Okay, so p2p could have licensed the tech. I guess that point isn't valid.
And I agree, the case about two songs sounding the same is unlikely to pop-up very often.
I just didn't want to say the obvious, this would reduce p2p into an exclusive list of things which only copyright owners are allowed to opt-in data for the network. Such a network seems insane to me because it assumes everyone is guilty and requires proving innocence. I can only begin to imagine the mess that would happen with this p2p being used for things beyond music, though. P2p with source code would be awful messy (just imagine Linux being distributed with the whole SCO mess). To me, any sort of legislature require this sort of setup seems paramount to a blocking of freedom of speech (yes, kazaa isn't used necessarily a lot for it, but freenet is), as it's an unreasonable burden to register everything you say to the government before being able to say it aloud.
The idea of "Race to the Bottom" seems to ignore that you still have workers that are fired. Comparative advantage just means we have to come up with new, innovative things we can do better. Maybe it's nothing in particular that we're good at directly, but include any overhead (timezone issues, culture barriers, shipping, etc) and there's going to be jobs here that you just can't export overseas. Multinational companies can extend work across the world. In the end, the race to the bottom only works if we accept crap from employers. They'll still hire people here because even if 5 billion people take crap jobs, the other 1 billion still make them money. I'd be most worried about legislation to prevent striking against crap jobs here or abroad--people in 3rd world countries have to demand to have 1st world conditions before they'll ever get it.
> The p2p companies have been claiming that it's impossible to implement this filtering. The point of the demo isn't to have p2p companies implement filtering, it's to establish (legally) that the p2p companies could implement the filtering and choose not to.
Yes and no. It's questionable whether this technology even works, but given that the company is likely to patent the technology, it's pretty credible to say that up until the point they start licensing out the technology the p2p networks can't pull off the nature of filtering necessary.
However, even *with* the filtering, there's no way for this to work properly. Why? Because the filter still can't determine copyright. The simplest example is two different bands playing a public domain work. Get enough bands to play a public domain work, and you're going to end up having a hash conflict. This ignores the whole point that two people could manage to produce the same song (not public domain) and copyright it, which would play pure hell on such a filter system.
> But in all seriousness, instead of location specific ones, wouldn't you rather have a personal ghost? You decide it's appearance on your PDA/Wearable Computer/Whatever, you adjust its personality via programming or learning capabilities.
Damn, straight. Except, they're called Personcons! And mine will say nothing but Chii.
> Michaelangelo got paid for painting the sistine chapel. Bach, Beethoven, Mozard, Handel, etc pretty much got paid for anything they ever did.
Michaelangelo was a software consultant. Bach, Beethoven, Mozart, Handel were sponsored artists. Their music was Free (no copyright), but them actually playing cost money (their time is a finite resource, music can be copied near infinitely). You can't really perform software, but those who made the software or study it well enough can help others to make their computers perform the software for a price, since their time is limited.
It's not a crazy idea even that people would write software and be hired because of what they wrote so a company has a jumpstart in support. And then the author can indirectly be paid for their efforts. Since all the above doesn't cover niche markets where a wide enough audience doesn't exist to provide for this hiring, I don't see it unreasonable that some software will remain proprietarily. The humorous part is, the market is deciding that open source is a viable solution for commodity software.
> All of those companies are bankrolling OSS/free software from their existing mountains of cash with the hopes that by offering it at a loss they can put some hurt on the Microsoft juggernaut, and I would wager that each of them is hemmoraging cash from the business unit in the process.
Yea, RedHat's whole purpose is to bleed Microsoft of money. That's one of the most paranoid things I've ever heard. RedHat is in a business, and Linux servers are selling while desktops just aren't well enough. It's smart business to focus on servers when that's the main market of Linux now.
> The long term view is that eventually by reclaiming the desktop they will be able to provide services, support, and administrative tools that will be profitable, but in the past 3 years and for the immediate future RedHat is spelling it out loud and clear : OSS on the desktop is not a profit driven business venture.
And of course everyone in the OS business is interested in taking monopoly control over desktops. Bleeding cash is not only not necessary for RedHat (since their development tweaking of Linux is cheaper than creating projects, so their total cost of production is cheaper than Microsoft), it'd be stupid considering RedHat doesn't have $40+ billion in the bank to wait out Microsoft. However, if they see a desktop market available, they'll take it. After all, there's 10x to 100x as many desktop users as server users.
