Hehehe, GPL, free software , all of it was about "code sharing" etc etc
No, no and no. GPL and free software are NOT just simply about 'code sharing'. They're about ensuring freedom.
In this case, it's quite unfortunate that the QPL and GPL are incompatible, but keep in mind that this is because the QPL places additional restrictions on what a user is allowed to do. The GPL does not allow one to compromise the users freedom.
Don't forget that it's optional - you can choose to use version 2.0 of the GPL, or at your option, any later verison.
So, if I were a proprietary developer, I'd choose GPL 3, fiddle a bit with the GPL'ed applications to give them a slightly different look, make them incompatible, and off I go selling them (without sources of course).
The GPL gives rights to the USER, not the programmer, in fact thats the whole idea behind it.
Maybe this is also the reason why people like RMS as he is. If he were less radical, or ever likely to change stances, we'd be taking a big risk.
Nearly all GPL'ed software uses a standard preamble which states:
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
So if RMS were to change the GPL license, it WOULD affect most free software that is out there.
Say one day he gets hit by lightning and suddenly starts to think that prorietary software is the way to go, he can change the GPL and affect nearly every GPL'ed work.
The reason for this clause is to allow the FSF to 'plug holes' in the license, and to be able to adjust to future developments (ASP comes to mind).
It would also be possible the fix the KDE/Qt problem this way. But that is not likely to ever happen. A good thing, because the when we choose the GPL for our software, we do it because we value freedom (Why KDE chose it is beyond me). Compromising that freedom would be a bad thing.
This is a good thing, and Red Hat deserves credit for this, but I'd really like to see one day is a more global way of providing legal support for free/open source developers.
Reverse engineering is one side of the story, but there are others too. Pursuing violations of your copyleft for example. As a small developer, whom do you turn to for your legal assistance? I don't think most developers can afford to pay lawyers if they see their licenses violated.
If you use a GNU license, you can theoretically call the FSF for support. Theoretically that is, because things tend to get complicated if you didn't sign over your copyright, which is what most people do.
And if you use another sort of free license, you're totally in your own...
It would be nice to have some kind of organistion, funded by companies like Red Hat, to turn to.
Network Boot ROM's do sit on network cards. Remember that the normal ROM would have to contain network card drivers for every card in existance if it were the other way around.
Probably memory issues...2MB is very little to run an OS and a webserver in. Not having to load the pages and software into that 2MB gives them some more room to play with.
Also, I think the ROM is a bit less power-hungry that your average network card...
Hmmm, is there anything that forces Microsoft to actually answer these questions?
After reading them, I don't think this is the kind of stuff they will want to respond to. For some reason, I don't even think Andover.net expects them to;)
Memory is not an issue to solving chess, as you can use a depth-first algorithm to traverse the tree, so the actual size of the tree doesn't matter for memory usage.
Time isn't either, if we're speaking about solving it. It doesn't matter how long it takes, just that it can be done in finite time. The required time can always be shortened by using more or faster comps.
Neither will solving chess make the game less interesting. There would be no way for a human to make use of the 'solution'.
In short, all what you've said was wrong.
-- GCP
Re:That's a pretty weak conclusion
on
Solving Chess?
·
· Score: 1
This kind of pruning techniques have been used in chessprograms for ages. Some of them, like alpha-beta pruning, are 'perfect', i.e. they never cut off anything they shouldn't have.
But the techniques you describe are forward pruning techniques.
NONE of those is perfect. Some of the more widely used ones (ie nullmove pruning) have well-known flaws. If you want to 'solve chess', you cannot use them.
'Eliminating stupid moves' is something that sounds easy, but in practise, simply isn't possible to do perfectly.
It isn't 'plainly obvious' at all. It just sounds that way.
-- GCP
Re:I don't think this should be determined
on
Solving Chess?
·
· Score: 1
Not at all. Solving chess means that we know whether the first players wins, draws or loses by perfect play.
The 'solution' would be a number, and a series of moves that establishes that result. (there may be several solutions, but only one would be needed)
Knowing the perfect move from every position is a practical problem that is not related to solving the game itself.
Not really. Knuth & Moore have proven the lower bound of the number of positions to be looked at, in order to prove the value (win/loss/draw) of a certain position as:
b^floor(d/2) + b^ceil(d/2) - 1
where b equals branching factor and d equals ply depth
If you know that b = 38 for chess, and the average chess game lasts 40 moves (d=80), you will see that even with a pruning technique that always reaches this minimum, the number is still terribly LARGE.
