The inline flag in C and C++ is a hint to the compiler that inlining this function is a good idea, just like register is a hint to the compiler.
GCC has always treated inline as such a hint, but the heuristics of how to use the hint has changed, so some functions that used to be inlined no longer is inlined.
The kernel has some function that *must* be inlined, not for speed but for correctness. GCC provide a difference way to specify this, a "inline this function or die" flag. Development kernels use this flag.
A bug in a deprecated GCC extension
on
GCC 3.3 Released
·
· Score: 2, Interesting
Treating casts as lvalues is a GCC extension, and an extension that has been deprecated for C++ since 3.0 because it causes problems for valid C++ code.
I believe the plan is to add a warning in 3.4 and remove it in 3.5.
Software projects tend to be released too early (before they are ready) or too late (later than initially expected).
It has nothing to do with whether the software is free or not, expect perhaps a preference of some free software managers towards "too late" over "too early".
I suspect this is because programming is still a creative craft, which makes it harder to predict.
The Danish constitution also protects free speach, but here it only means that the government cannot stop you from publishing your opinions. There are no constitunional guarentee that they will not put you in jail afterwards. And in fact they will, if your opinions are racist or against groups of specific ethnic, religious or sexual orientations.
The GPL is based on copyright. This means that you can read, use and study code covered by it, and use and techniques and knowledge you gained by doing so in your own work. The ideas behind a copyrighted work is yours to play with, no strings attached. Only the specific expression of the ideas are controlled by copyright. You cannot copy that expression without permission, and the GPL is a conditional permission.
An NDA is a contract that you sign where you promisse to keep information secrets, i.e. not just the particular expression as in copyright, but the ideas themselves. This is far more restritive than any copyright based license can ever be.
You are not allowed to take the GNU GPL, edit the sections you dislike, call the result MV GPL, and publish *your own* software under that new license.
The *text* of the GNU GPL is under a very restrictive license, at least as restrictive as the invariant sections of the GFDL.
The point is that if you really insist on being holier than the pope on the issue of free information, you cannot distribute any software under the GNU GPL. While the software covered by the GNU GPL is free accoding to the Debian free software guidelines, the text of the license itself is not free, and the license requires that it itself is distributed together with the software as an "invariant section".
What is strange is not German pascifism, but the American surprise by German pascifism.
A Danish journalist in the US explained it like this: "If you go into an American bookstore, you will find rows of books about the second world war, but only a few volumes about Germany post-WWII." I.e. American perception of Germany is still totally dominated by WWII.
Germany and Japan may be the two countries who suffered most under World War II. At the same time, it may also be the two countries that peace brought the greatest prosperity.
I think it is a small wonder that the population of these countries are more heeart-felt anti-war than UK or USA who has not suffered the trauma of occupation, nor had their cities anihilated by enemy bombs.
...for writing the series of Foundation short stories, and presumably he was satisfied with the rate since he volunteered to do the work. After (and while) writing the series, he continued to work for higher and higher rate, as his talent was obvious.
I don't see how it has promoted the state of science and art that he unexpectedly many years later got paid a second time for work already done when the short stories were collected and printed as a series.
He was also paid a third time by doing later sequels (of qustionable quality) in the same series, which I presume sold wll based on the popularity of the original stories. That is more reasonable, as this was new payment for new work.
If you look for something in the documentation and doesn't find it, or does not understand the explanation, a bug report about the documentation is very helpful. Assuming, of course, the project has any documentation to speak of.
Always state what the problem is and where you looked to find answers. Be specific, i.e. if you looked in the index, tell what words you looked for. If you did not understand the explanation, tell what paragraphs you found unclear.
If you actually somehow found an answer (possible in a support forum), try to suggest a formulation that you would understand.
As always with free software, try to be part of the solution. That is the only way free software will work. If you can't work like that, find a commercial alternative.
Except for some very specialized applications, GCC generated code is within 15% of ICC generated code on IA-32. My application is actually 5% faster when compiled with GCC.
I'd be very surprised if the same is true for IA-64.
The basic moral principle for RMS behind the GPL seems to be that people who use software should also be able to modify it (and share their modifications).
This would seem to imply that web-services should provide access to the source code.
The right thing would be to drop manned space exploration for the forseeable future, and put all the money on sondes and robotic crafts. This will give far more science for the money, and the spin-off techonolgy is also likely to be more valuable.
I know there is a lot of semtiments involved with manned space exploration, but I believe we need to put these aside, and consentrate on the science.
When the technology is cheap enough, or rich people are rich enough, a self-sufficient space turist industry may put back men in space, but until then, let the robots take over.
> Why do we keep assuming that opensource resources just transfer between products?
Because, to a large degree, it is true. And even to the degree it *isn't* true, we should try to encourage would-be developers to join an existing project, rather than start their own. There is no lack of free software projects, however there is a desperate lack if free software projects with enough developers to produce anything worthwhile.
