Well some of the staff may be the same, except for employee #1. Killed by a redneck in a speeding camaro. Of course he was another company's employee by then, but the point is still true.
You are confusing installation and configuration on the user's system with installation and configuration done at the factory.
The GPL refers to the tools used to install the software on the user's system by the user's own hand. The software in question here is embedded, that means all the software was installed at the factory and thus any tools used to install that software are not covered by the GPL.
Seems you haven't been paying attention to China for the last 20 years. They are no more socialist than America. They are certainly fascist, which is not a requirement for socialism. But, they are probably the largest free market in the world. No country, not even America has a truly free-market, the US government is always meddling as is the Chinese, but they have come a long, long way since trying to emulate the soviet Stalin.
Unfortunately, the US government seems to like what it sees there - free markets backed with fascist social policy and is moving willy-nilly to copy them. But at least it keeps the damn dirty socialist hippies in their place, right?
Hans - How is scalability? SGI seems to think that ReiserFS doesn't scale at all with multiple CPUs, unlike ext2. At least according to the paper they presented at the Ottawa linux symposium last month:
I believe death penalty executions ought to be on national television because it is far too easy to say, "let him fry" in this country without having to really face the consquences - kind of like how I also believe no member of congress should be allowed to vote for going to war unless they have a child in the military or are themselves in the reserves and no waivers for those reservists either.
But, I don't think there should be any ads during the proceedings because it would cheapen the situation, although it would probably get more people to tune in.
One major problem with relying on labor laws is that they can be changed at the whim of a corporate donor behind a couple of politicians. For example, see the recent change in federal labor laws that make it easier to screw over exempt employees (i.e. non-union) on overtime. Sure, union leadership can and has been bought too, but if that happens too blatantly somebody ends up wearing cement boots. Congressdroid gets too blatant and they just get a cushy corporate golfing, er lawyer, job after being voted out of office.
Also, "A veto override is a political catastrophe for a President"... Yes, when has something being a political catastrophe had any sway in changing Bush's mind?
Hey, that's not insanity, it's rewarding his sponsors, er POLITICAL LEADERSHIP, yeah, yeah he's leading the country for its own good, that's the ticket!
I picked up a Minolta Magic Color 2200DL (I think that's the model number, there are a couple that are pretty close) for just over $500 from a Dell Deal(tm) a few months back. I'm not a heavy duty user, my HP 5mp is still on its original toner cartridge. The Minolta lacks postscript, so it is 'windows only' but last I checked it looked like one of the ghostscript drivers and/or something from CUPS could be adapted to do the right thing.
Plus side: Takes standard PC100 or PC133 ram, so stuck some old dimms in it to take it up to 192MB or so. Down side: It doesn't come with much RAM to begin with.
Plus side: It comes with a 100baseT port built in. Down side: Speaks an officially undocumented, but apparently well-known queueing protocol.
Plus side: It was under $600 shipped. Down side: Comes with partially filled toner cartridges, good for like 2000 pages instead of 5000 or something equally unfull.
Plus side: You can buy individual toner carts, instead of all 4 CYMK carts at once. Down side: Toner costs a lot, like $125 per cartridge.
Plus side: Prints really fast, like a real 4 ppm color and a real 16 ppm b&w Down side: Takes like two minutes to warm up out of stand-by.
YMMV, I was too lazy to double-check my facts and just went from tequila-addled memory.
The Opteron competes against the Xeon, not the itanium.
That's what Intel wants you to think. Comparing the best Opteron systems versus the best Itanium2 systems in the SpecCPU database - Integer performance is roughly equal and for FP the Itanium2 is roughly twice the speed of the Opteron.
But, on a dollars per unit work basis, the Opteron stomps the Itanium2 for both integer and FP and that's the secret that Intel is working really hard to keep their Itanium2 customers from learning. Really, the only place Itanium2 beats opteron is watts expended per unit of work, the Itanium will put out more heat than any other chip available...
Powerstrip - but the nag screen is so unintrusive and the software is otherwise 100% functional, so whenever I reinstall, I never bother to even enter my registration code. It is a useful tool that I wanted to support, so I did.
Nowadays, as in with current compilers, wrong answers are probably due to broken code more often than broken compilers. Especially with C code, but it can happen with FORTRAN too. Typically the problem is that the code author violates part of the language spec and gets themselves into a situation of relying on undefined behaviour. It just so happens that with their particular compiler at low optimizations levels, the compiler's undefined behaviour happens to produce the answers the author wants, so they don't notice they are doing anything wrong. (When the undefined behaviour produces incorrect results, the author usually notices pretty quick, ends up scratching their head for a bit and then trying something else, sometimes they get it right and end up using valid code, sometimes they get it wrong again but this time the undefined behaviour ends up producing the results they expect).