A better pruning technique would improve pratical performance of chess-playing programs, but would not make the game more solvable.
The problem is that currently you can take a GPL'ed program, say the GIMP, stick your name all over it, remove the original authors credits and put it up as a server-based app.
You could even hack it up a little, so that it would only be fully usable with your proprietary client software, and so that other software would be more cumbersome or unstable for using it.
You're not distributing anything, so nobody can force you to give out sources.
Legally, what you're doing is fine. But I'm pretty sure the original authors won't be too happy with it.
a) Today free software has come to a point where lots of people are using it without understanding or caring about the underlying ideas. More and more commercial software is coming to platforms which previously solely consisted of free software. Projects like KDE flourish based on non-free foundations. Even in the comments to this/. article, many people don't even seem to understand the difference between free and open-source software.
When you try to focus attention to the free software idea, like in the GNU/Linux case, or the Library->Lesser GPL case, you are confronted with lack of understanding.
Considering this, do you think there is a realistic threat that in the near or distant future the free software ideas will be buried and forgotten ?
If so, how are you and the FSF going to stop this, especially considering that previous attempts seem to have had more of a negative than a positive effect ?
b) Sometimes you are called a communist, mostly by people that do not like you. But what *are* your views on the political and economical world ? How much do you think the free software philosophy is extendable outside of the software itself ?
I am not RMS, but I have discussed with him on this subject, so I can probably answer on some of the issues.
The GPL indeed has no power in the case of server software, that provides a service without being distributed.
There were thoughts about addressing this in GPL 3, but the FSF (and RMS) were unsure whether it was doable, the right thing to do, and whether it would be accepted by the community.
For that last issue, don't forget that most, if not all programs which are GPL'ed refer to GPL 2 'or (at your option) any later version'. So if a new version of the GPL comes out, it automatically has effect on all older GPL'ed software.
Publishing a new GPL is not something that will be done without *lots* of thought.
Slashdot Mirror
a) This is (adapted?) from Mark Twain, it's in
...but it keeps reminding me of 'Allo Allo?'
most fortunes.
b) No matter how funny it looks, if you read
it aloud its prefectly understandable...
c)
--
GCP
Nice link. I fell for it.
--
GCP
One word: LGPL
It wont allow you to get any license fees though.
--
GCP
No, no and no. GPL and free software are NOT just simply about 'code sharing'. They're about ensuring freedom.
In this case, it's quite unfortunate that the QPL and GPL are incompatible, but keep in mind that this is because the QPL places additional restrictions on what a user is allowed to do. The GPL does not allow one to compromise the users freedom.
-- GCP
So, if I were a proprietary developer, I'd choose GPL 3, fiddle a bit with the GPL'ed applications to give them a slightly different look, make them incompatible, and off I go selling them (without sources of course).
The GPL gives rights to the USER, not the programmer, in fact thats the whole idea behind it.
Maybe this is also the reason why people like RMS as he is. If he were less radical, or ever likely to change stances, we'd be taking a big risk.
-- GCP
Nearly all GPL'ed software uses a standard preamble which states:
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
So if RMS were to change the GPL license, it WOULD affect most free software that is out there.
Say one day he gets hit by lightning and suddenly starts to think that prorietary software is the way to go, he can change the GPL and affect nearly every GPL'ed work.
The reason for this clause is to allow the FSF to 'plug holes' in the license, and to be able to adjust to future developments (ASP comes to mind).
It would also be possible the fix the KDE/Qt problem this way. But that is not likely to ever happen. A good thing, because the when we choose the GPL for our software, we do it because we value freedom (Why KDE chose it is beyond me). Compromising that freedom would be a bad thing.
-- GCP
There was an attempt to replace Qt by a free clone called Harmony. But as far as I can tell, it's a dead project now.
If someone can demonstrate otherwise, I'd like a link.
--
GCP
This is a good thing, and Red Hat deserves credit for this, but I'd really like to see one day is a more global way of providing legal support for free/open source developers.
Reverse engineering is one side of the story, but there are others too. Pursuing violations of your copyleft for example. As a small developer, whom do you turn to for your legal assistance? I don't think most developers can afford to pay lawyers if they see their licenses violated.
If you use a GNU license, you can theoretically call the FSF for support. Theoretically that is, because things tend to get complicated if you didn't sign over your copyright, which is what most people do.
And if you use another sort of free license, you're totally in your own...