>>something not intended to run in safe mode all the time.
>What does "safe mode" have to do with XFree86?
Nothing really, but having XFree86 give you an option to run the Xserver as non-root (something I really would like to see) and giving an option to use vga/vesa if there is a problem wouldn't be bad ideas.
> The point is, "What constitutes 'free as in freedom'?" Is it only the GPL?
Anything that fulfills the idea of the GPL, to prevent the proprietizing of a work, is free as in freedom. BSD is open, not free. It's a sad fact that Microsoft says they're against open source software. I guess it'd look bad if they said they were against freedom source software.
Re:So they stick to the new license...
on
XFree86 4.4 Released
·
· Score: 2, Interesting
> Does the GPL mean whatever the FSF decides it means that week? Or, in a country governed by the rule of law, does it mean no more and no less that what it be proven to mean in court?
Fine, the FSF states that the GPL and such and such license are incompatible. Maybe you don't agree, and you're certainly allowed to go off and treat them as compatible. And if someone gets sued over it because of what you did, you'll be held responsible if they're proven to be incompatible along with anyone else who knew you included such and such licensed code with the GPL code. I'd personally rather side on caution, especially when what the FSF is saying doesn't sound like FUD. I don't need a court telling me using a stapler to commit murder is homicide.
> Would you be willing to hire a convicted Cyber Terrorist?
If every Tom, Dick, and Harry would pulls a prank which "endangers public safety" is labelled a Cyber Terrorist, it's going to be damn hard to find out who the real terrorists are. This is a case of the government crying wolf. At some point, people will ignore the government's claim and will end up hiring a real wolf.
> The "obscene" material was on the original hypthetical link, so I don't see how that would fall under this law.
That's my point, though. The link is still misleading (it's not Jackson's website), and it contains obscene material. It'd seem the law is worded in such a way that you could have people arrested for posting obscene material on a non-official url.
The only way for a url to not be misleading really is to, for example, have "goatse.cx" containing goat sex (an interesting point brought up by another poster). I wonder if phishing urls pointing to goatse.cx are now illegal. Or, really, any proxying url.
Say Janet Jackson decides to put up "obscene" pictures of herself on her website. Now, if someone else were to put up a "misleading" (ie, different way of writing Janet Jackson) and post the same pictures, would they be in violation of the law? I can see them at worst being in strongly violation of copyright law. Would having one porn site's mistyped address go to another porn site qualify?
I think it's more than that. The only secure thing to do in a lot of the vulnerabilities is to pull the plug on a box.
Lets assume for a moment that Microsoft decided to change its policy to encourage releasing security information whenever it's found. If a company has at least two sets of OSs running at a time setup such that if one group goes down the other takes over, then nearly every Windows vulnerability would mean a significant disparity in uptime. With a viable free (cost) OS like Linux or FreeBSD around, it wouldn't be at all surprising that sane companies would begin to have such a setup.
My point is, Microsoft's Security Chief is trying to make corporate consumers feel safe that all they need to do is patch, and they'll not need to worry about it. Corporations who have wised up to the utter breakdowns of have a monoculture might already be realizing this issue. So, it'd seem to follow Microsoft doesn't like the idea of halving sales (or worse, since Linux and FreeBSD would together provide a dual-culture).
> Actually GCC is a good example of why Java shouldn't go this route. Look at all the binary compatibility issues from the 2.96 compilers to the latest and greatest. Caused us all kinds of problems.
Do you expect a JVM v1.1 to run v1.4 programs? Forward and backward compatibility often get broken as a result of fixing design flaws. I'm happy they're fixing gcc now if it means it's a lot less likely to break in the future (not that it's actually, guaranteed..). Besides, 2.96 should never have been released as production since it was a pre3.0 beta. Now, complaining about the fact that there's a different abi between 3.0 and 3.2 is a valid complaint.
> Actually GCC is a good example of why Java shouldn't go this route. Look at all the binary compatibility issues from the 2.96 compilers to the latest and greatest. Caused us all kinds of problems.
Do you expect a JVM v1.1 to run v1.4 programs? Forward and backward compatibility often get broken as a result of fixing design flaws. I'm happy they're fixing gcc now if it means it's a lot less likely to break in the future (not that it's actually, guaranteed..). Besides, 2.96 should never have been released as production since it was a pre3.0 beta. Now, complaining about the fact that there's a different abi between 3.0 and 3.2 is a valid complaint.