When you crank up the optimization level, compilers tend to get a lot more strict about expecting standards conformant code and they start making assumptions that are true given strict adherence to the language spec. If the code had been relying on undefined behaviour, now that behaviour is different and the result is wrong answers.
It is common for the person who is trying a new platform/compiler or optimization level to be just as much a non-expert in the language as the original author (they may even be the original author) so they aren't able to debug the conformance issues and instead they turn to the easy scapegoat and say, "the compiler's broken - it used to work fine."
So, when you hear stories of modern compilers producing code that generates wrong answers, you need to take into account both the expertise of the "accuser" in the language itself as well as their experience with the particular piece of code. If they aren't expert in both, take the accusation with a huge grain of salt.
Note, I'm not saying the NASA guy does or doesn't know what he is doing, I haven't read the reference update yet to decide for myself. I'm just speaking about the generic problem of people reporting broken compilers when it is really the code that is broken.
Except that "300DPI" by itself does not communicate any information about the physical size of the image - it is only half of the required information. Is that 300DPI at 3" by 3" or 8" by 8" or what?
It is like walking into a paint store and telling them you need three gallons. You'll get three gallons allright, fuschia, aluminum and candy-apple red.
I've read the Book of Mormon, I don't remember any parts about space ships and phaser gun thingies.
I guess you missed the FAQ on the "Galactica Code" and how, when applied to the Book of Mormon, reveals the true origins of Man, God and that Moroni guy.
Man, was I glad to see that movie released on DVD. So many people hate it, but I think it is great - even brilliant as far as that genre goes, and I've never played any of the video games either. Wish it was anamorphic though.
Check out the prelimary specs for 802.1u - power over wireless ethernet. They are planning on using microwaves to send the data and power to wireless devices around the home and office.
> Well, is the game installed on any of their computers?
It is a browser-based "tactical RPG." So, it might be bookmarked but it won't be "installed." Furthermore as a tactical game, it is about as far from a first-person-shooter as you can get.
Re:Stuff from SF we should have.
on
Science Faction
·
· Score: 1
You forgot flying cars! Where's the flying cars? I want my flying car! Popular Science promised me and the Joneses down the road a flying car by 2000, it is already more than two years after that and I still can't buy a flying car! Seems like all there is are regular cars and that damned segway!
An argument could be made that open source contributes to the economic development of poor countries, giving them more of a chance to become self-sufficient and thus less of a need to fight with their neighbors for basic resources. Of course you could also say that the more computing power a nation has the more ability they have to make war.
Or, to put it in slashdot terms, you can't 100% compare a G5 mac with a P4 X86 - they just plain aren't the same animal, despite their both being at the top of the desktop technologial lader. (flame if you want, but it's a freakin' example. count to three first.)
Yeah, I'll take the bait. If history is any guide, it will probably get me modded down by the mac zealots for heresy rather than by rational mods for being off-topic.
I do that kind of comparison for a living and yes the G5 and the P4 can be "100% compared" as well as the Opteron and the Itanium and the PA8800 and the Power5 and the UltraSparc III, etc, etc. Or at least they can all be so compared in the systems in which they are sold - which is all anyone beyond cpu-design academia cares about. So far, it is just Apple's marketing department that can't do an even-handed comparison -- really no big suprise, all marketing departments like to pick and choose their benchmarks (for example Intel loves those encryption benchmarks because Itanium has a tweak to make many encryption algorithms really fly, far beyond the performance increase that any other kind of benchmark will see), the controversy around the G5 seems to be due to Apple being so transparently inept at the game that they really went overboard, almost as if they think their customer base is too ignorant to catch them.
Once the G5 Macs become readily available just about anyone who wants to will be able to do performance comparisons at both the macro level using synthetic benchmarks like LINPACK and generic application benchmarks like the Spec and TPC suites and at the micro-level with application specific benchmarks where you take the source of the particular application that you care about and compile it yourself (or if you are a commercial customer with the big bucks you get the vendor's pre-sales tech people to compile it with all the right optimizations, maybe even tweaking the code to match cpu characteristics like cache-line size, memory prefetch latency and the number and size of registers (among others) for the areas in which the application experiences bottlenecks).
The reason such comparisons can be made is that all of these cpus are very close cousins - they are all binary, they all are procedural, they all are deterministic, as well as all supporting the same abstractions like paging and TLBs and memory-mapped I/O and caches and registers and all the other features of a modern cpu architecture, not to mention they all run flavors of unix, they all have C and FORTRAN compilers, they all have tcp/ip stacks and so on and so on.
