Compiling separate independent source file to objs, yes. Linking and optimization over module borders, no. (Easy to parallelize.) The basic premise was anyway that the CPUs were not the bottleneck, when in fact they are.
Yeah, but if you just would take two Banias cores, overclock and overvoltage them enough to reach the same frequency, you would still have LESS processing power and a comparable total heat dissipation. The architecture has been widened as well. I'm certainly looking forward to getting a Merom machine sometime next year, unless AMD manages to turn the mobile market around completely.
Nope. It was a matter of bus width, the 386SX was basically still using a 16-bit bus to the outside world. (Or rather 24-some bits address and 16 bit data.) The 486SX, on the other hand, was a 486DX with a non-functioning FPU.
However, if we look at instructions per cycle, those transitions were less impressive, especially the Athlon that got thrown in there. There was some added efficiency, but MOST of the gain was from the basic fact that a process shrink (and a core adjusted to make use of it) allowed higher frequencies. The frequencies are certainly possible today, but the added performance is too low to make sense at the enormous cost in power (and its closest friend, Joulian heat).
How can we be sure they are not really transformed mice that are overlooking the running of their great computer? We can't. From our limited POV, we can find it highly unlikely. You also have to make that hypothesis more stringent, do you mean learning or do you mean evolving? The latter would be completely out of the question for them as a species, as the original changes wouldn't have had time to proliferate in the gene pool in any significant way.
Hotel, meals and the airline tickets were all paid for; not loss of income, though. (But the prize money would surely outweigh that for someone good enough.)
Well, the TCO had design and development competitions as well. Those are used in real applications. And, yes, there were quite a few finalists from the US this year, although none in development. The pay is higher for design, and I would also dare to say that the advantage of using your native language to write design documents is more significant there than in development or algorithmic competitions.
(And you can do rather well in the algorithmic ones without resorting to C macros, there are some C++ coders who don't, and also plenty of folks who are rated high while using Java.)
The best US team in the ICPC finals this year was "only" in 7th place, by the MIT. antimatter (Hubert Wang) was in that team, and also at the TCO. reid (Reid Barton), ex-IOI winner, was also present. Of course, there are lots of people that have done good in other programming contests that are not TopCoders, the 75 minute format stops some people who are smart, but not so keen on just writing code like crazy. Still, I haven't met people at other contests (like the ICPC) who consider TopCoder not credible. They're just as credible as any algorithmic programming contest, i.e. it doesn't tell you too much about the skills of a software developer, but quite a lot about the skills in algorithmic/mathematic problem solving.
(And, yeah, I was there, but only in the "component development" competition. There, I can agree with your points, there are little incentives for people from countries in the West in general to participate, unless you really love competing.)
But, in short, there is no lack of US participants. They usually dominate (in sheer numbers) in the weekly "Single Round Matches", although I think Indians might be competing for the top spot (in number of registrants).
Well, it generally does (ever heard of transplant rejection?). It might be the case that these lab mice are so closely related that the rejection effect is lower than normal. A bone-marrow transplant would be one way to do it, but that's a far more complex procedure.
The next issue is that making a large collection of blood cells from different people will make the risk of transmitting some pathogen far higher (think HIV).
Well, cancer not being a disease caused by a pathogen, that won't happen. There is also little reason for viruses that eventually cause cancer (or at least "helps" the cell some steps on the way towards it) would evolve in a direction that increased that ability, if that would mean combating a strengthened immunological response. On the other hand, the individual tumors would of course have a pretty significant selection towards not expressing the markers that trigger the response. This would only work on a individual level, and it adds a few more changes that would be needed to get a "successful" cancer.
No, there are quite a lof of standard interfaces. In fact, if you just use dialog boxes and get things like your tab order right, most tools will grab it right away. If you have a custom layout engine and/or custom widgets (like a word processor that does its own menu/toolbar rendering), you of course need to implement those interfaces, no matter if it's Word or OpenOffice. Then, some of the leading tools DO also perform special things with MS Office, to improve the end result. That's been the easiest way to find a differentiator against competing products.
Please enlighten me on why accessibility in XWindows or in the total OSS environment of your choice is any different/better here. If you just do random rendering calls to the screen, a screen reader has NO WAY to pick it up properly. Any platform that supports that will also require the applications to provide an alternate representation of that data.
Where would they fly with a plane which couldn't cross the Pacific and which was loud enough to not be allowed to go supersonic over land? Australia? (Hm, and don't even think of Japanese planes going to Hawaii...)
You're taking that quote out of context. They were starting to mention that it would be phased out back in September. GINA has been with us/Windows forever. Custom GINAs never worked well with fast user switching and "Run as" in XP, either, so it's not surprising that it's replaced.
