Does anyone know if SGI plans to use Precision Insight's direct rendering pipeline (XFree86 4.0, IOW) on the VW's, when the Linux support is ready? Or will all the work be done through a custom-built X server?
Without the pipeline infrastructure, it wouldn't be possible to render [as fast as possible] to an X window, wouldn't it?
I love this idea. Who of all people has shown more restraint, more dispassion, more level-headedness [when it comes to Linux] than Linus himself? I remember reading his response to the whole KDE/Gnome issue, and was astounded by how reasonable and impartial he was. And then, I realized, I myself had picked up a few of the strong feelings aired out in/. discussions.
Needless to say, it was a very enlightening moment.
Linus is not a zealot, but a true engineer. I, for one, can only hope to someday become as down-to-earth dispassionate as he has shown all along.
GQmpeg is nice, although it does have some issues with the frontend-to-mpg123 interface; namely, you start to hear MP3's about a half-second after they begin, and sometimes the buttons go dead for whatever odd reason. X11Amp didn't have at least the first problem when I tried it, IIRC . . .
Another one to mention is FreeAmp. It's also GPL'ed (sponsored by GoodNoise even), and has its own built-in decoder engine. A few ppl here mentioned Nitrane-- well, check this puppy out. It has a superhyperoptimized 128Kbps decoder: whereas mpg123 uses ~8-9% of my K6-2 300, freeamp uses just 0.1% (assuming top ain't lying). It should be even better now, now that NASM 0.98's been released . . .
Well, it looks like computers are finally getting to look as weird as they do in sci-fi films.
It's pretty interesting when you think that in 2010, your typical home computer will look something like that. (Of course, this impression came mostly from the "Intel concept PC" design-- the others look like the work of a lousy postmodernist. The bunnypeople didn't help either)
But dangit, I'm still waiting for someone to make a rad-hard case! I don't want any idiot with a EMP weapon to kill my uptime!
Re:What will the packages be called?
on
Corel Linux FAQ
·
· Score: 1
Obvious reasons, as in RedHat's marketshare. Everything comes out in.rpm.
As far as technical merit goes, I don't doubt.deb >.rpm. RPM's are easier to put together, however, and the single.spec file is a more elegant way of doing it than the control/copyright/dirs/rules/etc. assortment. IMHO, the ultimate package format would take the best qualities of those two (single definition file, rigid rules about specifying requirements, customizable scripting, all that good stuff)
(Oh, and it would NOT be named after any distro in existence;-)
What will the packages be called?
on
Corel Linux FAQ
·
· Score: 1
For some reason, I don't think they will be.deb's:-]
Incidentally, is anyone else out there somewhat dismayed at the fact that ever Linux package format since the tarball is named after the originating distro? RedHat has.rpm, Debian has.deb, Stampede has.slp... which, just from a political point of view, isn't going to help make any of them popular with others (although.rpm's have spread, for obvious reasons). Why not just come up with a package format to end all package formats, and call it.pkg or something similarly plain-vanilla neutral?
(At best, perhaps Corel could use the same exact format as.deb's but use a different extension... I wonder how well that would fly)
I suspect the primary danger of negative shareholder influence is not so much RedHat giving up on the GPL, as much as them giving up on the LSB efforts (and the goodwill toward it).
Right now, there are several commercial applications that are certified to work with "RedHat Linux." This doesn't mean they can't run on other distros, only that they have been tested with RedHat. But it would be in the interest of your typical greedy shareholder to have the system setup be a little more "unique," enough to break the 95%-100% compatibility there is now, such that major apps will ONLY run on RedHat. Other distros will either be left out in the cold, or be required to follow whatever convoluted/etc tricks et. al. RedHat does.
The obvious targets being PHT, OpenLinux and the other Linux competitors that have also been making inroads into the business field. Something like this would be a lot more subtle than an outright license change, and thereby a lot easier to bring about.
I just hope they keep their sights on NT marketshare. Young was right; there is still plenty there to go around [to all the Linux vendors].
Semi-on/off-topic here, but all the speak concerning super-proprietary DVD encoding specs got me wondering . . . .
Suppose Company X licenses the spec from the consortium, after signing the fifty or so NDA's necessary, to begin work on a decoder. Then, enter Bob, a Disgruntled Employee(tm) of Company X. Bob secretly makes a photocopy of the entire spec, packages it, and FedEx's it off to some country where no one really gives a damn about patents or copyrights. (Malaysia would probably be a candidate-- at least copyrights don't seem to matter much there).