Virtual PC, Bochs, and Qemu don't exist? If you mean implementing hardware, that's true. But, that's an issue with all hardware. Hardware costs materials. Software is prepaid HD space that can be reclaimed (plus whatever is charged for the software, backups, etc).
As an IU Alumni, I support the position of suing the RIAA and MPAA for emotional pain and suffering. Other than that, I don't think anything sort of just blocking users one at a time will work.
> Sun isn't going to Open source Java, it wont happen for the reason you mention. Incompatibility would ruin the cross-platform appeal of Java.
Like how gcc being open source causes incompatibilities with different gcc ports? Or how the x86 arch being open ruins the ability to write for it?
I really don't see how any of these arguments hold water except if you believe that somehow one company with proprietary control over a standard will do a better job than the open source community at writing to the spec.
The fact is, the Java spec is free as far as I know, and there's nothing really preventing someone from writing their own implementation and porting it to several platforms. Of course, they can't call it Java (Java is trademarked, as use of a programming language), but they can just as easily do everything but call it Java and have or not have compatibility problems. The only thing that Sun open sourcing Java would do is cut out a lot of the work developing the libraries and such which make up the bulk of Java's classes. Sun still owns the trademark and could write a license to prevent anyone using the term Java except Sun (they could even come up with a specific trademarks to clarify degrees of Java-ness which are allowed to be used by people). Open source doesn't mean Free software. And maybe people aren't happy with that. I personally don't care, since Free software wouldn't ever call it Java (maybe they'd call it Java-like), and I firmly believe that the core useful parts of Java will be written by the open source community eventually which will negate a need for Sun to do anything (well, except keep the specs open).
I just find it funny that the defense of keeping Java closed source is it's broken now, so having more people work on it will somehow make it worse. A company is like a bunch of cooks making one cake (or maybe two or three). The community at large is like a bunch of cooks make a bunch of cakes. If enough of the cooks get together, they can make a larger, better cake just like a company. But, there's a lot more cooks out there than there are in one company. And if what the community writes is utter crap, then no one uses it, and we'll stick to the old recipes. It's not like the open source community is a monopoly that can force you to use an inferior product. Only closed source setups work that way, thanks to things like forced bound upgrades from a single company.
No, don't quit. Do what any other good capitalist would do in a situation where you believe there's going to be a lot of loss do to disefficiencies, try to bleed money out of the company. No need to quit when you can just drag it out for months or years. Maybe they'll realize what a huge mistake a sudden switch is.
PS: I haven't a clue what the company all does, so maybe several of the desktop machines could be converted over without much retraining. I doubt there's any reason to switch though unless there's problems (or if this were Windows, an upgrade cycle). Any sort of switchover is bound to have problems and breed resentment with "loyalists". And maybe in the end, the company will be better off with Windows (anything's possible).
> IIRC, Celerons were basically Pentiums without the cache
The first Celerons were Pentium IIs with half the cache but at full speed (which ironically ended up making the Celerons faster than the Pentium IIs). Then Pentium IIs cache went full speed, so the Celerons became slower again.
> Thanks to the convolutedness of the x86 architecture, if you're running the CPU in 32-bit protected mode
The program is 16-bit DOS.
The bswap idea sounds good, but I'd rather if the program worked on all i386+ chips.
> If a C compiler was written that knew of the trick you mentioned, it would have to be able to tell which integers are 32-bits and which ones are 16-bits
No, that's not really necessary. Shorts are 16-bit integers. The idea really could be extended on Athlon 64's to make all the 64-bit registers behave like twice as many 32-bit registers, as I believe ints are still 32-bit on the Athlon 64. In any case, it'd be a nice hack to avoid the stack a little bit. I'm not sure how much of an improvement it'd really be, though.
>An example of this is some code that draws a scaled translucent sprite. Throughout the code, the scale will remain constant, and if the translucency is uniform, that will remain constant too. The code that does the translucent blitting will use the registers only for values that change during the sprite-drawing.
I did something like that, actually, though it was stretching a sprite.
>On an 80386, using this technique will cause a significant speed-increase in the code, but on 80486's and above where there are on-board L1 caches on the CPUs, the code-modification may cause cache-misses that may slow down the system - espcecially if it is run on an even newer x86 CPU that has a seperate program and data cache in the L1 cache.
