You don't understand this - these are no PCs you can run Windows on.
No, it's you who doesn't understand. What makes you think I'd want to run Windows on it? I'm sure the operating system overhead is lower than for Windows, but probably not enough to make up for a four-fold reduction in cycles.
RISC_OS
Heh, 4 full time developers?:-) Even the most obscure Linux port probably has four full-time-equivalent developers.
So it's another proprietary desktop operating system, but one that only runs on slow expensive machines, has an infinitesimal marketshare and development community, isn't unix-compatible, and has a graphic design stuck in the early 90s? I should care why?
Any few of those I could put up with (free/non-unix/marginal, or proprietary/gorgeous) but the combination is pretty damning.
I'm happy to admit that if I'd gone to school in the UK, I'd probably feel all nostalgic and want to buy one. But I didn't and I don't. And I especially don't want to support the kind of rat-bastardry discussed in the original story.
Yeah, and I see they're selling a *600MHz* 40GB 128MB machine (no peripherals) for UKP1250, or about AUD3750. That's pretty ridiculous; surely about five times the price you ought to be paying for that capability. No wonder they need to resort to stealing other people's code.
Heh, I see they don't even have the guts to run their web site on their own OS:
Here's a step-by-step walkthrough proving that a scumbag called Stephen Kapp wripped off some GPL'd code that I wrote.
You can see that the functionality is very similar, and that by using some simple tools to compare the binaries we can see they're the same on the inside. What we see here is mostly the names of functions which carry through from the source to the binary, to support debugging, run-time-linking, and similar things. The LKML post said that these strings were also found to be the same, and then Castle removed them, which is clearly evidence that they knew they'd been caught stealing.
Basically we're looking for patterns and similarities. Although somebody could (legally) write a program that had the same user interface, it's astronomically unlikely that they'd call their functions the same thing and have the exact same design.
Stephen Kapp crawled back under his rock once this was published. I expect Castle will do something similar: just continue to deny that it ever happened, but remove the code.
Similar things have happened to other major open-source projects. It hasn't been to court yet because, as far I know, every case has either been an honest misunderstanding, or an intentional violation but the perpetrator skulked away when challenged. I suppose in both cases it's not worth the FSF's time&money to take it further, but the drawback is that there's no clear example to others.
I really hope the FSF does help the copyright owners bring a lawsuit, it's time for a demonstration and I'd certainly throw in a hundred bucks to help fund it.
This isn't just a free software problem though: people who publish proprietary reusable code (development libraries,...) have lots of trouble preventing copyright infringement. I don't think I've ever seen a Windows developer's machine that didn't have pirated, or at least unregistered-shareware, software.
At the risk of stating the bloody obvious: nonverbal cues communicate different things in different cultures and subcultures.
Many of them are the same: respect for your audience at a hacker conference includes: good presentation, speaking clearly, making eye contact, being moderately entertaining, planning your talk, answering questions politely and attentively, etc.
Some nonverbal cues are interpreted differently: wearing a suit would send a message that you are a Suit, and not really part of the group you're speaking too. Perhaps that's the message you want! It might well be so, if your talk is about "A statistician's/lawyer's/CIO's look at Linux".
Respect implies *consideration* for your audience, not necessarily a tie as such.
I completely agree about erring on the side of trust.
However, I have been in touch with the LCA organizers -- who were naturally very concerned -- and it seems fairly clearly not true:
1. There's been no protest or comment from the German diplomatic corps, despite the claim they were called to the airport.
2. No German KDE developers were invited to the conference.
3. If somebody had been invited, or even registered, and not shown up, the organizers would have known. There was no such unexplained case. (It's hard to believe somebody would fly from Europe without registering first, especially since the conference sold out.)
4. They didn't say anything to the organizers, even on their return home. Their only apparent protest is an AC post on Slashdot.