The package somehow falls into the hands of MLUG, and they use it to write gdvd, a GPL'ed software DVD decoder. They put it online, and gdvd-0.8.tar.gz starts showing up on FTP servers all over the world.
Now, isn't this always a danger, for the DVD Consortium? They could nuke Company X, and do even worse things to Bob (*shudder*) but the proverbial genie would be out of the proverbial bottle. Would they be able to change the spec slightly, to make gdvd useless, while somehow retaining backward compatibility with other (legal) decoders? Would they be able to send a SWAT team into your house if they found out you'd downloaded it? Would they finally realize proprietary specs suck? What would happen?
Just a misgiving here, but if AOL has NullSoft, they would also own the copyrights associated with the WinAmp player . . . right?
Given that X11Amp uses the exact same user interface as WA, it seems conceivable that AOL could order a "cease and desist" for use of their UI graphic design (whereas Justin/former_NullSoft would probably not have cared, given that Windoze is his territory)
If AOL is seeking exclusivity over the WinAMP brand name/image, that might be a likely Step 2 . . . .
Gnu + penguin concept?
on
GNU Inside?
·
· Score: 2
Has anyone seen the logo that Debian distros put on the default Apache top-level page? It's a very beautiful silhouette of a gnu and a penguin, with "Debian GNU/Linux" across the top.
As good as this design is, I came across it only by chance, and have never seen it anywhere else. If it could be cannibalized (remove "Debian," maybe touch up here and there) it would make a splendid logo for the purpose at hand . . .
(P.S.: And if "GNU Inside" is too Intel-ish-- and I suspect it is-- why not use, say, "GNU System" or "The GNU System?")
What I'm wondering is how enthusiastically the developers are making this release. If I understand their previous dilemma correctly, the MOSIX package was basically something they could not do anything with (outside of their own internal use) without running smack into the GPL. Given the lack of ballyhoo on the site, it doesn't appear anyone at Hebrew is very happy about the tarball being online (done with sort of a "better than nothing" acerbity).
Anyone have the MOSIX-list posting available? That should have been insightful . . . .
(Of course, as another poster has mentioned, everything is now in place to show 'em how just good going GPL can be. Hopefully this can happen for more good software in the future)
but if you want to link your code with GPLed code, you need to license your code under the GPL. That doesn't keep you from also, say, licensing your code under the BSD license, or the MPL
I take it this refers only to the portion of code that is explicitly yours (and not a priori GPL'ed). But is it really possible to have different parts of a program under GPL + non-GPL, as long as the entire thing is already available under GPL? After all, it was noted:
you cannot further license the program or modify the GPL in any way
Of course, you could just distribute your code with the GPL'ed stuff ripped out-- a PITA, but maybe necessary for some reason . . . anyone know about this?
Code forking, when you think of it, is really a Good Thing(tm). It's what keeps OSS projects honest-- if anyone doesn't like what the original authors are doing (e.g. making it too complex, making the license non-free, etc.) then they can do something about it.
Take Mozilla, for example. Ever since they went under dual NPL/GPL licensing, they've opened themselves up to the danger of forking. But it hasn't happened, and probably won't. There is already much goodwill in the OSS community toward Mozilla.org because they are doing things correctly, and even more, RMS himself has asked that people work with (and not against) them.
Malicious forking? I would imagine this to be "hey, I'll fork the code tree for program X, add a feature or two, and put it out so that I can slap my name all over it." If X is already under GPL, then this is going to be a lot of trouble (maintaining an entire code tree) for comparatively little return (whee! see my kewl splash screen!) I doubt this would ever happen.
(The whole NeoPlanet/ActiveX control flap comes close... but that was module ownership, not code forking)
Awesome! Looks like these will be the ones who'll be rolling all those IRIX goodies SGI promised into the kernel! Just imagine: better scalability, cc:NUMA, a better MIPS codebase, etc. etc. *ooh! warm fuzzies*
This ranks second in coolness only to the development of a REAL libGL.so for Linux }:-)
Linus's Law (as you describe it, anyway) is just a simplifed version of Maslow's Hierarchy of Needs. I don't remember how it goes exactly, but at the bottom you have survival (food, sex, etc.), a couple steps up you have emotional security, and then at the very top there's self-actualization. There's five levels in all, but the others escape me at this moment:-]
I have to agree, things may very well be moving toward the higher, more interesting levels. The way I see it, technology can either 1) make us into more productive workers, and/or 2) ease the drudgery of life to let us focus more energy on fun things (like Linux!)
From the Industrial Revolution to now, it's been #1 for the most part-- but maybe, just maybe, the tables are turning . . . .