Actually, I had a 433Mhz Celeron, and I believe it was faster to do the overwrite (I only changed the immediates on O(1)). Eventually, I switched away from doing this to use 32-bit registered to behave like 2 16-bit registered to avoid doing lots of pushes. Just had to do rol 16 at appropriate times. I'm sort of surprised there isn't a compiler optimization to do this, as the results were nearly as fast as self-modifying code. Here's the program, if at all interested: spcx.
> Merely making stack pages non-executable doesn't prevent return-into-libc exploits for example where you use the global offset table to jump into arbitrary code by overwriting the entry for a library call like printf(3).
I'd be interested to know how a return-into-libc exploit works. The only thought that comes to mind as an exploit is putting args onto the stack and calling some function to do bad things (like remove memory protection) then continue the exploit into the stack space. Of course locking the stack one way or another and disallowing changing permissions would block that attack. So, how's return-into-libc exploit work?
> Also, I suspect AMD possibly suffers from the poor reputations of previous Intel competitors who truly did have unreliable, inferior products. I for one had trouble for a while remembering which of AMD and Cyrix was the one to avoid, thus for the average consumer choosing the always reliable Intel makes some sense.
*Also, I suspect Windows possibly suffers from the poor reputations of previous OS/2 competitors who truly did have unreliable, inferior products. I for one had trouble for a while remembering which of Windows 95 and Windows NT was the one to avoid, thus for the average consumer choosing the always reliable OS/2 makes more sense.*
And Windows 95 and Windows NT sound a lot more alike...
The truth is, Intel is better at marketing the whole '"Pentium" and "Intel Inside" == a compatible computer' for the non-techy who thinks he is a techy. Most people who used AMD or Cyrix computers still didn't notice anything but the price. Intel still has good OEM connections (look at Dell, for example). Most people still don't notice the difference (and starting with the K7 there's really no noticeable difference). I don't think marketing is really the way to go. It's getting better OEM tie-ins. How else do you think the WinTel grew so well? OEMs make two monopolies (admittedly, Intel is less of one now) look like they give you choice.
Slashdot Mirror
Okay, so p2p could have licensed the tech. I guess that point isn't valid.
And I agree, the case about two songs sounding the same is unlikely to pop-up very often.
I just didn't want to say the obvious, this would reduce p2p into an exclusive list of things which only copyright owners are allowed to opt-in data for the network. Such a network seems insane to me because it assumes everyone is guilty and requires proving innocence. I can only begin to imagine the mess that would happen with this p2p being used for things beyond music, though. P2p with source code would be awful messy (just imagine Linux being distributed with the whole SCO mess). To me, any sort of legislature require this sort of setup seems paramount to a blocking of freedom of speech (yes, kazaa isn't used necessarily a lot for it, but freenet is), as it's an unreasonable burden to register everything you say to the government before being able to say it aloud.
The idea of "Race to the Bottom" seems to ignore that you still have workers that are fired. Comparative advantage just means we have to come up with new, innovative things we can do better. Maybe it's nothing in particular that we're good at directly, but include any overhead (timezone issues, culture barriers, shipping, etc) and there's going to be jobs here that you just can't export overseas. Multinational companies can extend work across the world. In the end, the race to the bottom only works if we accept crap from employers. They'll still hire people here because even if 5 billion people take crap jobs, the other 1 billion still make them money. I'd be most worried about legislation to prevent striking against crap jobs here or abroad--people in 3rd world countries have to demand to have 1st world conditions before they'll ever get it.
> The p2p companies have been claiming that it's impossible to implement this filtering. The point of the demo isn't to have p2p companies implement filtering, it's to establish (legally) that the p2p companies could implement the filtering and choose not to.
Yes and no. It's questionable whether this technology even works, but given that the company is likely to patent the technology, it's pretty credible to say that up until the point they start licensing out the technology the p2p networks can't pull off the nature of filtering necessary.
However, even *with* the filtering, there's no way for this to work properly. Why? Because the filter still can't determine copyright. The simplest example is two different bands playing a public domain work. Get enough bands to play a public domain work, and you're going to end up having a hash conflict. This ignores the whole point that two people could manage to produce the same song (not public domain) and copyright it, which would play pure hell on such a filter system.
> But in all seriousness, instead of location specific ones, wouldn't you rather have a personal ghost? You decide it's appearance on your PDA/Wearable Computer/Whatever, you adjust its personality via programming or learning capabilities.