It would be nice to have some kind of organistion, funded by companies like Red Hat, to turn to.
Would it be feasible? I don't know.
--
GCP
Network Boot ROM's do sit on network cards. Remember that the normal ROM would have to contain network card drivers for every card in existance if it were the other way around.
Probably memory issues...2MB is very little to run an OS and a webserver in. Not having to load the pages and software into that 2MB gives them some more room to play with.
Also, I think the ROM is a bit less power-hungry that your average network card...
Hmmm, is there anything that forces Microsoft to actually answer these questions?
;)
After reading them, I don't think this is the kind of stuff they will want to respond to. For some reason, I don't even think Andover.net expects them to
--
GCP
Memory is not an issue to solving chess, as you can use a depth-first algorithm to traverse the tree, so the actual size of the tree doesn't matter for memory usage.
Time isn't either, if we're speaking about solving it. It doesn't matter how long it takes, just that it can be done in finite time. The required time can always be shortened by using more or faster comps.
Neither will solving chess make the game less interesting. There would be no way for a human to make use of the 'solution'.
In short, all what you've said was wrong.
--
GCP
This kind of pruning techniques have been used in chessprograms for ages. Some of them, like alpha-beta pruning, are 'perfect', i.e. they never cut off anything they shouldn't have.
But the techniques you describe are forward pruning techniques.
NONE of those is perfect. Some of the more widely used ones (ie nullmove pruning) have well-known flaws. If you want to 'solve chess', you cannot use them.
'Eliminating stupid moves' is something that sounds easy, but in practise, simply isn't possible to do perfectly.
It isn't 'plainly obvious' at all. It just sounds that way.
--
GCP
Not at all. Solving chess means that we know whether the first players wins, draws or loses by perfect play.
The 'solution' would be a number, and a series of moves that establishes that result. (there may be several solutions, but only one would be needed)
Knowing the perfect move from every position is a practical problem that is not related to solving the game itself.
--
GCP
Not really. Knuth & Moore have proven the lower bound of the number of positions to be looked at, in order to prove the value (win/loss/draw) of a certain position as:
b^floor(d/2) + b^ceil(d/2) - 1
where b equals branching factor and d equals ply depth
If you know that b = 38 for chess, and the average chess game lasts 40 moves (d=80), you will see that even with a pruning technique that always reaches this minimum, the number is still terribly LARGE.
A better pruning technique would improve pratical performance of chess-playing programs, but would not make the game more solvable.
--
GCP
The problem is that currently you can take a GPL'ed program, say the GIMP, stick your name all over it, remove the original authors credits and put it up as a server-based app.
You could even hack it up a little, so that it would only be fully usable with your proprietary client software, and so that other software would be more cumbersome or unstable for using it.
You're not distributing anything, so nobody can force you to give out sources.
Legally, what you're doing is fine. But I'm pretty sure the original authors won't be too happy with it.
--
GCP
2 questions:
/. article, many people don't even seem to understand the difference between free and open-source software.
a) Today free software has come to a point where lots of people are using it without understanding or caring about the underlying ideas. More and more commercial software is coming to platforms which previously solely consisted of free software. Projects like KDE flourish based on non-free foundations. Even in the comments to this
When you try to focus attention to the free software idea, like in the GNU/Linux case, or the Library->Lesser GPL case, you are confronted with lack of understanding.
Considering this, do you think there is a realistic threat that in the near or distant future the free software ideas will be buried and forgotten ?
If so, how are you and the FSF going to stop this, especially considering that previous attempts seem to have had more of a negative than a positive effect ?
b) Sometimes you are called a communist, mostly by people that do not like you. But what *are* your views on the political and economical world ? How much do you think the free software philosophy is extendable outside of the software itself ?
--
GCP
Why not put _your_ path to free software on your webpage?
I may not be interesting to as many people as RMS's would, but it would provide interesting reading anyway.
--
GCP
I am not RMS, but I have discussed with him on this subject, so I can probably answer on some of the issues.
The GPL indeed has no power in the case of server software, that provides a service without being distributed.
There were thoughts about addressing this in GPL 3, but the FSF (and RMS) were unsure whether it was doable, the right thing to do, and whether it would be accepted by the community.
For that last issue, don't forget that most, if not all programs which are GPL'ed refer to GPL 2 'or (at your option) any later version'. So if a new version of the GPL comes out, it automatically has effect on all older GPL'ed software.
Publishing a new GPL is not something that will be done without *lots* of thought.
--
GCP