Something more like the analogy you are trying to make would be to compare any one of those cpus to a quantum processor or even one of those fantastical sounding chromosomal/biological processor where many of the basic precepts that we take for granted today really are different. Give it another 10 years and people will be trying to make those kinds of comparisons.
For now though, I think simply saying that when comparing sports tech you have to consider the interaction with the individual human element as much as the specific tech itself is enough to get your point across to even someone as physically inept as myself.
Sorry, but "the world" is wrong. The companies that are experiencing these problems need to adapt. If they don't want to adapt, then they are welcome to not support linux - and accept the consquences of that, be they large or small.
In the long term, "the world" will be better off by changing to the linux, or rather the free software, paradigm.
For starters, and I'm sure a little more creative googling would be illuminating.
Furthermore, I have personal experience in that a very close relative got "the shakedown" by a podunk prosecutor - pay the $1000 fine and get no jail time, or go to court, spend more than $1000 on just the attorney and if you lose in front of the local judge who could easily be the prosecutors inlaw then spend at least 90 days in jail. Not the same scale as admitting to murder, but if you don't think that federal and local prosecutors make deals to get convictions instead of getting justice then you must not watch be watching enough news on television.
Ashcroft is such a huge hypocrite (does that make him a hippocrit?) - "States Rights" is his catch-all excuse for supporting states that do what he wants (biggest example is his pro-gun agenda, which I happen to agree with) - but when it comes to laws he doesn't like such as sex, drugs and death (aka gay rights, medical marijuana and assisted suicide), "States Rights" are quickly ushered out of the room.
Another example of his hypocrisy is his former strong opposition to the Clipper chip and statements made supporting the individual's right to privacy. But as soon as he was appointed to office as attorney general all that rhetoric went out the window and he quickly set to dismantling as many privacy rights as he could, and 9/11 only made that crusade easier when he was able to personally draft the USA-PATRIOT act.
Slashdot Mirror
Well some of the staff may be the same, except for employee #1. Killed by a redneck in a speeding camaro. Of course he was another company's employee by then, but the point is still true.
You are confusing installation and configuration on the user's system with installation and configuration done at the factory.
The GPL refers to the tools used to install the software on the user's system by the user's own hand. The software in question here is embedded, that means all the software was installed at the factory and thus any tools used to install that software are not covered by the GPL.
Seems you haven't been paying attention to China for the last 20 years. They are no more socialist than America. They are certainly fascist, which is not a requirement for socialism. But, they are probably the largest free market in the world. No country, not even America has a truly free-market, the US government is always meddling as is the Chinese, but they have come a long, long way since trying to emulate the soviet Stalin.
Unfortunately, the US government seems to like what it sees there - free markets backed with fascist social policy and is moving willy-nilly to copy them. But at least it keeps the damn dirty socialist hippies in their place, right?
I didn't "fail" - I chose not to. It isn't a competition with ext*, it is a question about implementation.
Hans - How is scalability? SGI seems to think that ReiserFS doesn't scale at all with multiple CPUs, unlike ext2. At least according to the paper they presented at the Ottawa linux symposium last month:
(see page 9 of this PDF for the graph)
The implication is a lack of fine-grain locking. Does this new all-atomic, all-the-time implementation automagically bring better locking too?
I believe death penalty executions ought to be on national television because it is far too easy to say, "let him fry" in this country without having to really face the consquences - kind of like how I also believe no member of congress should be allowed to vote for going to war unless they have a child in the military or are themselves in the reserves and no waivers for those reservists either.
But, I don't think there should be any ads during the proceedings because it would cheapen the situation, although it would probably get more people to tune in.
Every OS makes use of TrueType fonts.
Damn! CP/M sure looks sweet with those true-type fonts!
One major problem with relying on labor laws is that they can be changed at the whim of a corporate donor behind a couple of politicians. For example, see the recent change in federal labor laws that make it easier to screw over exempt employees (i.e. non-union) on overtime. Sure, union leadership can and has been bought too, but if that happens too blatantly somebody ends up wearing cement boots. Congressdroid gets too blatant and they just get a cushy corporate golfing, er lawyer, job after being voted out of office.
Also, "A veto override is a political catastrophe for a President"... Yes, when has something being a political catastrophe had any sway in changing Bush's mind?
Hey, that's not insanity, it's rewarding his sponsors, er POLITICAL LEADERSHIP, yeah, yeah he's leading the country for its own good, that's the ticket!