Damn, straight. Except, they're called Personcons! And mine will say nothing but Chii.
> Michaelangelo got paid for painting the sistine chapel. Bach, Beethoven, Mozard, Handel, etc pretty much got paid for anything they ever did.
Michaelangelo was a software consultant. Bach, Beethoven, Mozart, Handel were sponsored artists. Their music was Free (no copyright), but them actually playing cost money (their time is a finite resource, music can be copied near infinitely). You can't really perform software, but those who made the software or study it well enough can help others to make their computers perform the software for a price, since their time is limited.
It's not a crazy idea even that people would write software and be hired because of what they wrote so a company has a jumpstart in support. And then the author can indirectly be paid for their efforts. Since all the above doesn't cover niche markets where a wide enough audience doesn't exist to provide for this hiring, I don't see it unreasonable that some software will remain proprietarily. The humorous part is, the market is deciding that open source is a viable solution for commodity software.
> You're giving away half of what you have to offer. They'll find someone who will be cheaper to do the other half.
:)
They'll find someone who will pay to do the other half?
> All of those companies are bankrolling OSS/free software from their existing mountains of cash with the hopes that by offering it at a loss they can put some hurt on the Microsoft juggernaut, and I would wager that each of them is hemmoraging cash from the business unit in the process.
Yea, RedHat's whole purpose is to bleed Microsoft of money. That's one of the most paranoid things I've ever heard. RedHat is in a business, and Linux servers are selling while desktops just aren't well enough. It's smart business to focus on servers when that's the main market of Linux now.
> The long term view is that eventually by reclaiming the desktop they will be able to provide services, support, and administrative tools that will be profitable, but in the past 3 years and for the immediate future RedHat is spelling it out loud and clear : OSS on the desktop is not a profit driven business venture.
And of course everyone in the OS business is interested in taking monopoly control over desktops. Bleeding cash is not only not necessary for RedHat (since their development tweaking of Linux is cheaper than creating projects, so their total cost of production is cheaper than Microsoft), it'd be stupid considering RedHat doesn't have $40+ billion in the bank to wait out Microsoft. However, if they see a desktop market available, they'll take it. After all, there's 10x to 100x as many desktop users as server users.
>>something not intended to run in safe mode all the time.
>What does "safe mode" have to do with XFree86?
Nothing really, but having XFree86 give you an option to run the Xserver as non-root (something I really would like to see) and giving an option to use vga/vesa if there is a problem wouldn't be bad ideas.
> The point is, "What constitutes 'free as in freedom'?" Is it only the GPL?
Anything that fulfills the idea of the GPL, to prevent the proprietizing of a work, is free as in freedom. BSD is open, not free. It's a sad fact that Microsoft says they're against open source software. I guess it'd look bad if they said they were against freedom source software.
> Does the GPL mean whatever the FSF decides it means that week? Or, in a country governed by the rule of law, does it mean no more and no less that what it be proven to mean in court?
Fine, the FSF states that the GPL and such and such license are incompatible. Maybe you don't agree, and you're certainly allowed to go off and treat them as compatible. And if someone gets sued over it because of what you did, you'll be held responsible if they're proven to be incompatible along with anyone else who knew you included such and such licensed code with the GPL code. I'd personally rather side on caution, especially when what the FSF is saying doesn't sound like FUD. I don't need a court telling me using a stapler to commit murder is homicide.
> Would you be willing to hire a convicted Cyber Terrorist?
If every Tom, Dick, and Harry would pulls a prank which "endangers public safety" is labelled a Cyber Terrorist, it's going to be damn hard to find out who the real terrorists are. This is a case of the government crying wolf. At some point, people will ignore the government's claim and will end up hiring a real wolf.
> The "obscene" material was on the original hypthetical link, so I don't see how that would fall under this law.
That's my point, though. The link is still misleading (it's not Jackson's website), and it contains obscene material. It'd seem the law is worded in such a way that you could have people arrested for posting obscene material on a non-official url.
The only way for a url to not be misleading really is to, for example, have "goatse.cx" containing goat sex (an interesting point brought up by another poster). I wonder if phishing urls pointing to goatse.cx are now illegal. Or, really, any proxying url.