5. The details in general are just a bit hard to believe, e.g. Qantas vs Customs.
"Did you hear the one about the rapist at the local shopping mall?"
Certainly people should feel pissed off with Howard's "Wonderful suck-holing" exercise, but not for this particular story.
He says that as a German citizen he was handed over to some kind of German consular official bloody and beaten. It seems a bit hard to believe that there would be no reaction of any kind from the German government. Even if he wanted to remain anonymous, you'd think they'd at least be able to confirm the general circumstances or demand an apology.
I wouldn't say this sort of thing never happens, but I think this one is a troll.
Your sympathy does you credit, but remember there is a lot of bullshit on slashdot. If it's true I would be aghast and disappointed, but there really isn't enough detail to say.
The original poster really should report this to, say, the Australian (major newspaper) or Four Corners (serious investigative TV programme). If there's any truth to it, they'll get the story out.
Paul Russell (ipchains author) described this kind of technique, and at least some countermeasures, about three years ago. I heard him verbally describe it, and I'm fairly sure it's in a FAQ somewhere.
Nice start. (Leaving aside the technical problems discussed below.)
What I'm *really* looking forward to is Microsoft's release of a "Linux Affinity Kit" (as for AIX, Solaris, HPUX, SCO,...) to allow Linux binaries to run directly on top of Windows. I reckon they'll release it sometime in the next four years.
The US didn't beat the Communist bloc by nuking it into the ground, they just gradually "persuaded" it of the advantages of freedom. It'll be like this for Linux and Microsoft.
I've never seen the AIX source, but I imagine it must be tens or even hundreds of millions of lines. Probably it depends upon IBM-specific tools for code generation, compilation, testing, or build management, just because they loved reinvention when it was written. ("Pels" not "pixels", etc...) So even making sense of it would depend on learning whatever wacky IBM alternatives to make, bison, as, gcc, etc, were used.
If they just stuck it up on an FTP site nobody would be even able to build it, let alone usefully contribute back. Imagine the lag that Mozilla has had in getting outside developers in, but multiplied by a factor of ten. Sure, perhaps after a few years some hardy souls would have started to make headway, but it would be more or less irrelevant by then.
Nah, they're going about it the right way: move *money*, *mindshare*, *people* and *ideas* from AIX to Linux. That's far more useful than sourcecode.
One of the things I thought was cool about OpenMosix was the general purpose approach it takes. If I can merely write my program to spawn many subprocesses (without shared memory) then I can easily and transparently leverage OpenMosix just as easily as an SMP box.
I agree that's a very cool feature. Of course the performance characteristics may be quite different to an SMP box, so getting it running is not the same as getting it running well. If memory is bounced between machines over Fast Ethernet it'll be much slower than anything inside a PC.
Of course although OpenMOSIX is more general than distcc, there are some thing that it won't handle well. If you need a 1TB single image, then a pile of PCs doesn't do it.
A massive POV rendering job,...
I don't know for sure, but I think the folks who do CGI for movies have their own special-purpose distribution systems conceptually similar to distcc. On a job that size, a few programmer-months to write it would pay off nicely. And of course they probably use something like RenderMan rather than POV.f
Re:Quick Summary... and a Why?
on
Xbox Linux Cluster
·
· Score: 0, Flamebait
At least in Australia, plenty of people seem to be giving them away for free as prizes, or whatever. (Presumably M$ are trying to prime demand for games...)
If you got one at no cost it might be good to try something like this.
Well, because for the particular case of building software, distcc probably does a better job than OpenMOSIX -- although to be fair nobody has done back-to-back tests yet. Have a look on the web site for more details, but briefly:
distcc is easier to install: you don't need a kernel patch, you don't need root, you don't even need all the machines to be running Linux or the same CPU architecture.
distcc is optimized for compilation: it's not clear that MOSIX's process migration/IO redirection stuff is really very efficient for this type of work. For some projects, distcc gets over 90% of the theoretical max scalability, which is pretty good.
