Some GNU software is written by staff of the Free Software Foundation, but most GNU software is contributed by volunteers. Some contributed software is copyrighted by the Free Software Foundation; some is copyrighted by the contributors who wrote it.
Woohoo! As I suspected - under their definition of 'GNU Software' they define it that some is written by the FSF but most is written by volunteers.
OK - fair enough. You did say this a few posts up. RMS has said it too. You're right it's irrelevant - I find it more an attack on Linus's character than anything.
Care to respond to the other point? The body of work by the GNU project themselves - excluding GPL software written people whose opinions have not been poled - is surely a tiny amount of any distribution, right?
I certainly don't believe that RMS can speak for those who write GPL software, and infer that they
want representation under the GNU project. It's like socialists speaking for the unemployed, or the right-wing speaking for businesses. It's not scientific - there's no data or polls to say that the people they claim to represent have that view.
I really think it's a stretch to say that because someone releases software under a licence you can infer support and specific actions for an organisation. The GPL is a licence. It has terms and conditions. It's not a way of life - but regardless I think the GPL is the most free licence.
He claims the name because a vast majority of what you see on "linux" is actually stuff developed by the GNU project
LGPL/GPL stuff written specifically by the GNU project and not including software under the GPL by other authors?... well, that'd be tiny ammount - right? Are there any statistics?
how would you feel if i took something you wrote and named it all after myself? And when you ask for credit, i label you a 'fanatic'...
Well, it's not Linus at fault here - it's the distro makers, so why does RMS target Linus specifically? (and continue to hint at an ego that he named it after himself when it was someone else that chose the kernel's name).
I write GPL software, and I don't want Linux to be called GNU/Linux.
The GPL doesn't say that names have to be prefixed with GNU. As such, it's about (1) whether RMS can claim to have the authority of speaking for the majority of authors who put works out under the GPL, (2) what percentage of work is under the GPL, and (3) whether the percentage of work should warrant a name change.
3. Names rarely bare any relation to their target. The recent and odd proliferation of "XP" in the name should have taught that.
2. It's the majority licence, but not over 50% of software in Redhat.
1. As I said at the beginning, I write GPL software, and I prefer the name Linux. RMS doesn't speak for me, and just because he wrote a licence. I suspect that many GPL authors prefer the Linux name. It's foggy whether he can claim to speak for authors wanting credit for their work under the GNU moniker. I very much doubt he has any weight here.
You're missing the key feature of patents here. The one "proven" to invent first by virtue of a patent gets the privilege of setting arbitrary license terms for any use of the covered invention.
You know, I could have sworn I said that - oh yes, I sez:
"having a patent means that you get to decide the rules under which the technology [...] is used"
Thus, a patent has everything to do with licensing, and it may be "closed" at the whim of the patent holder at any time
I didn't say otherwise. I did say that it had nothing to do with specific types of licencing ("Patents have nothing to do with free, open or closedness licencing.") as the original post said "the only thing a patent is good for is to stop people from using an invention".
The fact that RedHat holds a few patents will not stop them from being sued by someone else for patent violation.
Well obviously... anyone can sue anyone for anything - but a few patents under your belt makes your case stronger and clearer. It proves that your company went through a generally respected process (not respected by the public - but respected by the courts, definately)
There is no difference between offensive patents and defensive patents - they're just patents, and it depends on how are used.
If someone achieves a patent, but you have prior art, it's more difficult to prove that you were first because you didn't go via The System[TM]. Disproving a patent via prior art is significantly more difficult than disproving a patent via other patents.
Patents have nothing to do with free, open or closedness licencing. They are merely to do with a system whose intention is to prove who invented what first.
If you don't apply for a patent and you use 'your technology' then someone else could more easily take legal action upon you for using 'their technology'.
In this way having a patent means that you get to decide the rules under which the technology (kill me now for using that word) is used. A good patent owner will licence it under good rules, and a bad patent owner will licence under bad rules.
So it all comes down to how we think the owners of this patent will act upon uses of their 'technology'.
It's fast at simple pages. But try saving a 300 comment nested view slashdot page locally, and then view that in Opera or Mozilla. Opera will take _much_ longer to render that page than Mozilla or IE (Mozilla is the undisputed king of rendering speed - it's the interface that's sluggish)
Translucent windows can be used to show inactive windows. This is more natural and clear than a window manager border colour change (though maybe a grey tone over a window would convey the idea better).