Say Janet Jackson decides to put up "obscene" pictures of herself on her website. Now, if someone else were to put up a "misleading" (ie, different way of writing Janet Jackson) and post the same pictures, would they be in violation of the law? I can see them at worst being in strongly violation of copyright law. Would having one porn site's mistyped address go to another porn site qualify?
I think it's more than that. The only secure thing to do in a lot of the vulnerabilities is to pull the plug on a box.
Lets assume for a moment that Microsoft decided to change its policy to encourage releasing security information whenever it's found. If a company has at least two sets of OSs running at a time setup such that if one group goes down the other takes over, then nearly every Windows vulnerability would mean a significant disparity in uptime. With a viable free (cost) OS like Linux or FreeBSD around, it wouldn't be at all surprising that sane companies would begin to have such a setup.
My point is, Microsoft's Security Chief is trying to make corporate consumers feel safe that all they need to do is patch, and they'll not need to worry about it. Corporations who have wised up to the utter breakdowns of have a monoculture might already be realizing this issue. So, it'd seem to follow Microsoft doesn't like the idea of halving sales (or worse, since Linux and FreeBSD would together provide a dual-culture).
> Actually GCC is a good example of why Java shouldn't go this route. Look at all the binary compatibility issues from the 2.96 compilers to the latest and greatest. Caused us all kinds of problems.
Do you expect a JVM v1.1 to run v1.4 programs? Forward and backward compatibility often get broken as a result of fixing design flaws. I'm happy they're fixing gcc now if it means it's a lot less likely to break in the future (not that it's actually, guaranteed..). Besides, 2.96 should never have been released as production since it was a pre3.0 beta. Now, complaining about the fact that there's a different abi between 3.0 and 3.2 is a valid complaint.
> Actually GCC is a good example of why Java shouldn't go this route. Look at all the binary compatibility issues from the 2.96 compilers to the latest and greatest. Caused us all kinds of problems. Do you expect a JVM v1.1 to run v1.4 programs? Forward and backward compatibility often get broken as a result of fixing design flaws. I'm happy they're fixing gcc now if it means it's a lot less likely to break in the future (not that it's actually, guaranteed..). Besides, 2.96 should never have been released as production since it was a pre3.0 beta. Now, complaining about the fact that there's a different abi between 3.0 and 3.2 is a valid complaint.
Virtual PC, Bochs, and Qemu don't exist? If you mean implementing hardware, that's true. But, that's an issue with all hardware. Hardware costs materials. Software is prepaid HD space that can be reclaimed (plus whatever is charged for the software, backups, etc).
As an IU Alumni, I support the position of suing the RIAA and MPAA for emotional pain and suffering. Other than that, I don't think anything sort of just blocking users one at a time will work.
> Sun isn't going to Open source Java, it wont happen for the reason you mention. Incompatibility would ruin the cross-platform appeal of Java.
Like how gcc being open source causes incompatibilities with different gcc ports? Or how the x86 arch being open ruins the ability to write for it?
I really don't see how any of these arguments hold water except if you believe that somehow one company with proprietary control over a standard will do a better job than the open source community at writing to the spec.
The fact is, the Java spec is free as far as I know, and there's nothing really preventing someone from writing their own implementation and porting it to several platforms. Of course, they can't call it Java (Java is trademarked, as use of a programming language), but they can just as easily do everything but call it Java and have or not have compatibility problems. The only thing that Sun open sourcing Java would do is cut out a lot of the work developing the libraries and such which make up the bulk of Java's classes. Sun still owns the trademark and could write a license to prevent anyone using the term Java except Sun (they could even come up with a specific trademarks to clarify degrees of Java-ness which are allowed to be used by people). Open source doesn't mean Free software. And maybe people aren't happy with that. I personally don't care, since Free software wouldn't ever call it Java (maybe they'd call it Java-like), and I firmly believe that the core useful parts of Java will be written by the open source community eventually which will negate a need for Sun to do anything (well, except keep the specs open).
I just find it funny that the defense of keeping Java closed source is it's broken now, so having more people work on it will somehow make it worse. A company is like a bunch of cooks making one cake (or maybe two or three). The community at large is like a bunch of cooks make a bunch of cakes. If enough of the cooks get together, they can make a larger, better cake just like a company. But, there's a lot more cooks out there than there are in one company. And if what the community writes is utter crap, then no one uses it, and we'll stick to the old recipes. It's not like the open source community is a monopoly that can force you to use an inferior product. Only closed source setups work that way, thanks to things like forced bound upgrades from a single company.