A small, specific tool generally beats a complex generalized one. OpenMOSIX is very cool, but if you want faster builds easily distcc is the way to go.
If you buy a bitkeeper licence, then presumably you're fine. This debate is about the no-cost licence.
That licence gives lm the right to change the licence in the future and to revoke your licence. He has demonstrated publicly that he is very willing to do this to people who annoy him. The licence is essentially the "don't piss off larry" licence.
lm insists that since he's allowing people to use it at no charge, he is able to impose whatever conditions he wants. That may or may not be morally or legally fair. All I'm suggesting is that it's a bad idea to take him up on the offer.
I have plenty of notice to extract revision history from it and feed it into my new software before it stops working.
No, you may not. For example, with no warning, Larry announced that people who have sent patches to arch, cvs, or subversion are not allowed to use bk. If you're part-way through a project that depends on it, you're screwed.
Even if you don't fit into this category, then I think you ought to be concerned that you might be next.
Suppose RMS was always changing the GPL to ban Microsoft, or Amazon, or whoever he dislikes this week. Even though I'm not at Microsoft, it would make me pretty worried about the potential future changes.
If I'm a good developer and back up my working copy daily then I'll always have a recent snapshot of my source anyway.
Sure, but merely having your working copy is a poor compensation for losing your change history. If that was all you wanted, and you made backups every day, you wouldn't need a CM system at all.
If you want a proprietary system, look at BK: it's good technology, although not as novel as lm says it is. However, if you want to do free software, I think it's a pretty poor idea.
Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it.
Getting so many people out was a pretty impressive effort, I agree. In the newspaper tables this morning of previous fires that destroyed a similar number of houses the number of fatalities was sadly substantially higher -- I think about 60 for Ash Wednesday. Having good, relatively empty roads, good communication, and emergency services people doing what they could covers most of it.
Whether or not it is true that there is nothing to be done, I'm sure it would be politically unacceptable for Stanhope to do nothing. "These things happen", however true, is not going to go down well. Knowing that Something Must Be Done, I'm curious what that Something will be.
Aussies might make a bitter joke of it, but not such a lame joke as the parent post. Real spirit is shown by the gentleman quoted in the SMH:
Greg Matthews does not rise from the wooden garden bench where he sits beside his father, on the footpath, facing his burnt-out home. "G'day", he deadpans, "we'd offer you a cuppa, but we're out of milk."
The seat is somehow undamaged by the flames; so are Mr Matthews and his family, but that is about all. Everything else is gone -- the house, its entire contents, three dogs, the parrot, and about 700 tropical fish.
"It was just terrible, mate," he says, pointing to Mt Arawang, 1.5 kilometers to the west, "from the top of that hill to the bottom of our street, no word of a lie, the flames took no more than about five minutes. The noise was deafening. It sounded like a Boeing jet."
And insurance?
"My insurance was a dog in the backyard," he said. That, and a fire hydrant on the footpath, not 10 meters from where he now sits. But that proved to be no more protection than the dog....
The fire chief was quoted saying it would have taken about 500 trucks to properly fight the fire.
Of course, the problem is, if you buy enough equipment to deal with a 99th-percentile fire, you will have 500 trucks sitting idle most of the time.
I'm interested to see what will change in future years. Even if this was an unpreventable fluke, I don't think it will be politically acceptable to do nothing. Perhaps stricter building codes? (Fire shutters? No wood houses?) Perhaps more fire trucks, or emergency water reservoirs?
You know, looking yesterday at the eerie black smoke and red light in the sky, and the strings of helicopters going across, I thought that this must have been a *tiny* taste of what it was like in the Gulf. Thank goodness they were dropping water not bombs.
I am a bit hopeful that this will make Australians think twice about participating in a slaughter in Iraq. Losing four people is bad enough -- the US & friends killed about 100,000 Iraqis last time. If there is any reasonable way to avoid it, we should.