Basically, what's this is all about is that the way to alter the login process in Windows, all the way back to NT 3.1, has been a custom "GINA", that replaced part of the Ctrl-Alt-Del login process. Naturally, a lengthy biometric process migth be fine if you do it once a day, but it will both need new software and possibly some thought to work well with a LUA approach, where you need to repeat your credentials more frequently for specific operations. This is basically no different from using sudo or doing admin operations in MacOS X. It's also no different from that you can't use a custom GINA to run a specific app as admin in current Windows versions.
Massive caches and using ear plugs while the tech fans roar about it. In addition, things like more flexible load reordering and the general wideness of the architecture might alleviate the latency problem. (And, of course, this time the Intel cores won't synchronize over the FSB, but rather share the complete L2.)
Yeah, cause we all remember how a 386SX compared to a 386DX was nothing like a 486SX compared to a 486DX and what a 486DX2 was. I doubt anyone could call the reusage of the SX/DX moniker to mean something completely different "logical".
Move Creative out of that lot. All the others made their own chips (ATI Rage 128, S3 Savage, Matrox G400 or something like that at the time). Creative were just licensing S3, nVidia and 3dfx chips, a retail board maker based on the chips from others.
As I've both a 3dfx Voodoo 2 and a GeForce 256, I am not sure what you try to say about the open/closed thing. 3dfx and nVidia both sold chips and reference designs to other OEMs. ATI, on the other hand, has had a rather bumpy ride between ATI-branded retail cards, cheap OEM solutions and reatil cards made by other companies, although right now the latter dominates for both the major chip makers.
To some degree, that tells something about the inherent problems in writing a spec years before someone has even designed, or written, code that would be able to support said spec. 2D layout is often mistaken to be an easy problem. It isn't. If you add some performance concerns, and any ability to render anything before the complete file, with all dependencies (CSS, images,...) are loaded, it won't get easier.
On the other hand, what reasonably complex system is available in several standard-compliant implementations? There are deviations from even some of the fairly basic RFCs if we start looking at odd (or not so odd) cases. Some of them even make sense. Welcome to the life of writing code (even when it isn't Turing complete).
If the involved factories indeed were producing under what they thought were real licensing contracts, then it's more than just faking a consumer brand. (Now, of course the involved factories and other parties can just benefit by claiming that they were in good faith, but there might be some truth to it in some cases, who knows?)
Hey, Banias/Dothan ARE overclocked P6 chips. The changes have been gradual, but I think there is no doubt about that. They've kept the number of execution units and many other characteristics. The cache and branch predictor have indeed changed, but it did so many times before in P6 (even within the same moniker [Pentium !!!], like Katmai to Coppermine).
That's some creative history. There was a race to 1 GHz, the Intel release was a paper one. They then rushed out the 1.13 GHz, which they retired after Tom's hardware and some other sources claimed it to be unstable. It didn't return until the Tualatin die shrink. In short, Intel was hard pressed regarding the performance crown even back then. They had been all since the original Athlon at something like 600 MHz was released. They countered with the original Coppermine (the first PIII that made sense, L2 at chip clock and all other niceties).
I don't think the Borg appearance in Enterprise (considering First Contact) was the worst thing about it. It was a mere incident. There are lots of larger holes that it takes much more time to explain away.
It might be rather true for the DVD renting market, though. Just like scientific papers and the libraries that keep them have been far more influenced by electronic distribution than traditional hardcopy books.
The truth is that Apple has always been the friend of Big Brother. You just remember incorrectly. Those links are obviously fake, as they contradict the truth.
Slashdot Mirror
Compiling separate independent source file to objs, yes. Linking and optimization over module borders, no. (Easy to parallelize.) The basic premise was anyway that the CPUs were not the bottleneck, when in fact they are.
Yeah, but if you just would take two Banias cores, overclock and overvoltage them enough to reach the same frequency, you would still have LESS processing power and a comparable total heat dissipation. The architecture has been widened as well. I'm certainly looking forward to getting a Merom machine sometime next year, unless AMD manages to turn the mobile market around completely.
However, if we look at instructions per cycle, those transitions were less impressive, especially the Athlon that got thrown in there. There was some added efficiency, but MOST of the gain was from the basic fact that a process shrink (and a core adjusted to make use of it) allowed higher frequencies. The frequencies are certainly possible today, but the added performance is too low to make sense at the enormous cost in power (and its closest friend, Joulian heat).
How can we be sure they are not really transformed mice that are overlooking the running of their great computer? We can't. From our limited POV, we can find it highly unlikely. You also have to make that hypothesis more stringent, do you mean learning or do you mean evolving? The latter would be completely out of the question for them as a species, as the original changes wouldn't have had time to proliferate in the gene pool in any significant way.
Hotel, meals and the airline tickets were all paid for; not loss of income, though. (But the prize money would surely outweigh that for someone good enough.)