Transluceny is an important part of anti-aliasing, and this allows anti-aliasing windows with other windows. For readability it might not be good, but the edges of curved windows could be made smoother by anti-aliasing.
Consider an window state that isn't reveal/hidden but it's stages between. You could have a program open that is unselectable but that fades into view when you have email.
Translucency is also to do with Window shadows (as used in Mac-OSX) that are about the best way I have seen of expressing the layer order of windows. The drop shadows are cast by a window and are various levels translucency of a shadow colour.
X is a protocol, and X can run windowed inside other environments (ie, framebuffer).
Specifically, toolkits usually write to an X abstraction library which becomes X the protocol and then a data stream is rendered by an X client. The client can be a window on a framebuffer.
Unless the software ontop of X uses XFree specific code then you can run it on a framebuffer with X ontop.
The main problem is with X extensions but these are usually quite well programmed and most have framebuffer ports. GTK already has a framebuffer port.
I think the main point though is that it won't be much faster. X isn't slow, and there isn't any proof (only rumour) towards that argument.
Dude, OSX has worse font rendering than OS9, and pretty much any GUI (even '98). They cut many corners and it produces a blurry rendering of the font that discards kerning and some specific parts of hinting. Freetype is nice, and it renders better than XP or OSX (of course most distros fuck it up, but that's besides the point)
Well, you've also got to remember that what works best in browsers and what's part of the standard are two very different things. Netscape 4 has four attributes to control page margins which aren't part of any standard.
Documents should be made out of H1, H2, H3 and down with paragraph tags. Don't skip H1 and H2 and start at H3. Build your template around this.
Acronyms and abbreviations should only be defined at the first occurence. This is part of the HTML 4.01, XHTML spec.
I've seen css classes named 'red' (I'm not kidding - they were that dumb). Use semantic descriptions to describe the function of that marked up bit of text. Before you do any work decide on some CSS class names. Too often I've seen dozens of unique CSS classes made without any planning.
Often people put the title of the document in an H1 - this is wrong. H1's are supposed to occur many times in a document (besides, to use XPath notation, that's what html/head/title tag is for). If a book were to be marked up H1 would be each chapter's title. The markup of a document is independant from to how many pages it's spread across (ie, just because you put ecah chapter on another page don't make H1 into H2 so that there are many H1s).
Basically - it requires an understanding of your information, and software can't do that.
Comparing HTML to a programming langauge is ridiculous. When's the last time you could still use a bit of software and the compiler could cleanly ignore newer code it didn't understand (say, like, XHTML).
XML fixes most of the problems, but even so compliance is over-hyped. People think that so long as validator.w3.org passes you then you must have done everything right - when there can still be a million things wrong with your code.
I said it's equivilent. The emphasis was on the simplicity of revealing/hiding a DIV. Hiermenus are much more. They understand browser rendering bugs better than any other and they work around them for the users sake.
Slashdot Mirror
Perhaps they want to be everyone's foe as well.
Care to respond to the other point? The body of work by the GNU project themselves - excluding GPL software written people whose opinions have not been poled - is surely a tiny amount of any distribution, right?
I certainly don't believe that RMS can speak for those who write GPL software, and infer that they
want representation under the GNU project. It's like socialists speaking for the unemployed, or the right-wing speaking for businesses. It's not scientific - there's no data or polls to say that the people they claim to represent have that view.
I really think it's a stretch to say that because someone releases software under a licence you can infer support and specific actions for an organisation. The GPL is a licence. It has terms and conditions. It's not a way of life - but regardless I think the GPL is the most free licence.
Well, it's not Linus at fault here - it's the distro makers, so why does RMS target Linus specifically? (and continue to hint at an ego that he named it after himself when it was someone else that chose the kernel's name).
The GPL doesn't say that names have to be prefixed with GNU. As such, it's about (1) whether RMS can claim to have the authority of speaking for the majority of authors who put works out under the GPL, (2) what percentage of work is under the GPL, and (3) whether the percentage of work should warrant a name change.
3. Names rarely bare any relation to their target. The recent and odd proliferation of "XP" in the name should have taught that.
2. It's the majority licence, but not over 50% of software in Redhat.
1. As I said at the beginning, I write GPL software, and I prefer the name Linux. RMS doesn't speak for me, and just because he wrote a licence. I suspect that many GPL authors prefer the Linux name. It's foggy whether he can claim to speak for authors wanting credit for their work under the GNU moniker. I very much doubt he has any weight here.