Isn't this already done by the ICQ/AIM clients? Of course, this sort of advertising would work especially well with Net-aware apps (softupdate!)
But hey, if it's closed-source, there's still work to do anyway, ah? What would be funny is if a software house decided to release an app under OSS, and threw ad blurbs all over the source code. (How about that for targetting a technically literate market segment?:-)
That's a pretty interesting way of putting a very important point. So often, some of us fall into the trap of thinking one operating system can/will do it all. But no, more is better, just as the availability of more than one soft drink is a Good Thing(tm).
There was an article just yesterday somewhere where the CEO of SCO (IIRC) was going on about how he loved Linux, and how it was improving SCO's bottom line.
The best line was near the end, where he stated: "We love Linux. Because it's free, we can steal anything we want. We can keep in step with it."
/usr/local/* is the perfect place to install non-RPM programs (.tgz's et. al.) If RPM put binaries into/usr/local, I wouldn't know right offhand whether I could just delete random files from that tree without breaking anything in/var/lib/rpm/*, instead of having to confirm the fact manually with 'rpm -qf '. The thought of mixing referenced and unreferenced files (esp. where multiple directories are involved) gives me the willies.
I know, this is just a small convenience in the greater scheme of things, but as with your example, it is also well thought-out. Just . . . well-thought out in a different way:-)
(And just not too long ago I was annoyed finding my RPM'ed GQmpeg binary in/usr/local/bin)
If they follow standard building-dedication procedures, they'll be sure to put up a nice, big 'ol 4'x5' oil-pastel portrait of him in the lobby. Everyone who works in that building is going to see his ugly mug whenever they walk in.
But hey, OTOH, it could become the greatest hack magnet ever since the Great Dome }:-)
Location via internic.net lookups? Somewhat simplistic a solution, no?
sighhh... it'd be nice if programs like xtraceroute could work without having to use massive IP location tables. For some reason I can't help but feel like a James Bond villian when I see my network hops neatly displayed on a map of the earth }:-)
I hear there's an RFC in the works that will help with this (or is it part of IPv6?)
Slashdot Mirror
Does anyone know if SGI plans to use Precision Insight's direct rendering pipeline (XFree86 4.0, IOW) on the VW's, when the Linux support is ready? Or will all the work be done through a custom-built X server?
Without the pipeline infrastructure, it wouldn't be possible to render [as fast as possible] to an X window, wouldn't it?
I love this idea. Who of all people has shown more restraint, more dispassion, more level-headedness [when it comes to Linux] than Linus himself? I remember reading his response to the whole KDE/Gnome issue, and was astounded by how reasonable and impartial he was. And then, I realized, I myself had picked up a few of the strong feelings aired out in /. discussions.
Needless to say, it was a very enlightening moment.
Linus is not a zealot, but a true engineer. I, for one, can only hope to someday become as down-to-earth dispassionate as he has shown all along.
GQmpeg is nice, although it does have some issues with the frontend-to-mpg123 interface; namely, you start to hear MP3's about a half-second after they begin, and sometimes the buttons go dead for whatever odd reason. X11Amp didn't have at least the first problem when I tried it, IIRC . . .
Another one to mention is FreeAmp. It's also GPL'ed (sponsored by GoodNoise even), and has its own built-in decoder engine. A few ppl here mentioned Nitrane-- well, check this puppy out. It has a superhyperoptimized 128Kbps decoder: whereas mpg123 uses ~8-9% of my K6-2 300, freeamp uses just 0.1% (assuming top ain't lying). It should be even better now, now that NASM 0.98's been released . . .
Well, it looks like computers are finally getting to look as weird as they do in sci-fi films.
It's pretty interesting when you think that in 2010, your typical home computer will look something like that. (Of course, this impression came mostly from the "Intel concept PC" design-- the others look like the work of a lousy postmodernist. The bunnypeople didn't help either)
But dangit, I'm still waiting for someone to make a rad-hard case! I don't want any idiot with a EMP weapon to kill my uptime!
Obvious reasons, as in RedHat's marketshare. Everything comes out in .rpm.
.deb > .rpm. RPM's are easier to put together, however, and the single .spec file is a more elegant way of doing it than the control/copyright/dirs/rules/etc. assortment. IMHO, the ultimate package format would take the best qualities of those two (single definition file, rigid rules about specifying requirements, customizable scripting, all that good stuff)
;-)
As far as technical merit goes, I don't doubt
(Oh, and it would NOT be named after any distro in existence
For some reason, I don't think they will be .deb's :-]
.rpm, Debian has .deb, Stampede has .slp... which, just from a political point of view, isn't going to help make any of them popular with others (although .rpm's have spread, for obvious reasons). Why not just come up with a package format to end all package formats, and call it .pkg or something similarly plain-vanilla neutral?
