Dataflow is old news, but there were some fuzzy words in the article which seemed to imply that they were doing some sort of lazy partial evaluation in hardware. That seems like an interesting idea, and one generally applicable to any ISA: Imagine that your compiler could mark the interesting output registers for a basic block, and then the chip could optimize away all of the side-effects! The power savings alone would be enormous, plus you could fill pipelines with just those ops which were actually useful.
Alcohol does degrade performance at effective doses, but at therapeutic doses for depression, heroin does not degrade performance of dexterity or cognition tasks.
You can simulate a brain now. It's a question of resolving power and predictive accuracy. There are good reasons to think that you will never be able to simulate a brain with fidelity adequate for a variety of useful or interesting purposes.
> 1. To produce mechanical power > 2. To intelligently apply that mechanical power
This "physics" viewpoint is pretty limited.
3. Mass-market capitalism requires a lot of consumers and very few concentrators. 4. Republican politics requires cannon fodder to keep the oil revenues flowing. 5. The stock market requires a lot of small losers to make a few big winners. 6. The CIA requires ghetto crack dealers to provide revenue for black ops. 7. Getting my glands drained requires desperate crack whores.
Poverty is just too useful to ever become obsolete.
You will gain more trust by telling people what they want to hear than you will by telling them the unvarnished truth. This because humans are strongly preferrential to reinforcement stimuli. If the truth does not fit into the world-view of the audience, the message is probably lost. They need to have categories pre-installed that adequately parse the message. You just can't tell them the raw truth unless you have established a rapor and a vocabulary. If the cateogories required to parse the message are only recently acquired, the understanding will be poor and the retention will be poor.
You're an existence proof for a Windows admin capable of operating a stable web server (under some conditions).
It's a weak argument.
The myth of Crest Electronics is offered as an existence proof for an incompetent Linux admin capable of operating an unstable server (under some conditions).
Whether the users in question were downloaders or merely cd rippers is irrelevant. The RIAA sued because they were uploading, i.e. offering the copyrighted content to share.
These were the biggest areas of pain for me: Header file dependency changes, STL namespace issues, and linking model.
The only benefit I found was smaller object code size. Oh, and debugging is somewhat improved. If you were going full-on.NET the benefits would be substantial, of course.
IANAL. The US courts have consistently held that the act of downloading is not making a copy and hence cannot violate copyright. The act of uploading is the act analogous to print publication, and uploading copyrighted material in the absence of permit or license is a violation of the copyright. I.e. leeching is legal, sharing is not.
I'd say the lesson is never to base your livelihood on selling software that runs on Windows, because if you become successful, Microsoft will destroy you. Windows is an appealling platform for ISVs only because it has so many users -- but that is merely bait for the trap, because it is precisely the vastness of its market and warchest that enables Microsoft to eat up every market sector which ISVs take the risk and expense of developing into a profitable business. It's okay, if they decide to buy you out, but more often they'll buy a cheaper competitor and throw a few million code-monkeys at it and integrate it with the platform and lower the price and modify the platform to insure that you can't compete.
No, really, Chinese is not a problem for the flag icon. The PRC flag and the Taiwanese flag and the Singaporean flag all clearly and unambiguously specify Putonghua. There just aren't any flags which are ambiguous with respect to language, not even the Indian flag, despite the fact that Hindi is a minority language.
There are, however, loads of languages which have no identifiable flag. But until you start implementing localizations to Marathi, or Hmong or Yakima, flags do work. I'm skeptical that anyone will ever localize a piece of software to an actual solely spoken language with a graphic orthography such that said language does not have a flag or suitably flag-like icon which will clearly identify that language to a primary speaker, since the use of flags for non-national ethnicities is so widespread.
An alternative to the complexity of creating a web of state modalities (volatile, nonce, automatic, private, public, protected, static, persistent, etc.) and event model entities (mutex, monitor, listener, adapter, signal, slot, section, notification) which all have to be drawn in painstaking detail by under-educated and overpaid slackers with asperger's is to eliminate state and implement systems which can be reasoned about by infallible computer programs, instead of UTA grad school drop-outs.
As long as you are restricting yourself to languages which can be unambiguously identified by flags, flags are an excellent iconic identifier. The British flag unambiguously identifies English, the French flag, French, and the Chinese flag, Chinese. Some flags may be too obscure, however. How many Spanish speakers will recognize the Spanish flag, for example? Not 100% in any case. But the real trouble comes when you want to support languages without a national flag association, of which there are thousands.
One thing seems a foregone conclusion: The winner will be an ethnic Malay, given the racial apartheid system used to suppress the 40% of the Malaysian population which is not Malay.
Slashdot Mirror
> Don't Christians have the right to defend themselves when Muslims attack them?
Not according to Jesus. (Matthew 5:38)
Dataflow is old news, but there were some fuzzy words in the article which seemed to imply that they were doing some sort of lazy partial evaluation in hardware. That seems like an interesting idea, and one generally applicable to any ISA: Imagine that your compiler could mark the interesting output registers for a basic block, and then the chip could optimize away all of the
side-effects! The power savings alone would be enormous, plus you could fill pipelines with just those ops which were actually useful.
Alcohol does degrade performance at effective doses, but at therapeutic doses for depression, heroin does not degrade performance of dexterity or cognition tasks.
> oil has been better than horsepower
tell that to the horses.