2) My cake isn't round or square, it is raised in the middle and then loops down on itself. With one cut I make 8 equal parts with icing.
3) See #2.
Cube sez "5".
In two days there will be NO CUBE!
1) I can do it with one cube. Sides 1-6 are for monday-saturday. When I'm hiding the cube you can assume it's sunday.
2) 0, apparently the 3 switches and lights are on the outside of the room.
3) Just build it so the diameter equals the length from one corner to the diagonal opposite
__
/--\
||
And God's name, is Linus.
"having a patent means that you get to decide the rules under which the technology [...] is used"
I didn't say otherwise. I did say that it had nothing to do with specific types of licencing ("Patents have nothing to do with free, open or closedness licencing.") as the original post said "the only thing a patent is good for is to stop people from using an invention".
Ok, bye now!
There is no difference between offensive patents and defensive patents - they're just patents, and it depends on how are used.
If someone achieves a patent, but you have prior art, it's more difficult to prove that you were first because you didn't go via The System[TM]. Disproving a patent via prior art is significantly more difficult than disproving a patent via other patents.
If you don't apply for a patent and you use 'your technology' then someone else could more easily take legal action upon you for using 'their technology'.
In this way having a patent means that you get to decide the rules under which the technology (kill me now for using that word) is used. A good patent owner will licence it under good rules, and a bad patent owner will licence under bad rules.
So it all comes down to how we think the owners of this patent will act upon uses of their 'technology'.
I certainly trust Redhat.
Yes. It has go-faster stripes.
It's fast at simple pages. But try saving a 300 comment nested view slashdot page locally, and then view that in Opera or Mozilla. Opera will take _much_ longer to render that page than Mozilla or IE (Mozilla is the undisputed king of rendering speed - it's the interface that's sluggish)
Transluceny is an important part of anti-aliasing, and this allows anti-aliasing windows with other windows. For readability it might not be good, but the edges of curved windows could be made smoother by anti-aliasing.
Consider an window state that isn't reveal/hidden but it's stages between. You could have a program open that is unselectable but that fades into view when you have email.
Translucency is also to do with Window shadows (as used in Mac-OSX) that are about the best way I have seen of expressing the layer order of windows. The drop shadows are cast by a window and are various levels translucency of a shadow colour.
(asdf jkl; asdf jkl; asdf jkl;)
Specifically, toolkits usually write to an X abstraction library which becomes X the protocol and then a data stream is rendered by an X client. The client can be a window on a framebuffer.
Unless the software ontop of X uses XFree specific code then you can run it on a framebuffer with X ontop.
The main problem is with X extensions but these are usually quite well programmed and most have framebuffer ports. GTK already has a framebuffer port.
I think the main point though is that it won't be much faster. X isn't slow, and there isn't any proof (only rumour) towards that argument.
Read this: http://www.osnews.com/comment.php?news_id=760&limi t=no
Dude, OSX has worse font rendering than OS9, and pretty much any GUI (even '98). They cut many corners and it produces a blurry rendering of the font that discards kerning and some specific parts of hinting. Freetype is nice, and it renders better than XP or OSX (of course most distros fuck it up, but that's besides the point)
See here
Well, you've also got to remember that what works best in browsers and what's part of the standard are two very different things. Netscape 4 has four attributes to control page margins which aren't part of any standard.
Documents should be made out of H1, H2, H3 and down with paragraph tags. Don't skip H1 and H2 and start at H3. Build your template around this.
Acronyms and abbreviations should only be defined at the first occurence. This is part of the HTML 4.01, XHTML spec.
I've seen css classes named 'red' (I'm not kidding - they were that dumb). Use semantic descriptions to describe the function of that marked up bit of text. Before you do any work decide on some CSS class names. Too often I've seen dozens of unique CSS classes made without any planning.
Often people put the title of the document in an H1 - this is wrong. H1's are supposed to occur many times in a document (besides, to use XPath notation, that's what html/head/title tag is for). If a book were to be marked up H1 would be each chapter's title. The markup of a document is independant from to how many pages it's spread across (ie, just because you put ecah chapter on another page don't make H1 into H2 so that there are many H1s).
Basically - it requires an understanding of your information, and software can't do that.
XML fixes most of the problems, but even so compliance is over-hyped. People think that so long as validator.w3.org passes you then you must have done everything right - when there can still be a million things wrong with your code.
If you can do better - I'd love to see it.