Of course, hobbyist programmers should do whatever they think is fun, even if it never produce anything useful to others. But that should not stop us oldbies encouraging them to join existing larger projects that may already have produced, or is likely to end up producing, something worthwhile. There is also a great fun seeing your code getting used. And while working with others can be annoying, it can also be rewarding, and it is a valuable skill to learn.
Note that far the most contribution to the Linux kernel come from developers who are *not* working for free, but for various companies such as Red Hat, SGI, SUSE, etc.
In some ways, Linux and the GPL provide the Unix companies with what Unix International and OSF failed to crate, a platform where the various Unix vendors all share improvements.
Of course, some companies share more than others. IBM is one of those companies that contribute the most
It would establish QuickTime as the premier portable media format.
While Linux isn't a force on the desktop, it is pretty important among the technicians in the server rooms, and these guys often get to make technical decisions such as "which media format to use". As long as it is convenient to play on MS Windows, their bosses don't really care. Unless they have just been visited by a MS sales representive, they will leave the decision to the technicians.
Of course people are bitching that they can't get others to fix their problems for free. As a programmer I find it hard to sympatise, I they expect me to fix their problems, they better expect to pay me for the priviledge. I don't expect them to work for free for me either.
I *do* distribute some software for free though, but that is just because it was fun to write, it solved problems for me, or (for most of it) other people paid me to write it, and there wasn't any reason not to distribute it for free.
If someone think I should write more free software, they better make sure one of these three motivations apply. Bitching isn't helping there.
Very little free software is written for political reasons, so trying to say how the world "should" be is useless.
If you want new or better software there are two useful options: Write it yourself, or pay someone to write it for you. Bitching wont change a thing. Nobody is obliged to program for you for free, if you think so you have misunderstood what the free in free software means.
Basically, you can only post on advogato if some trusted users claim you are ok. It is all backed by a complex web of trust algorithm which makes it hard to attack.
This makes Advogato an "invitation only" club.
Obviously, this only works when the people you want to use the forum know each others, at least by reputation.
Slashdot Mirror
The inline flag in C and C++ is a hint to the compiler that inlining this function is a good idea, just like register is a hint to the compiler.
GCC has always treated inline as such a hint, but the heuristics of how to use the hint has changed, so some functions that used to be inlined no longer is inlined.
The kernel has some function that *must* be inlined, not for speed but for correctness. GCC provide a difference way to specify this, a "inline this function or die" flag. Development kernels use this flag.
Treating casts as lvalues is a GCC extension, and an extension that has been deprecated for C++ since 3.0 because it causes problems for valid C++ code.
I believe the plan is to add a warning in 3.4 and remove it in 3.5.
Software projects tend to be released too early (before they are ready) or too late (later than initially expected).
It has nothing to do with whether the software is free or not, expect perhaps a preference of some free software managers towards "too late" over "too early".
I suspect this is because programming is still a creative craft, which makes it harder to predict.
The Danish constitution also protects free speach, but here it only means that the government cannot stop you from publishing your opinions. There are no constitunional guarentee that they will not put you in jail afterwards. And in fact they will, if your opinions are racist or against groups of specific ethnic, religious or sexual orientations.
Do you have any idea what a NDA is?
The GPL is based on copyright. This means that you can read, use and study code covered by it, and use and techniques and knowledge you gained by doing so in your own work. The ideas behind a copyrighted work is yours to play with, no strings attached. Only the specific expression of the ideas are controlled by copyright. You cannot copy that expression without permission, and the GPL is a conditional permission.
An NDA is a contract that you sign where you promisse to keep information secrets, i.e. not just the particular expression as in copyright, but the ideas themselves. This is far more restritive than any copyright based license can ever be.
You are not allowed to take the GNU GPL, edit the sections you dislike, call the result MV GPL, and publish *your own* software under that new license.
The *text* of the GNU GPL is under a very restrictive license, at least as restrictive as the invariant sections of the GFDL.
The point is that if you really insist on being holier than the pope on the issue of free information, you cannot distribute any software under the GNU GPL. While the software covered by the GNU GPL is free accoding to the Debian free software guidelines, the text of the license itself is not free, and the license requires that it itself is distributed together with the software as an "invariant section".
At least not compared to the history of ordinary computers. The theory of computation predates actual hardware with more than a decade.
What is strange is not German pascifism, but the American surprise by German pascifism.
A Danish journalist in the US explained it like this: "If you go into an American bookstore, you will find rows of books about the second world war, but only a few volumes about Germany post-WWII." I.e. American perception of Germany is still totally dominated by WWII.
Germany and Japan may be the two countries who suffered most under World War II. At the same time, it may also be the two countries that peace brought the greatest prosperity.
I think it is a small wonder that the population of these countries are more heeart-felt anti-war than UK or USA who has not suffered the trauma of occupation, nor had their cities anihilated by enemy bombs.
I'd strongly recommend against using that for anything but historical reference, the languag has changed a lot meanwhile.
However, it was the year C got standardized, so a C book from 1989 is still relevant (C99 isn't widely supported yet).