.deb's but use a different extension... I wonder how well that would fly)
Incidentally, is anyone else out there somewhat dismayed at the fact that ever Linux package format since the tarball is named after the originating distro? RedHat has
(At best, perhaps Corel could use the same exact format as
I suspect the primary danger of negative shareholder influence is not so much RedHat giving up on the GPL, as much as them giving up on the LSB efforts (and the goodwill toward it).
/etc tricks et. al. RedHat does.
Right now, there are several commercial applications that are certified to work with "RedHat Linux." This doesn't mean they can't run on other distros, only that they have been tested with RedHat. But it would be in the interest of your typical greedy shareholder to have the system setup be a little more "unique," enough to break the 95%-100% compatibility there is now, such that major apps will ONLY run on RedHat. Other distros will either be left out in the cold, or be required to follow whatever convoluted
The obvious targets being PHT, OpenLinux and the other Linux competitors that have also been making inroads into the business field. Something like this would be a lot more subtle than an outright license change, and thereby a lot easier to bring about.
I just hope they keep their sights on NT marketshare. Young was right; there is still plenty there to go around [to all the Linux vendors].
Semi-on/off-topic here, but all the speak concerning super-proprietary DVD encoding specs got me wondering . . . .
Suppose Company X licenses the spec from the consortium, after signing the fifty or so NDA's necessary, to begin work on a decoder. Then, enter Bob, a Disgruntled Employee(tm) of Company X. Bob secretly makes a photocopy of the entire spec, packages it, and FedEx's it off to some country where no one really gives a damn about patents or copyrights. (Malaysia would probably be a candidate-- at least copyrights don't seem to matter much there).
The package somehow falls into the hands of MLUG, and they use it to write gdvd, a GPL'ed software DVD decoder. They put it online, and gdvd-0.8.tar.gz starts showing up on FTP servers all over the world.
Now, isn't this always a danger, for the DVD Consortium? They could nuke Company X, and do even worse things to Bob (*shudder*) but the proverbial genie would be out of the proverbial bottle. Would they be able to change the spec slightly, to make gdvd useless, while somehow retaining backward compatibility with other (legal) decoders? Would they be able to send a SWAT team into your house if they found out you'd downloaded it? Would they finally realize proprietary specs suck? What would happen?
Just a misgiving here, but if AOL has NullSoft, they would also own the copyrights associated with the WinAmp player . . . right?
Given that X11Amp uses the exact same user interface as WA, it seems conceivable that AOL could order a "cease and desist" for use of their UI graphic design (whereas Justin/former_NullSoft would probably not have cared, given that Windoze is his territory)
If AOL is seeking exclusivity over the WinAMP brand name/image, that might be a likely Step 2 . . . .
Has anyone seen the logo that Debian distros put on the default Apache top-level page? It's a very beautiful silhouette of a gnu and a penguin, with "Debian GNU/Linux" across the top.
In case no one has seen it, I have it here.
As good as this design is, I came across it only by chance, and have never seen it anywhere else. If it could be cannibalized (remove "Debian," maybe touch up here and there) it would make a splendid logo for the purpose at hand . . .
(P.S.: And if "GNU Inside" is too Intel-ish-- and I suspect it is-- why not use, say, "GNU System" or "The GNU System?")
What I'm wondering is how enthusiastically the developers are making this release. If I understand their previous dilemma correctly, the MOSIX package was basically something they could not do anything with (outside of their own internal use) without running smack into the GPL. Given the lack of ballyhoo on the site, it doesn't appear anyone at Hebrew is very happy about the tarball being online (done with sort of a "better than nothing" acerbity).
Anyone have the MOSIX-list posting available? That should have been insightful . . . .
(Of course, as another poster has mentioned, everything is now in place to show 'em how just good going GPL can be. Hopefully this can happen for more good software in the future)
This is why works of fiction are not GPL'ed.
Artistic integrity is a very different animal from closed source.
but if you want to link your code with GPLed code, you need to license your code under the GPL. That doesn't keep you from also, say, licensing your code under the BSD license, or the MPL
I take it this refers only to the portion of code that is explicitly yours (and not a priori GPL'ed). But is it really possible to have different parts of a program under GPL + non-GPL, as long as the entire thing is already available under GPL? After all, it was noted:
you cannot further license the program or modify the GPL in any way
Of course, you could just distribute your code with the GPL'ed stuff ripped out-- a PITA, but maybe necessary for some reason . . . anyone know about this?