Slashdot Mirror
You don't understand this - these are no PCs you can run Windows on.
:-) Even the most obscure Linux port probably has four full-time-equivalent developers.
No, it's you who doesn't understand. What makes you think I'd want to run Windows on it? I'm sure the operating system overhead is lower than for Windows, but probably not enough to make up for a four-fold reduction in cycles.
RISC_OS
Heh, 4 full time developers?
So it's another proprietary desktop operating system, but one that only runs on slow expensive machines, has an infinitesimal marketshare and development community, isn't unix-compatible, and has a graphic design stuck in the early 90s? I should care why?
Any few of those I could put up with (free/non-unix/marginal, or proprietary/gorgeous) but the combination is pretty damning.
I'm happy to admit that if I'd gone to school in the UK, I'd probably feel all nostalgic and want to buy one. But I didn't and I don't. And I especially don't want to support the kind of rat-bastardry discussed in the original story.
Yeah, and I see they're selling a *600MHz* 40GB 128MB machine (no peripherals) for UKP1250, or about AUD3750. That's pretty ridiculous; surely about five times the price you ought to be paying for that capability. No wonder they need to resort to stealing other people's code.
Heh, I see they don't even have the guts to run their web site on their own OS:
Server: Apache/1.3.27 (Unix) mod_throttle/3.1.2 PHP/4.1.2 mod_perl/1.27
217.204.38.36:80 * Novell Netware TCP/IP
Here's a step-by-step walkthrough proving that a scumbag called Stephen Kapp wripped off some GPL'd code that I wrote.
...) have lots of trouble preventing copyright infringement. I don't think I've ever seen a Windows developer's machine that didn't have pirated, or at least unregistered-shareware, software.
You can see that the functionality is very similar, and that by using some simple tools to compare the binaries we can see they're the same on the inside. What we see here is mostly the names of functions which carry through from the source to the binary, to support debugging, run-time-linking, and similar things. The LKML post said that these strings were also found to be the same, and then Castle removed them, which is clearly evidence that they knew they'd been caught stealing.
Basically we're looking for patterns and similarities. Although somebody could (legally) write a program that had the same user interface, it's astronomically unlikely that they'd call their functions the same thing and have the exact same design.
Stephen Kapp crawled back under his rock once this was published. I expect Castle will do something similar: just continue to deny that it ever happened, but remove the code.
Similar things have happened to other major open-source projects. It hasn't been to court yet because, as far I know, every case has either been an honest misunderstanding, or an intentional violation but the perpetrator skulked away when challenged. I suppose in both cases it's not worth the FSF's time&money to take it further, but the drawback is that there's no clear example to others.
I really hope the FSF does help the copyright owners bring a lawsuit, it's time for a demonstration and I'd certainly throw in a hundred bucks to help fund it.
This isn't just a free software problem though: people who publish proprietary reusable code (development libraries,
At the risk of stating the bloody obvious: nonverbal cues communicate different things in different cultures and subcultures.
Many of them are the same: respect for your audience at a hacker conference includes: good presentation, speaking clearly, making eye contact, being moderately entertaining, planning your talk, answering questions politely and attentively, etc.
Some nonverbal cues are interpreted differently: wearing a suit would send a message that you are a Suit, and not really part of the group you're speaking too. Perhaps that's the message you want! It might well be so, if your talk is about "A statistician's/lawyer's/CIO's look at Linux".
Respect implies *consideration* for your audience, not necessarily a tie as such.
I completely agree about erring on the side of trust.
However, I have been in touch with the LCA organizers -- who were naturally very concerned -- and it seems fairly clearly not true:
1. There's been no protest or comment from the German diplomatic corps, despite the claim they were called to the airport.
2. No German KDE developers were invited to the conference.
3. If somebody had been invited, or even registered, and not shown up, the organizers would have known. There was no such unexplained case. (It's hard to believe somebody would fly from Europe without registering first, especially since the conference sold out.)