Every week, and most are (still) 9 PM EDT.
(And you can do rather well in the algorithmic ones without resorting to C macros, there are some C++ coders who don't, and also plenty of folks who are rated high while using Java.)
(And, yeah, I was there, but only in the "component development" competition. There, I can agree with your points, there are little incentives for people from countries in the West in general to participate, unless you really love competing.)
But, in short, there is no lack of US participants. They usually dominate (in sheer numbers) in the weekly "Single Round Matches", although I think Indians might be competing for the top spot (in number of registrants).
The next issue is that making a large collection of blood cells from different people will make the risk of transmitting some pathogen far higher (think HIV).
Well, cancer not being a disease caused by a pathogen, that won't happen. There is also little reason for viruses that eventually cause cancer (or at least "helps" the cell some steps on the way towards it) would evolve in a direction that increased that ability, if that would mean combating a strengthened immunological response. On the other hand, the individual tumors would of course have a pretty significant selection towards not expressing the markers that trigger the response. This would only work on a individual level, and it adds a few more changes that would be needed to get a "successful" cancer.
Please enlighten me on why accessibility in XWindows or in the total OSS environment of your choice is any different/better here. If you just do random rendering calls to the screen, a screen reader has NO WAY to pick it up properly. Any platform that supports that will also require the applications to provide an alternate representation of that data.
Where would they fly with a plane which couldn't cross the Pacific and which was loud enough to not be allowed to go supersonic over land? Australia? (Hm, and don't even think of Japanese planes going to Hawaii...)
You're taking that quote out of context. They were starting to mention that it would be phased out back in September. GINA has been with us/Windows forever. Custom GINAs never worked well with fast user switching and "Run as" in XP, either, so it's not surprising that it's replaced.
Basically, what's this is all about is that the way to alter the login process in Windows, all the way back to NT 3.1, has been a custom "GINA", that replaced part of the Ctrl-Alt-Del login process. Naturally, a lengthy biometric process migth be fine if you do it once a day, but it will both need new software and possibly some thought to work well with a LUA approach, where you need to repeat your credentials more frequently for specific operations. This is basically no different from using sudo or doing admin operations in MacOS X. It's also no different from that you can't use a custom GINA to run a specific app as admin in current Windows versions.
Massive caches and using ear plugs while the tech fans roar about it. In addition, things like more flexible load reordering and the general wideness of the architecture might alleviate the latency problem. (And, of course, this time the Intel cores won't synchronize over the FSB, but rather share the complete L2.)
Yeah, cause we all remember how a 386SX compared to a 386DX was nothing like a 486SX compared to a 486DX and what a 486DX2 was. I doubt anyone could call the reusage of the SX/DX moniker to mean something completely different "logical".
As I've both a 3dfx Voodoo 2 and a GeForce 256, I am not sure what you try to say about the open/closed thing. 3dfx and nVidia both sold chips and reference designs to other OEMs. ATI, on the other hand, has had a rather bumpy ride between ATI-branded retail cards, cheap OEM solutions and reatil cards made by other companies, although right now the latter dominates for both the major chip makers.
On the other hand, what reasonably complex system is available in several standard-compliant implementations? There are deviations from even some of the fairly basic RFCs if we start looking at odd (or not so odd) cases. Some of them even make sense. Welcome to the life of writing code (even when it isn't Turing complete).
If the involved factories indeed were producing under what they thought were real licensing contracts, then it's more than just faking a consumer brand. (Now, of course the involved factories and other parties can just benefit by claiming that they were in good faith, but there might be some truth to it in some cases, who knows?)
Hey, Banias/Dothan ARE overclocked P6 chips. The changes have been gradual, but I think there is no doubt about that. They've kept the number of execution units and many other characteristics. The cache and branch predictor have indeed changed, but it did so many times before in P6 (even within the same moniker [Pentium !!!], like Katmai to Coppermine).
That's some creative history. There was a race to 1 GHz, the Intel release was a paper one. They then rushed out the 1.13 GHz, which they retired after Tom's hardware and some other sources claimed it to be unstable. It didn't return until the Tualatin die shrink. In short, Intel was hard pressed regarding the performance crown even back then. They had been all since the original Athlon at something like 600 MHz was released. They countered with the original Coppermine (the first PIII that made sense, L2 at chip clock and all other niceties).
I don't think the Borg appearance in Enterprise (considering First Contact) was the worst thing about it. It was a mere incident. There are lots of larger holes that it takes much more time to explain away.
It might be rather true for the DVD renting market, though. Just like scientific papers and the libraries that keep them have been far more influenced by electronic distribution than traditional hardcopy books.
The truth is that Apple has always been the friend of Big Brother. You just remember incorrectly. Those links are obviously fake, as they contradict the truth.
Only in Korea.