Re: Support for VIA driver, patch against SVN HEAD
On Wed, Jun 02, 2004 at 08:30:00PM -0400, Nathanael Nerode wrote: Branden Robinson wrote: This leaves a few questions open: A) What does David Dawes regard as attribution to [him]?[1] If the responsibility for the CVS commit is attributed to him, as it appears to be in several of the above, does he consider the XFree86 1.1 license to attach to those changes? I would really like to know this. I have not included such commits in my most recent summary of potential XFree86 1.1 license issues in X.org. I don't know the answer. Perhaps you could ask him, as my efforts over the past or so to obtain useful information from Mr. Dawes have been in vain. The changes would generally be copyright their authors. I would think so, yes. If the authors were contacted directly and released the patches, Dawes could be ignored. I would think so, yes. B) What is Marc La France's policy regarding the application of the XFree86 1.1 license to modifications he makes to files that bear no copyright notice by him or by the XFree86 Project, Inc.? Maybe someone could ask him personally? (Or is he not responding to your mail either?) There were quite a lot of these, and likewise I have not included them in that summary. I haven't asked Mr. La France personally, but given some of his messages[1][2] to the X.Org Foundation list, which are vigorously in favor of the new XFree86 license policy, I would wager he has adopted a policy similar to Mr. Dawes's. C) Does the XFree86 Project, Inc., consider patches submitted to their Bugzilla system to have any applicable copyrights therein assigned to them? It had better not; to my knowledge, that doesn't actually happen under US law. There's no way that a voluntarily submitted thing from an outsider is a work for hire, and copyright assignments have to be signed and on paper, last time I checked. That's the best of your and my knowledge; it has not been established that The XFree86 Project, Inc., its Board of Directors, or its President knows this. Is there someone who'd like to broach these questions with the XFree86 Project? My mails seem to go unanswered. Don't mail them. Mail the individual authors, perhaps? Sure; it could be done. At any rate, for any of the above where we can get the licensing straightened out, I'd be happly to apply the relevant patches. It may be worth contacting Thomas Hellström and Luc Verhaegen to inquire as to the provenance of their patches. Would it be worth contacting the others? Given a few anecdotes since the relicensing, I'd say yes. [1] Message-ID: [EMAIL PROTECTED] http://www.opengroup.org/sophocles/show_mail.tpl?source=Llistname=xorg_foundationid=100 [2] Message-ID: [EMAIL PROTECTED] http://www.opengroup.org/sophocles/show_mail.tpl?source=Llistname=xorg_foundationid=140 -- G. Branden Robinson| If you're handsome, it's flirting. Debian GNU/Linux | If you're a troll, it's sexual [EMAIL PROTECTED] | harassment. http://people.debian.org/~branden/ | -- George Carlin signature.asc Description: Digital signature
Re: Support for VIA driver, patch against SVN HEAD
Branden Robinson wrote: This leaves a few questions open: A) What does David Dawes regard as attribution to [him]?[1] If the responsibility for the CVS commit is attributed to him, as it appears to be in several of the above, does he consider the XFree86 1.1 license to attach to those changes? I would really like to know this. I have not included such commits in my most recent summary of potential XFree86 1.1 license issues in X.org. The changes would generally be copyright their authors. If the authors were contacted directly and released the patches, Dawes could be ignored. B) What is Marc La France's policy regarding the application of the XFree86 1.1 license to modifications he makes to files that bear no copyright notice by him or by the XFree86 Project, Inc.? Maybe someone could ask him personally? (Or is he not responding to your mail either?) There were quite a lot of these, and likewise I have not included them in that summary. C) Does the XFree86 Project, Inc., consider patches submitted to their Bugzilla system to have any applicable copyrights therein assigned to them? It had better not; to my knowledge, that doesn't actually happen under US law. There's no way that a voluntarily submitted thing from an outsider is a work for hire, and copyright assignments have to be signed and on paper, last time I checked. Is there someone who'd like to broach these questions with the XFree86 Project? My mails seem to go unanswered. Don't mail them. Mail the individual authors, perhaps? At any rate, for any of the above where we can get the licensing straightened out, I'd be happly to apply the relevant patches. It may be worth contacting Thomas Hellström and Luc Verhaegen to inquire as to the provenance of their patches. Would it be worth contacting the others? For the others, I am willing to write up a plain-English description as I did for the AT ioctl() issue. [1] http://www.mail-archive.com/devel%40xfree86.org/msg05906.html -- There are none so blind as those who will not see.
