about chips that are already so fast that no one cares--except a certain fanboy crowd--so now we're talking about the difference between 10x more speed than I know what to do with and 20x more speed than I know what to do with. Sure, games and some crazy high-end airflow simulation, but this begs the question of "Is it worth overturning the entire PC market just for those two minorities?"
Desktop computers that come already put together with software installed are only for sale to the masses. If you know more about computers than the average person there is no reason for you not to build your own computer.
Sure there is: It's very time consuming to track down components, learn to troubleshoot the hardware, and generally fiddle around until you get it all working. The last time I did this it took me days of web surfing, half a dozen runs to local stores, and about two weeks of total time. I have also ordered a PC from Dell in the past and it was a snap to order and set-up.
I'm always annoyed when I see the term "masses" used on Slashdot. Do you build your own car, or do you drive one of those built for "the masses." Do you drink Coke, watch The Simpsons, and buy clothes from a store? So do "the masses."