No, don't quit. Do what any other good capitalist would do in a situation where you believe there's going to be a lot of loss do to disefficiencies, try to bleed money out of the company. No need to quit when you can just drag it out for months or years. Maybe they'll realize what a huge mistake a sudden switch is.
PS: I haven't a clue what the company all does, so maybe several of the desktop machines could be converted over without much retraining. I doubt there's any reason to switch though unless there's problems (or if this were Windows, an upgrade cycle). Any sort of switchover is bound to have problems and breed resentment with "loyalists". And maybe in the end, the company will be better off with Windows (anything's possible).
I stand corrected. I forget, the first Celeron (266) had no L2 cache. They overcompensated with the 300A. Well, anyways... :)
> IIRC, Celerons were basically Pentiums without the cache
The first Celerons were Pentium IIs with half the cache but at full speed (which ironically ended up making the Celerons faster than the Pentium IIs). Then Pentium IIs cache went full speed, so the Celerons became slower again.
> Thanks to the convolutedness of the x86 architecture, if you're running the CPU in 32-bit protected mode
The program is 16-bit DOS.
The bswap idea sounds good, but I'd rather if the program worked on all i386+ chips.
> If a C compiler was written that knew of the trick you mentioned, it would have to be able to tell which integers are 32-bits and which ones are 16-bits
No, that's not really necessary. Shorts are 16-bit integers. The idea really could be extended on Athlon 64's to make all the 64-bit registers behave like twice as many 32-bit registers, as I believe ints are still 32-bit on the Athlon 64. In any case, it'd be a nice hack to avoid the stack a little bit. I'm not sure how much of an improvement it'd really be, though.
>An example of this is some code that draws a scaled translucent sprite. Throughout the code, the scale will remain constant, and if the translucency is uniform, that will remain constant too. The code that does the translucent blitting will use the registers only for values that change during the sprite-drawing.
I did something like that, actually, though it was stretching a sprite.
>On an 80386, using this technique will cause a significant speed-increase in the code, but on 80486's and above where there are on-board L1 caches on the CPUs, the code-modification may cause cache-misses that may slow down the system - espcecially if it is run on an even newer x86 CPU that has a seperate program and data cache in the L1 cache.
Actually, I had a 433Mhz Celeron, and I believe it was faster to do the overwrite (I only changed the immediates on O(1)). Eventually, I switched away from doing this to use 32-bit registered to behave like 2 16-bit registered to avoid doing lots of pushes. Just had to do rol 16 at appropriate times. I'm sort of surprised there isn't a compiler optimization to do this, as the results were nearly as fast as self-modifying code. Here's the program, if at all interested: spcx.
> Merely making stack pages non-executable doesn't prevent return-into-libc exploits for example where you use the global offset table to jump into arbitrary code by overwriting the entry for a library call like printf(3).
I'd be interested to know how a return-into-libc exploit works. The only thought that comes to mind as an exploit is putting args onto the stack and calling some function to do bad things (like remove memory protection) then continue the exploit into the stack space. Of course locking the stack one way or another and disallowing changing permissions would block that attack. So, how's return-into-libc exploit work?
> Also, I suspect AMD possibly suffers from the poor reputations of previous Intel competitors who truly did have unreliable, inferior products. I for one had trouble for a while remembering which of AMD and Cyrix was the one to avoid, thus for the average consumer choosing the always reliable Intel makes some sense.
*Also, I suspect Windows possibly suffers from the poor reputations of previous OS/2 competitors who truly did have unreliable, inferior products. I for one had trouble for a while remembering which of Windows 95 and Windows NT was the one to avoid, thus for the average consumer choosing the always reliable OS/2 makes more sense.*
And Windows 95 and Windows NT sound a lot more alike...
The truth is, Intel is better at marketing the whole '"Pentium" and "Intel Inside" == a compatible computer' for the non-techy who thinks he is a techy. Most people who used AMD or Cyrix computers still didn't notice anything but the price. Intel still has good OEM connections (look at Dell, for example). Most people still don't notice the difference (and starting with the K7 there's really no noticeable difference). I don't think marketing is really the way to go. It's getting better OEM tie-ins. How else do you think the WinTel grew so well? OEMs make two monopolies (admittedly, Intel is less of one now) look like they give you choice.