| |||||||||||||
|
| |||||||||||||
GPL concerns halt Kororaa live CD (NewsForge)GPL concerns halt Kororaa live CD (NewsForge)Posted May 16, 2006 2:59 UTC (Tue) by pheldens (guest, #19366)Parent article: GPL concerns halt Kororaa live CD (NewsForge)
It's good they are held to their responsibility.
(Log in to post comments)
GPL concerns halt Kororaa live CD (NewsForge) Posted May 16, 2006 4:25 UTC (Tue) by Arker (guest, #14205) [Link] Matrox cards are also well supported, and the Intel and Via onboards.
It would be very nice to be able to use the beefier videocards, but for very few people is it in any way essential.
ATI driver call from the FSF Posted May 16, 2006 4:42 UTC (Tue) by dwheeler (guest, #1216) [Link] The FSF's High Priority Free Software Projects list includes, as one of their top priority items, a 3D driver for ATI graphics cards (was here, moved to Freedesktop.org). I suspect they started with ATI because there were at least FLOSS drivers to start with. If good 3D ATI drivers get developed, that could be a wedge to get others to join... e.g., then NV either joins the party, or everyone stops buying their cards (since others wil be more supportable).I would love to see OSDL, Ubuntu, or other organiations work to help flesh out the ATI cards, PRONTO. If anyone can lend time to the ATI driver project, please do so.
ATI driver call from the FSF Posted May 16, 2006 10:53 UTC (Tue) by tajyrink (subscriber, #2750) [Link] There's also a new try for 3D NVIDIA drivers at: http://nouveau.freedesktop.org/wiki/
Mostly just planning, but anyway it's good that there's 3D life after Utah-GLX for NVIDIA cards, too.
grow up :-/ Posted May 16, 2006 8:33 UTC (Tue) by gvy (guest, #11981) [Link] What exactly responsibility? Why those who whine about licenses and responsibilities are routinely not even reading those licenses? The drivers are fine to ship.
Re "libre", I'm running "ati" on RM7000 with X.Org 7.1RC2 and you know what? Previous reply to you was eaten up by a frozen X server (together with another 20 tabs, half an hour and ~1.5 megs of GPRS traffic).
$subject
grow up :-/ Posted May 16, 2006 17:20 UTC (Tue) by Arker (guest, #14205) [Link] "The drivers are fine to ship."
Have you really thought this through?
You can ship a binary blob on the same cd with GPL software. You can choose to link that blob into the software, and run it and use it yourself. That's within your rights.
But, you *can't* turn around and redistribute that once you do. You can only redistribute the GPL part if you can give folks the source to the whole thing, and if part of it is a binary blob, you can't give the source to that, so you can't ship it.
So the key question here is which is actually happening. It's my understanding, and I've seen no reason to think otherwise, that in terms of a live CD it's not an option to simply package the blob and compile instructions - it all needs to be compiled and ready to run for the live cd concept to work. And if that's the case, it sure looks like a legal problem to me. I wouldn't touch it without a law firm behind me.
grow up :-/ Posted May 16, 2006 18:29 UTC (Tue) by khim (subscriber, #9252) [Link] The drivers are fine to ship... 100%, no problem (at least if they really have no GPLed code in binary blob) ...the kernel is fine to ship... 100% correct, kernel is kernel, GPL-licensed and all that ...on the same CD. Um, no - that'll be GPL violation. Why those who whine about licenses and responsibilities are routinely not even reading those licenses? If you have read GPL and know it so well then explain how this CD (derived work of Linux kernel!!!) can be distributed if part of it (binary drivers) are not GPLed ? How can you comply with item 2 of GPL requirement ? Note: as special exception GPLv2 allows creation of works where GPLed code and unrelated non-GPLed code are put on the same physical medium. This is indulgence of CPLv2 - GPLv1 disallowed such distribution (you were unable to distribute gcc with *bsd kernel back then). Can you honestly say that kernel and kernel driver are unrelated works ??? That kernel driver (with glue code, of course!) is "work not based on the kernel" ? This is kind of hard to justify, you know... Once more. There are three components:
too narrow Posted May 16, 2006 19:09 UTC (Tue) by sepreece (subscriber, #19270) [Link] I think it would depend on how you distributed. If you package the three separately, or if you prelinked the shim to the kernel (1+2) with the blob (3) simply on the disk, then you should be OK, because the blob is not "a work based on the program" and is allowed to be aggregated. However, if you prelink the blob to the shim (2+3), then you can't ship them on the same media with the kernel, because 2+3 IS a work based on 1. I don't know how the Kororaa distribution is structured.
Actually, after looking at the language again, it's not clear to me why you can't aggregate an unmodified binary kernel with anything you like (including things that would modify that kernel post-distribution), but that would depend on exactly how a court read the phrase "under the terms of Sections 1 and 2" that appear in section 3. Since section 2 only talks about distribution of modified versions, I don't see why it would bear on distributing an unmodified binary; I'll have to go see if Rosen says anything about how this interpretation works.
|
|
||||||||||||