Code forking, when you think of it, is really a Good Thing(tm). It's what keeps OSS projects honest-- if anyone doesn't like what the original authors are doing (e.g. making it too complex, making the license non-free, etc.) then they can do something about it.
Take Mozilla, for example. Ever since they went under dual NPL/GPL licensing, they've opened themselves up to the danger of forking. But it hasn't happened, and probably won't. There is already much goodwill in the OSS community toward Mozilla.org because they are doing things correctly, and even more, RMS himself has asked that people work with (and not against) them.
Malicious forking? I would imagine this to be "hey, I'll fork the code tree for program X, add a feature or two, and put it out so that I can slap my name all over it." If X is already under GPL, then this is going to be a lot of trouble (maintaining an entire code tree) for comparatively little return (whee! see my kewl splash screen!) I doubt this would ever happen.
(The whole NeoPlanet/ActiveX control flap comes close... but that was module ownership, not code forking)
Awesome! Looks like these will be the ones who'll be rolling all those IRIX goodies SGI promised into the kernel! Just imagine: better scalability, cc:NUMA, a better MIPS codebase, etc. etc. *ooh! warm fuzzies*
This ranks second in coolness only to the development of a REAL libGL.so for Linux }:-)
Why the deuce was this comment moderated down? It was relevant, and made a good point!
Linus's Law (as you describe it, anyway) is just a simplifed version of Maslow's Hierarchy of Needs. I don't remember how it goes exactly, but at the bottom you have survival (food, sex, etc.), a couple steps up you have emotional security, and then at the very top there's self-actualization. There's five levels in all, but the others escape me at this moment :-]
I have to agree, things may very well be moving toward the higher, more interesting levels. The way I see it, technology can either 1) make us into more productive workers, and/or 2) ease the drudgery of life to let us focus more energy on fun things (like Linux!)
From the Industrial Revolution to now, it's been #1 for the most part-- but maybe, just maybe, the tables are turning . . . .
Isn't this already done by the ICQ/AIM clients? Of course, this sort of advertising would work especially well with Net-aware apps (softupdate!)
:-)
But hey, if it's closed-source, there's still work to do anyway, ah? What would be funny is if a software house decided to release an app under OSS, and threw ad blurbs all over the source code. (How about that for targetting a technically literate market segment?
That's a pretty interesting way of putting a very important point. So often, some of us fall into the trap of thinking one operating system can/will do it all. But no, more is better, just as the availability of more than one soft drink is a Good Thing(tm).
:-)
I will have to remember that
Pricey package, and definitely not open-source, but there are tons of programs that can import the format ;-)
There was an article just yesterday somewhere where the CEO of SCO (IIRC) was going on about how he loved Linux, and how it was improving SCO's bottom line.
The best line was near the end, where he stated: "We love Linux. Because it's free, we can steal anything we want. We can keep in step with it."
[rolls up GPL into a stick]
[THWACK]
/usr/local/* is the perfect place to install non-RPM programs (.tgz's et. al.) If RPM put binaries into /usr/local, I wouldn't know right offhand whether I could just delete random files from that tree without breaking anything in /var/lib/rpm/*, instead of having to confirm the fact manually with 'rpm -qf '. The thought of mixing referenced and unreferenced files (esp. where multiple directories are involved) gives me the willies.
:-)
/usr/local/bin)
I know, this is just a small convenience in the greater scheme of things, but as with your example, it is also well thought-out. Just . . . well-thought out in a different way
(And just not too long ago I was annoyed finding my RPM'ed GQmpeg binary in
If they follow standard building-dedication procedures, they'll be sure to put up a nice, big 'ol 4'x5' oil-pastel portrait of him in the lobby. Everyone who works in that building is going to see his ugly mug whenever they walk in.
But hey, OTOH, it could become the greatest hack magnet ever since the Great Dome }:-)
If y'all want a really good-looking Linux mascot, check out the lynx.
(bigger version here)
Now's there's something I want gracing my Linux box };-)
(apologies to Tux)
Location via internic.net lookups? Somewhat simplistic a solution, no?
sighhh... it'd be nice if programs like xtraceroute could work without having to use massive IP location tables. For some reason I can't help but feel like a James Bond villian when I see my network hops neatly displayed on a map of the earth }:-)
I hear there's an RFC in the works that will help with this (or is it part of IPv6?)