I picked up a Minolta Magic Color 2200DL (I think that's the model number, there are a couple that are pretty close) for just over $500 from a Dell Deal(tm) a few months back. I'm not a heavy duty user, my HP 5mp is still on its original toner cartridge. The Minolta lacks postscript, so it is 'windows only' but last I checked it looked like one of the ghostscript drivers and/or something from CUPS could be adapted to do the right thing.
Plus side: Takes standard PC100 or PC133 ram, so stuck some old dimms in it to take it up to 192MB or so.
Down side: It doesn't come with much RAM to begin with.
Plus side: It comes with a 100baseT port built in.
Down side: Speaks an officially undocumented, but apparently well-known queueing protocol.
Plus side: It was under $600 shipped.
Down side: Comes with partially filled toner cartridges, good for like 2000 pages instead of 5000 or something equally unfull.
Plus side: You can buy individual toner carts, instead of all 4 CYMK carts at once.
Down side: Toner costs a lot, like $125 per cartridge.
Plus side: Prints really fast, like a real 4 ppm color and a real 16 ppm b&w
Down side: Takes like two minutes to warm up out of stand-by.
YMMV, I was too lazy to double-check my facts and just went from tequila-addled memory.
The Opteron competes against the Xeon, not the itanium.
That's what Intel wants you to think. Comparing the best Opteron systems versus the best Itanium2 systems in the SpecCPU database - Integer performance is roughly equal and for FP the Itanium2 is roughly twice the speed of the Opteron.
But, on a dollars per unit work basis, the Opteron stomps the Itanium2 for both integer and FP and that's the secret that Intel is working really hard to keep their Itanium2 customers from learning. Really, the only place Itanium2 beats opteron is watts expended per unit of work, the Itanium will put out more heat than any other chip available...
Powerstrip - but the nag screen is so unintrusive and the software is otherwise 100% functional, so whenever I reinstall, I never bother to even enter my registration code. It is a useful tool that I wanted to support, so I did.
Nowadays, as in with current compilers, wrong answers are probably due to broken code more often than broken compilers. Especially with C code, but it can happen with FORTRAN too. Typically the problem is that the code author violates part of the language spec and gets themselves into a situation of relying on undefined behaviour. It just so happens that with their particular compiler at low optimizations levels, the compiler's undefined behaviour happens to produce the answers the author wants, so they don't notice they are doing anything wrong. (When the undefined behaviour produces incorrect results, the author usually notices pretty quick, ends up scratching their head for a bit and then trying something else, sometimes they get it right and end up using valid code, sometimes they get it wrong again but this time the undefined behaviour ends up producing the results they expect).
When you crank up the optimization level, compilers tend to get a lot more strict about expecting standards conformant code and they start making assumptions that are true given strict adherence to the language spec. If the code had been relying on undefined behaviour, now that behaviour is different and the result is wrong answers.
It is common for the person who is trying a new platform/compiler or optimization level to be just as much a non-expert in the language as the original author (they may even be the original author) so they aren't able to debug the conformance issues and instead they turn to the easy scapegoat and say, "the compiler's broken - it used to work fine."
So, when you hear stories of modern compilers producing code that generates wrong answers, you need to take into account both the expertise of the "accuser" in the language itself as well as their experience with the particular piece of code. If they aren't expert in both, take the accusation with a huge grain of salt.
Note, I'm not saying the NASA guy does or doesn't know what he is doing, I haven't read the reference update yet to decide for myself. I'm just speaking about the generic problem of people reporting broken compilers when it is really the code that is broken.
One project is working on a new standard for memory in DIMM form - the HCC DIMM - Hacker Checking and Correcting memory.
Except that "300DPI" by itself does not communicate any information about the physical size of the image - it is only half of the required information. Is that 300DPI at 3" by 3" or 8" by 8" or what?
It is like walking into a paint store and telling them you need three gallons. You'll get three gallons allright, fuschia, aluminum and candy-apple red.
I've read the Book of Mormon, I don't remember any parts about space ships and phaser gun thingies.
I guess you missed the FAQ on the "Galactica Code" and how, when applied to the Book of Mormon, reveals the true origins of Man, God and that Moroni guy.
Trust the Fungus -Luigi
Man, was I glad to see that movie released on DVD. So many people hate it, but I think it is great - even brilliant as far as that genre goes, and I've never played any of the video games either. Wish it was anamorphic though.
Check out the prelimary specs for 802.1u - power over wireless ethernet. They are planning on using microwaves to send the data and power to wireless devices around the home and office.
> Well, is the game installed on any of their computers?