You can simulate a brain now. It's a question of resolving power and predictive accuracy. There are good reasons to think that you will never be able to simulate a brain with fidelity adequate for a variety of useful or interesting purposes.
Name a task you consider impossible of accomplishment, and I will define a schedule and a budget and a plan to complete it.
> 1. To produce mechanical power
> 2. To intelligently apply that mechanical power
This "physics" viewpoint is pretty limited.
3. Mass-market capitalism requires a lot of consumers and very few concentrators.
4. Republican politics requires cannon fodder to keep the oil revenues flowing.
5. The stock market requires a lot of small losers to make a few big winners.
6. The CIA requires ghetto crack dealers to provide revenue for black ops.
7. Getting my glands drained requires desperate crack whores.
Poverty is just too useful to ever become obsolete.
I'd rather live 10 glorious years of love and apotheosis in a Cebu City garbage heap than eek out 100 years of misery and decay in a Soho townhouse.
You will gain more trust by telling people what they want to hear than you will by telling them the unvarnished truth. This because humans are strongly preferrential to reinforcement stimuli. If the truth does not fit into the world-view of the audience, the message is probably lost. They need to have categories pre-installed that adequately parse the message. You just can't tell them the raw truth unless you have established a rapor and a vocabulary. If the cateogories required to parse the message are only recently acquired, the understanding will be poor and the retention will be poor.
Or at least remove them once they kill their first hundred thousand.
Search is how you find what you want. A browsable DAG is how you find out what it was that you wanted.
Like keeping a man-eating tiger alive by feeding it aborted fetuses helps the farmers in a nearby village... right....
Having just done a comparison of JWChat and JBother, the Java applet is orders of magnitude superior to the scrubbing cleanser version.
I would rephrase that as the spectacular success of Microsoft in their effort to make Java suck.
The only thing wrong with Java 5 is that it's not installed.
TPJ? That's a bit of private language in an otherwise clear discourse. Please disambiguate.
You're an existence proof for a Windows admin capable of operating a stable web server (under some conditions).
It's a weak argument.
The myth of Crest Electronics is offered
as an existence proof for an incompetent Linux admin capable of
operating an unstable server (under some conditions).
That's also a weak argument.
Nothing to see here.
1) Sony is in the business for profits
TRUE
2)profits come from title licenses for commercial PSP media
TRUE
3) To get media endorsement for the PSP, Sony has to demonstrate that the media and platform are a perfect lock-in solution.
FALSE
Whether the users in question were downloaders or merely cd rippers is irrelevant. The RIAA sued because they were uploading, i.e. offering the copyrighted content to share.
These were the biggest areas of pain for me: Header file dependency changes, STL namespace issues, and linking model.
.NET the benefits would be substantial, of course.
The only benefit I found was smaller object code size. Oh, and debugging is somewhat improved. If you were going full-on
IANAL. The US courts have consistently held that the act of downloading is not making a copy and hence cannot violate copyright. The act of uploading is the act analogous to print publication, and uploading copyrighted material in the absence of permit or license is a violation of the copyright. I.e. leeching is legal, sharing is not.
I'd say the lesson is never to base your livelihood on selling software that runs on Windows, because if you become successful, Microsoft will destroy you. Windows is an appealling platform for ISVs only because it has so many users -- but that is merely bait for the trap, because it is precisely the vastness of its market and warchest that enables Microsoft to eat up every market sector which ISVs take the risk and expense of developing into a profitable business. It's okay, if they decide to buy you out, but more often they'll buy a cheaper competitor and throw a few million code-monkeys at it and integrate it with the platform and lower the price and modify the platform to insure that you can't compete.
No, really, Chinese is not a problem for the flag icon. The PRC flag and the Taiwanese flag and the Singaporean flag all clearly and unambiguously specify Putonghua. There just aren't any flags which are ambiguous with respect to language, not even the Indian flag, despite the fact that Hindi is a minority language.
There are, however, loads of languages which have no identifiable flag. But until you start implementing localizations to Marathi, or Hmong or Yakima, flags do work. I'm skeptical that anyone will ever localize a piece of software to an actual solely spoken language with a graphic orthography such that said language does not have a flag or suitably flag-like icon which will clearly identify that language to a primary speaker, since the use of flags for non-national ethnicities is so widespread.
Ebonic seems like the best candidate.
An alternative to the complexity of creating a web of state modalities (volatile, nonce, automatic, private, public, protected, static, persistent, etc.) and event model entities (mutex, monitor, listener, adapter, signal, slot, section, notification) which all have to be drawn in painstaking detail by under-educated and overpaid slackers with asperger's is to eliminate state and implement systems which can be reasoned about by infallible computer programs, instead of UTA grad school drop-outs.
As long as you are restricting yourself to languages which can be unambiguously identified by flags, flags are an excellent iconic identifier. The British flag unambiguously identifies English, the French flag, French, and the Chinese flag, Chinese. Some flags may be too obscure, however. How many Spanish speakers will recognize the Spanish flag, for example? Not 100% in any case. But the real trouble comes when you want to support languages without a national flag association, of which there are thousands.
As for political sensitivity, who gives a fuck?
One thing seems a foregone conclusion: The winner will be an ethnic Malay, given the racial apartheid system used to suppress the 40% of the Malaysian population which is not Malay.