...for writing the series of Foundation short stories, and presumably he was satisfied with the rate since he volunteered to do the work. After (and while) writing the series, he continued to work for higher and higher rate, as his talent was obvious.
I don't see how it has promoted the state of science and art that he unexpectedly many years later got paid a second time for work already done when the short stories were collected and printed as a series.
He was also paid a third time by doing later sequels (of qustionable quality) in the same series, which I presume sold wll based on the popularity of the original stories. That is more reasonable, as this was new payment for new work.
If you look for something in the documentation and doesn't find it, or does not understand the explanation, a bug report about the documentation is very helpful. Assuming, of course, the project has any documentation to speak of.
Always state what the problem is and where you looked to find answers. Be specific, i.e. if you looked in the index, tell what words you looked for. If you did not understand the explanation, tell what paragraphs you found unclear.
If you actually somehow found an answer (possible in a support forum), try to suggest a formulation that you would understand.
As always with free software, try to be part of the solution. That is the only way free software will work. If you can't work like that, find a commercial alternative.
Nowehere does the article suggest that it would be a good idea for bug reporters to read any form for documentation before reporting a bug.
/. moderators.
I understand that such suggestions is considered unbearable rude by the new generation of software users, but no such suggestion was made.
I, however, would suggest that moderators should read the article before moderating comments that claim to summarize the article as "insightful".
Hopefully such a suggestions are not yet considered unbearable rude by
Only copyright law make it possible for terrorists to earn money on unautorized copying. So if you support copyright law, you support terrorism.
Same logic applies to illegal drugs.
Does it produce code that is any good?
Except for some very specialized applications, GCC generated code is within 15% of ICC generated code on IA-32. My application is actually 5% faster when compiled with GCC.
I'd be very surprised if the same is true for IA-64.
The basic moral principle for RMS behind the GPL seems to be that people who use software should also be able to modify it (and share their modifications).
This would seem to imply that web-services should provide access to the source code.
The right thing would be to drop manned space exploration for the forseeable future, and put all the money on sondes and robotic crafts. This will give far more science for the money, and the spin-off techonolgy is also likely to be more valuable.
I know there is a lot of semtiments involved with manned space exploration, but I believe we need to put these aside, and consentrate on the science.
When the technology is cheap enough, or rich people are rich enough, a self-sufficient space turist industry may put back men in space, but until then, let the robots take over.
> Why do we keep assuming that opensource resources just transfer between products?
Because, to a large degree, it is true. And even to the degree it *isn't* true, we should try to encourage would-be developers to join an existing project, rather than start their own. There is no lack of free software projects, however there is a desperate lack if free software projects with enough developers to produce anything worthwhile.
Of course, hobbyist programmers should do whatever they think is fun, even if it never produce anything useful to others. But that should not stop us oldbies encouraging them to join existing larger projects that may already have produced, or is likely to end up producing, something worthwhile. There is also a great fun seeing your code getting used. And while working with others can be annoying, it can also be rewarding, and it is a valuable skill to learn.
Note that far the most contribution to the Linux kernel come from developers who are *not* working for free, but for various companies such as Red Hat, SGI, SUSE, etc.
In some ways, Linux and the GPL provide the Unix companies with what Unix International and OSF failed to crate, a platform where the various Unix vendors all share improvements.
Of course, some companies share more than others. IBM is one of those companies that contribute the most
It would establish QuickTime as the premier portable media format.
While Linux isn't a force on the desktop, it is pretty important among the technicians in the server rooms, and these guys often get to make technical decisions such as "which media format to use". As long as it is convenient to play on MS Windows, their bosses don't really care. Unless they have just been visited by a MS sales representive, they will leave the decision to the technicians.
Tell your grandma to pay someone for software she can use. It may be Red Hat or Microsoft, whoever currently offer the best value for the money.
Of course people are bitching that they can't get others to fix their problems for free. As a programmer I find it hard to sympatise, I they expect me to fix their problems, they better expect to pay me for the priviledge. I don't expect them to work for free for me either.
I *do* distribute some software for free though, but that is just because it was fun to write, it solved problems for me, or (for most of it) other people paid me to write it, and there wasn't any reason not to distribute it for free.
If someone think I should write more free software, they better make sure one of these three motivations apply. Bitching isn't helping there.
Very little free software is written for political reasons, so trying to say how the world "should" be is useless.
If you want new or better software there are two useful options: Write it yourself, or pay someone to write it for you. Bitching wont change a thing. Nobody is obliged to program for you for free, if you think so you have misunderstood what the free in free software means.
Meaning you could play movies from within an Emacs buffer for the Emacs version JWZ initiated (Lucid GNU Emacs).
I'm pretty sure that feature is broken in the current version (XEmac), and that nobody cares.
Basically, you can only post on advogato if some trusted users claim you are ok. It is all backed by a complex web of trust algorithm which makes it hard to attack.
This makes Advogato an "invitation only" club.
Obviously, this only works when the people you want to use the forum know each others, at least by reputation.