It is a browser-based "tactical RPG." So, it might be bookmarked but it won't be "installed." Furthermore as a tactical game, it is about as far from a first-person-shooter as you can get.
You forgot flying cars! Where's the flying cars? I want my flying car! Popular Science promised me and the Joneses down the road a flying car by 2000, it is already more than two years after that and I still can't buy a flying car! Seems like all there is are regular cars and that damned segway!
An argument could be made that open source contributes to the economic development of poor countries, giving them more of a chance to become self-sufficient and thus less of a need to fight with their neighbors for basic resources. Of course you could also say that the more computing power a nation has the more ability they have to make war.
Or, to put it in slashdot terms, you can't 100% compare a G5 mac with a P4 X86 - they just plain aren't the same animal, despite their both being at the top of the desktop technologial lader. (flame if you want, but it's a freakin' example. count to three first.)
Yeah, I'll take the bait. If history is any guide, it will probably get me modded down by the mac zealots for heresy rather than by rational mods for being off-topic.
I do that kind of comparison for a living and yes the G5 and the P4 can be "100% compared" as well as the Opteron and the Itanium and the PA8800 and the Power5 and the UltraSparc III, etc, etc. Or at least they can all be so compared in the systems in which they are sold - which is all anyone beyond cpu-design academia cares about. So far, it is just Apple's marketing department that can't do an even-handed comparison -- really no big suprise, all marketing departments like to pick and choose their benchmarks (for example Intel loves those encryption benchmarks because Itanium has a tweak to make many encryption algorithms really fly, far beyond the performance increase that any other kind of benchmark will see), the controversy around the G5 seems to be due to Apple being so transparently inept at the game that they really went overboard, almost as if they think their customer base is too ignorant to catch them.
Once the G5 Macs become readily available just about anyone who wants to will be able to do performance comparisons at both the macro level using synthetic benchmarks like LINPACK and generic application benchmarks like the Spec and TPC suites and at the micro-level with application specific benchmarks where you take the source of the particular application that you care about and compile it yourself (or if you are a commercial customer with the big bucks you get the vendor's pre-sales tech people to compile it with all the right optimizations, maybe even tweaking the code to match cpu characteristics like cache-line size, memory prefetch latency and the number and size of registers (among others) for the areas in which the application experiences bottlenecks).
The reason such comparisons can be made is that all of these cpus are very close cousins - they are all binary, they all are procedural, they all are deterministic, as well as all supporting the same abstractions like paging and TLBs and memory-mapped I/O and caches and registers and all the other features of a modern cpu architecture, not to mention they all run flavors of unix, they all have C and FORTRAN compilers, they all have tcp/ip stacks and so on and so on.
Something more like the analogy you are trying to make would be to compare any one of those cpus to a quantum processor or even one of those fantastical sounding chromosomal/biological processor where many of the basic precepts that we take for granted today really are different. Give it another 10 years and people will be trying to make those kinds of comparisons.
For now though, I think simply saying that when comparing sports tech you have to consider the interaction with the individual human element as much as the specific tech itself is enough to get your point across to even someone as physically inept as myself.
Sorry, but "the world" is wrong. The companies that are experiencing these problems need to adapt. If they don't want to adapt, then they are welcome to not support linux - and accept the consquences of that, be they large or small.
In the long term, "the world" will be better off by changing to the linux, or rather the free software, paradigm.
For starters, and I'm sure a little more creative googling would be illuminating.
Furthermore, I have personal experience in that a very close relative got "the shakedown" by a podunk prosecutor - pay the $1000 fine and get no jail time, or go to court, spend more than $1000 on just the attorney and if you lose in front of the local judge who could easily be the prosecutors inlaw then spend at least 90 days in jail. Not the same scale as admitting to murder, but if you don't think that federal and local prosecutors make deals to get convictions instead of getting justice then you must not watch be watching enough news on television.
Ashcroft is such a huge hypocrite (does that make him a hippocrit?) - "States Rights" is his catch-all excuse for supporting states that do what he wants (biggest example is his pro-gun agenda, which I happen to agree with) - but when it comes to laws he doesn't like such as sex, drugs and death (aka gay rights, medical marijuana and assisted suicide), "States Rights" are quickly ushered out of the room.
Another example of his hypocrisy is his former strong opposition to the Clipper chip and statements made supporting the individual's right to privacy. But as soon as he was appointed to office as attorney general all that rhetoric went out the window and he quickly set to dismantling as many privacy rights as he could, and 9/11 only made that crusade easier when he was able to personally draft the USA-PATRIOT act.