4. They didn't say anything to the organizers, even on their return home. Their only apparent protest is an AC post on Slashdot.
5. The details in general are just a bit hard to believe, e.g. Qantas vs Customs.
"Did you hear the one about the rapist at the local shopping mall?"
Certainly people should feel pissed off with Howard's "Wonderful suck-holing" exercise, but not for this particular story.
He says that as a German citizen he was handed over to some kind of German consular official bloody and beaten. It seems a bit hard to believe that there would be no reaction of any kind from the German government. Even if he wanted to remain anonymous, you'd think they'd at least be able to confirm the general circumstances or demand an apology.
I wouldn't say this sort of thing never happens, but I think this one is a troll.
Matt,
Your sympathy does you credit, but remember there is a lot of bullshit on slashdot. If it's true I would be aghast and disappointed, but there really isn't enough detail to say.
The original poster really should report this to, say, the Australian (major newspaper) or Four Corners (serious investigative TV programme). If there's any truth to it, they'll get the story out.
Paul Russell (ipchains author) described this kind of technique, and at least some countermeasures, about three years ago. I heard him verbally describe it, and I'm fairly sure it's in a FAQ somewhere.
"rigid like diarrhea"? That's the wierdest metaphor I've heard for a while.
Well, closed source is filled with self-righteous, immature college dropouts too. So what?
Nice start. (Leaving aside the technical problems discussed below.)
...) to allow Linux binaries to run directly on top of Windows. I reckon they'll release it sometime in the next four years.
What I'm *really* looking forward to is Microsoft's release of a "Linux Affinity Kit" (as for AIX, Solaris, HPUX, SCO,
The US didn't beat the Communist bloc by nuking it into the ground, they just gradually "persuaded" it of the advantages of freedom. It'll be like this for Linux and Microsoft.
I've never seen the AIX source, but I imagine it must be tens or even hundreds of millions of lines. Probably it depends upon IBM-specific tools for code generation, compilation, testing, or build management, just because they loved reinvention when it was written. ("Pels" not "pixels", etc...) So even making sense of it would depend on learning whatever wacky IBM alternatives to make, bison, as, gcc, etc, were used.
If they just stuck it up on an FTP site nobody would be even able to build it, let alone usefully contribute back. Imagine the lag that Mozilla has had in getting outside developers in, but multiplied by a factor of ten. Sure, perhaps after a few years some hardy souls would have started to make headway, but it would be more or less irrelevant by then.
Nah, they're going about it the right way: move *money*, *mindshare*, *people* and *ideas* from AIX to Linux. That's far more useful than sourcecode.
One of the things I thought was cool about OpenMosix was the general purpose approach it takes. If I can merely write my program to spawn many subprocesses (without shared memory) then I can easily and transparently leverage OpenMosix just as easily as an SMP box.
I agree that's a very cool feature. Of course the performance characteristics may be quite different to an SMP box, so getting it running is not the same as getting it running well. If memory is bounced between machines over Fast Ethernet it'll be much slower than anything inside a PC.
Of course although OpenMOSIX is more general than distcc, there are some thing that it won't handle well. If you need a 1TB single image, then a pile of PCs doesn't do it.
A massive POV rendering job,...
I don't know for sure, but I think the folks who do CGI for movies have their own special-purpose distribution systems conceptually similar to distcc. On a job that size, a few programmer-months to write it would pay off nicely. And of course they probably use something like RenderMan rather than POV.f
At least in Australia, plenty of people seem to be giving them away for free as prizes, or whatever. (Presumably M$ are trying to prime demand for games...)
If you got one at no cost it might be good to try something like this.