Re: Support for VIA driver, patch against SVN HEAD
On Mon, Mar 29, 2004 at 10:28:30AM +1000, Adam Conrad wrote: As a follow-up to my patch posting for VIA support, here's a quick discussion on licensing from IRC: infinity = 'Adam Conrad' overfiend = 'Branden Robinson' infinity No comment on the VIA driver patch? overfiend haven't done anything with it yet, sorry infinity I didn't write about licensing, but I just duoble-checked, and everything is clean. infinity CVS rev 1.9 of the Imakefile claims the changes between 1.8 and 1.9 are under the 1.1 license. Ironically, those changes break the compile and I rolled back to version 1.8 anyway. infinity All the other files in via/* are under the old license (and won't be relicensed without consent from VIA/S3), and everything outside of via/ was changed by hand by me, so no changes were pulled from upstream. overfiend okay. You don't want to post to -x about this? infinity I'll just post this IRC log. :) overfiend okay :) In light of recent messages by David Dawes to the XFree86 devel list[1][2][3], I'm not sure it is safe to include anything from XFree86 CVS after 2003-02-11, even in the via driver. [1] http://www.mail-archive.com/devel%40xfree86.org/msg05906.html [2] http://www.mail-archive.com/devel%40xfree86.org/msg05959.html [3] http://www.mail-archive.com/devel%40xfree86.org/msg05939.html -- G. Branden Robinson| A fundamentalist is someone who Debian GNU/Linux | hates sin more than he loves [EMAIL PROTECTED] | virtue. http://people.debian.org/~branden/ | -- John H. Schaar signature.asc Description: Digital signature
Re: Support for VIA driver, patch against SVN HEAD
On Mon, May 10, 2004 at 05:45:47PM -0500, Branden Robinson wrote: On Mon, Mar 29, 2004 at 10:28:30AM +1000, Adam Conrad wrote: As a follow-up to my patch posting for VIA support, here's a quick discussion on licensing from IRC: infinity = 'Adam Conrad' overfiend = 'Branden Robinson' infinity No comment on the VIA driver patch? overfiend haven't done anything with it yet, sorry infinity I didn't write about licensing, but I just duoble-checked, and everything is clean. infinity CVS rev 1.9 of the Imakefile claims the changes between 1.8 and 1.9 are under the 1.1 license. Ironically, those changes break the compile and I rolled back to version 1.8 anyway. infinity All the other files in via/* are under the old license (and won't be relicensed without consent from VIA/S3), and everything outside of via/ was changed by hand by me, so no changes were pulled from upstream. overfiend okay. You don't want to post to -x about this? infinity I'll just post this IRC log. :) overfiend okay :) In light of recent messages by David Dawes to the XFree86 devel list[1][2][3], I'm not sure it is safe to include anything from XFree86 CVS after 2003-02-11, even in the via driver. The simple reality of copyright law is that David Dawes cannot relicense something on behalf of other contributors, and that whatever copyright the authors (in this case, VIA/S3) have put on the code, stands. Whether or not it was later merged into XFree86 by an external party (Mr. Dawes), is completely irrelevant. -- Daniel Stone[EMAIL PROTECTED] Debian: the universal operating system http://www.debian.org signature.asc Description: Digital signature
Re: Support for VIA driver, patch against SVN HEAD
Using my sanitized XFree86 CVS tree, I have prepared my own 000_stolen_from_HEAD_via_driver.diff, which stops at 2003-02-12, the day before the application of the XFree86 1.1 to the tree. This lost 3 small commits: 1) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/02/20 13:46:36 Log message: 814. Fix missing ';' in via_driver.c hidden by an empty macro (Bugzilla #813, Luc Verhaegen). Modified files: xc/programs/Xserver/hw/xfree86/: CHANGELOG xc/programs/Xserver/hw/xfree86/doc/sgml/: RELNOTES.sgml xc/programs/Xserver/hw/xfree86/drivers/via/: via_driver.c Revision ChangesPath 3.3152+3 -1 xc/programs/Xserver/hw/xfree86/CHANGELOG 1.114 +2 -1 xc/programs/Xserver/hw/xfree86/doc/sgml/RELNOTES.sgml 1.29 +2 -2 xc/programs/Xserver/hw/xfree86/drivers/via/via_driver.c 2) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/02/20 13:50:07 Log message: 815. Fix for detection of hardwired panel for the via driver (Bugzilla #813, Luc Verhaegen, reported by Johannes Specht). Modified files: xc/programs/Xserver/hw/xfree86/: CHANGELOG xc/programs/Xserver/hw/xfree86/drivers/via/: via_bios.c Revision ChangesPath 3.3153+3 -1 xc/programs/Xserver/hw/xfree86/CHANGELOG 1.11 +19 -1 xc/programs/Xserver/hw/xfree86/drivers/via/via_bios.c 3) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/03/15 07:04:22 Log message: Fix driver name Modified files: xc/programs/Xserver/hw/xfree86/drivers/via/: via_driver.c Revision ChangesPath 1.30 +2 -4 xc/programs/Xserver/hw/xfree86/drivers/via/via_driver.c There have been four commits since then: 4) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/03/17 23:07:14 Log message: A second invocation of 'make' shouldn't change anything. -- These changes are Copyright (c) 2004 The XFree86 Project, Inc Rights as per version 1.1 of the XFree86 License (http://www.xfree86.org/legal/licenses.html). Modified files: [SNIP a whole bunch of files] 1.9 +2 -2 xc/programs/Xserver/hw/xfree86/drivers/via/Imakefile [SNIP more] That is the commit you said broke the build, right? 5) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/03/29 08:25:22 Log message: 40. Address Xv segfaults with Option XaaNoSolidFillRect by reducing output driver dependence on xaalocal.h (Marc La France). Modified files: [SNIP another gigantic commit] 1.12 +1 -2 xc/programs/Xserver/hw/xfree86/drivers/via/via_accel.c 1.5 +1 -2 xc/programs/Xserver/hw/xfree86/drivers/via/via_dga.c 1.31 +1 -2 xc/programs/Xserver/hw/xfree86/drivers/via/via_driver.c 1.18 +5 -38 xc/programs/Xserver/hw/xfree86/drivers/via/via_video.c 6) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/04/03 13:39:55 Log message: 56. Via driver updates: - Workaround for not properly detected memory allocation failure of the drm module. - In case of drm memory allocation failure, added a fallback to linear allocation in the pixmap cache when XFREE86_44 is set. - Increased the pixmap cache size to the hardware blitting limit if XFREE86_44 is set. - Restructuring and cleanups in via_accel.c (Removed unecessary comments and moved ViaInitLinear to via_memory.c) - Fixed some warnings about unresolved drm symbols. (Bugzilla #1320, Thomas Hellstr?m). Modified files: xc/programs/Xserver/hw/xfree86/: CHANGELOG xc/programs/Xserver/hw/xfree86/drivers/via/: via_accel.c via_driver.c via_driver.h via_memory.c Revision ChangesPath 3.3206+12 -1 xc/programs/Xserver/hw/xfree86/CHANGELOG 1.13 +12 -54 xc/programs/Xserver/hw/xfree86/drivers/via/via_accel.c 1.32 +3 -1 xc/programs/Xserver/hw/xfree86/drivers/via/via_driver.c 1.14 +10 -4 xc/programs/Xserver/hw/xfree86/drivers/via/via_driver.h 1.6 +72 -28 xc/programs/Xserver/hw/xfree86/drivers/via/via_memory.c 7) CVSROOT:/home/x-cvs Module name:xc Changes by: [EMAIL PROTECTED] 04/04/03 14:05:02 Log message: 59. Cleanup the via driver's primary DDC/EDID retrieval: -
Re: Support for VIA driver, patch against SVN HEAD
As a follow-up to my patch posting for VIA support, here's a quick discussion on licensing from IRC: infinity = 'Adam Conrad' overfiend = 'Branden Robinson' infinity No comment on the VIA driver patch? overfiend haven't done anything with it yet, sorry infinity I didn't write about licensing, but I just duoble-checked, and everything is clean. infinity CVS rev 1.9 of the Imakefile claims the changes between 1.8 and 1.9 are under the 1.1 license. Ironically, those changes break the compile and I rolled back to version 1.8 anyway. infinity All the other files in via/* are under the old license (and won't be relicensed without consent from VIA/S3), and everything outside of via/ was changed by hand by me, so no changes were pulled from upstream. overfiend okay. You don't want to post to -x about this? infinity I'll just post this IRC log. :) overfiend okay :) ... Adam -- backup [n] (bak'up): The duplicate copy of crucial data that no one bothered to make; used only in the abstract. 1024D/C6CEA0C9 C8B2 CB3E 3225 49BB 5ED2 0002 BE3C ED47 C6CE A0C9
Support for VIA driver, patch against SVN HEAD
I have made an SVN diff[1] between trunk rev 1172 and my working copy
which adds the VIA driver. Changelog entry follows:
* Backported VIA driver from 4.4 HEAD, 20040324. This driver adds
support
for the VIA CLE266 and related northbridge integrated chipsets,
typically
found in VIA EPIA-M Mini-ITX systems. (Adam Conrad
[EMAIL PROTECTED])
- debian/patches/000_stolen_from_HEAD_via_driver.diff
+ adds xc/programs/Xserver/hw/xfree86/drivers/via/*
+ adds via to xc/programs/Xserver/hw/xfree86/Registry
+ adds via to XF86CardDrivers for ia32 in
xc/config/cf/xfree86.cf
- Add support in debian/ directory for the new VIA driver:
+ add via to DRIVER_LIST for i386 in xserver-xfree86.config.in
+ add /usr/X11R6/lib/{modules/drivers/via_drv.o,man/man4/via.4x}
to xserver-xfree86.install.i386 and MANIFEST.i386
---
This has been tested with a HEAD build on Sid, as well as with an
application of this patch against the Woody 4.3.0 backport sources.
Both worked flawlessly. This patch only adds the 2D driver. If there
is demand/need for the 3D/dri driver, I can add that too, however we
don't use it in-house, and I didn't want to touch more things than I had
to (especially since I'm hoping that this driver will make it into the
sarge release).
... Adam
[1] http://loki.0c3.net/~adconrad/xfree86/1172-wc.diff.gz
--
backup [n] (bak'up): The duplicate copy of crucial data that no one
bothered to make; used only in the abstract.
1024D/C6CEA0C9 C8B2 CB3E 3225 49BB 5ED2 0002 BE3C ED47 C6CE A0C9