Well, because for the particular case of building software, distcc probably does a better job than OpenMOSIX -- although to be fair nobody has done back-to-back tests yet. Have a look on the web site for more details, but briefly:
distcc is easier to install: you don't need a kernel patch, you don't need root, you don't even need all the machines to be running Linux or the same CPU architecture.
distcc is optimized for compilation: it's not clear that MOSIX's process migration/IO redirection stuff is really very efficient for this type of work. For some projects, distcc gets over 90% of the theoretical max scalability, which is pretty good.
A small, specific tool generally beats a complex generalized one. OpenMOSIX is very cool, but if you want faster builds easily distcc is the way to go.
If you buy a bitkeeper licence, then presumably you're fine. This debate is about the no-cost licence.
That licence gives lm the right to change the licence in the future and to revoke your licence. He has demonstrated publicly that he is very willing to do this to people who annoy him. The licence is essentially the "don't piss off larry" licence.
lm insists that since he's allowing people to use it at no charge, he is able to impose whatever conditions he wants. That may or may not be morally or legally fair. All I'm suggesting is that it's a bad idea to take him up on the offer.
I have plenty of notice to extract revision history from it and feed it into my new software before it stops working.
No, you may not. For example, with no warning, Larry announced that people who have sent patches to arch, cvs, or subversion are not allowed to use bk. If you're part-way through a project that depends on it, you're screwed.
Even if you don't fit into this category, then I think you ought to be concerned that you might be next.
Suppose RMS was always changing the GPL to ban Microsoft, or Amazon, or whoever he dislikes this week. Even though I'm not at Microsoft, it would make me pretty worried about the potential future changes.
If I'm a good developer and back up my working copy daily then I'll always have a recent snapshot of my source anyway.
Sure, but merely having your working copy is a poor compensation for losing your change history. If that was all you wanted, and you made backups every day, you wouldn't need a CM system at all.
If you want a proprietary system, look at BK: it's good technology, although not as novel as lm says it is. However, if you want to do free software, I think it's a pretty poor idea.
The changes you suggest would enormously reduce clarity. ntofs(x) or ntofs(x, x) does not indicate that x is updated.
If the macro produces
x = x;
then the compiler can optimize it out.
Obfuscating the code to work around a checker bug (not understanding the conditionality of the macro) is stupid.
And on a similar note:
Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it.
-- Brian W. Kernighan
This is just begging for a rotten.com link, isn't it? I've seen photos of pyrotechnics injuries, and they're not pretty.
Getting so many people out was a pretty impressive effort, I agree. In the newspaper tables this morning of previous fires that destroyed a similar number of houses the number of fatalities was sadly substantially higher -- I think about 60 for Ash Wednesday. Having good, relatively empty roads, good communication, and emergency services people doing what they could covers most of it.
Whether or not it is true that there is nothing to be done, I'm sure it would be politically unacceptable for Stanhope to do nothing. "These things happen", however true, is not going to go down well. Knowing that Something Must Be Done, I'm curious what that Something will be.
I was thinking about volunteering once things have settled down. I suppose they'll have more volunteers than equipment.
You poor bastard.
The fire chief was quoted saying it would have taken about 500 trucks to properly fight the fire.
Of course, the problem is, if you buy enough equipment to deal with a 99th-percentile fire, you will have 500 trucks sitting idle most of the time.
I'm interested to see what will change in future years. Even if this was an unpreventable fluke, I don't think it will be politically acceptable to do nothing. Perhaps stricter building codes? (Fire shutters? No wood houses?) Perhaps more fire trucks, or emergency water reservoirs?
Please don't link to the canberraconnect web site -- it seems to already be overloaded just with people who have a bona fide interest.
You know, looking yesterday at the eerie black smoke and red light in the sky, and the strings of helicopters going across, I thought that this must have been a *tiny* taste of what it was like in the Gulf. Thank goodness they were dropping water not bombs.
I am a bit hopeful that this will make Australians think twice about participating in a slaughter in Iraq. Losing four people is bad enough -- the US & friends killed about 100,000 Iraqis last time. If there is any reasonable way to avoid it, we should.