The following files are impacted:
Jelmer, is the GPL license intentional in this case?
In order to change the pytalloc license to LGPL, the following developers would need to agree:
Jelmer Vernooij <email@example.com>
Andrew Tridgell <firstname.lastname@example.org>
Stefan Metzmacher <email@example.com>
(In reply to comment #0)
> The following files are impacted:
> Jelmer, is the GPL license intentional in this case?
> In order to change the pytalloc license to LGPL, the following developers would
> need to agree:
> Jelmer Vernooij <firstname.lastname@example.org>
> Andrew Tridgell <email@example.com>
> Stefan Metzmacher <firstname.lastname@example.org>
Would be ok for me...
What about the others python bindings of tdb, tevent, ldb?
(In reply to comment #1)
> > In order to change the pytalloc license to LGPL, the following developers would
> > need to agree:
> > Jelmer Vernooij <email@example.com>
> > Andrew Tridgell <firstname.lastname@example.org>
> > Stefan Metzmacher <email@example.com>
> Would be ok for me...
> What about the others python bindings of tdb, tevent, ldb?
pytdb.c, pytevent.c and pyldb.c are all LGPL.
Created attachment 8945 [details]
I'm happy to see my pytalloc work relicensed as LGPLv3 or later.
We are going backwards on this one.
Author: Andrew Bartlett <firstname.lastname@example.org>
Author: Andrew Tridgell <email@example.com>
Author: Björn Jacke <firstname.lastname@example.org>
Author: Douglas Bagnall <email@example.com> *
Author: Garming Sam <firstname.lastname@example.org>
Author: Gary Lockyer <email@example.com>
Author: Jelmer Vernooij <firstname.lastname@example.org> *
Author: Kamen Mazdrashki <email@example.com>
Author: Kristján Valur <firstname.lastname@example.org>
Author: Lukas Slebodnik <email@example.com>
Author: Noel Power <firstname.lastname@example.org>
Author: Petr Viktorin <email@example.com>
Author: Stefan Metzmacher <firstname.lastname@example.org> *
Author: Timur I. Bakeyev <timur@iXsystems.com>
For my part, I am happy to see pytalloc relicensed as LGPLv3.
The authors marked with asterisks above have indicated consent in the comments here.
I'm also ok with pytalloc being relicensed as LGPLv3.
Fine by me, I'm happy for pytalloc to be LGPLv3
To the people just added to the CC list:
Talloc is LGPL, while pytalloc is GPL. We want to regularise that by making pytalloc LGPL. (precisely, version 3 or greater in each case).
YOU are a pytalloc contributor (thank you!) so we need your assent.
See the proposed patch.
+1 for the change
I can't think of a good reason why pytalloc should be GPL rather than LGPL; it's almost certainly just oversight on my part. Maybe pytalloc started its live outside of lib/talloc, somewhere else in Samba and only made it in later?
Either way, +1 on making it LGPL.
Just FYI - anyone who signed the Samba (C) statement here:
already implicitly gave consent to re-license under LGPLv3.
also fine for me for lgpl
As email@example.com is no longer a valid email address (and hence the bugzilla account has disappeared), I contacted Kristján Valur Jónsson using the email address on the gitlab account from which we got the patches. He replied with:
> I approve of the proposed licencing change of Pytalloc.
> Kristján Valur Jónsson, formerly firstname.lastname@example.org
Fine by me.
+1 from me.
(In reply to Jelmer Vernooij from comment #10)
> Maybe pytalloc started its live outside of lib/talloc, somewhere else in Samba and only made it in later?
Yep. It was source4/scripting/python/pytalloc.[ch] until commit 0f043c197c473c801fc32c727194b5a2d6ae232f ("Move pytalloc to talloc directory.", October 2008).
(In reply to Jeremy Allison from comment #12)
> Just FYI - anyone who signed the Samba (C) statement here:
> already implicitly gave consent to re-license under LGPLv3.
Thanks. I'm trying to get explicit consent also, because it seems possible and better. And I'm not sure that all of the remaining three will have signed the policy.
If you need my blessing - here it is :)
On 2/11/20 10:22 am, Andrew Tridgell wrote:
> Hi Douglas,
>> You contributed quite a bit to the Python bindings. Do you mind if we
>> relicense as LGPL?
> that is OK by me.
> Cheers, Tridge
Created attachment 16321 [details]
patch v2 with signed-off-by, BUG:, etc
Our patch conventions have changed. I've added a new version with:
* Signed-off-by David Disseldorp
* Reviewed-by all those who clicked the bugzilla review thing
* BUG: line
and not as a purely procedural thing, but because the information is on hand:
* a note based on comment #17, explaining how this came about.
Comment on attachment 16321 [details]
patch v2 with signed-off-by, BUG:, etc
Looks fine to me, Douglas. I'd also be fine if you used your authorship, given that you've done all the hard work to track down and get these acks.
The one author we haven't heard from added c0b9526aaf29442f85f62232e22aecfb484b1576 and 76a8ae9e0db0340bfa96c428d55afe9b885977e0.
According to git blame, these lines are gone, but when you look at the code, the essence of each commit remains (that is, talloc_unlink() and a char* cast, respectively).
I am not sure how to proceed here. Jeremy, did Kamen sign the copyright thing?
I don't have a record of his signing, but he has a @samba.org address, which means I don't think there's a conflict in org goals here.
(In reply to David Disseldorp from comment #22)
David: you keep authorship!
And since you are the author, are you able to work out if/how we add WHATSNEW entries for libtalloc releases?
(In reply to Douglas Bagnall from comment #25)
> (In reply to David Disseldorp from comment #22)
> David: you keep authorship!
> And since you are the author, are you able to work out if/how we add
> WHATSNEW entries for libtalloc releases?
IIUC, talloc "release notes" normally just get listed in the tagged release commit.
I am fine with pytalloc being relicensed as LGPLv3.
Apologies for super slow response
(In reply to Kamen Mazdrashki from comment #27)
thanks Kamen! That's everyone!
I have pushed to https://gitlab.com/samba-team/samba/-/merge_requests/1839
Because I checked and confused myself again, here is a note about the list of authors, and why it lacks the Tim Potter that you will find with `git log --follow lib/talloc/pytalloc.c`. That --follow follows pytalloc.c back to its beginnings as source4/scripting/python/pytalloc.c, and then decides that pytalloc.c was derived from source4/utils/setntacl.c by replacing everything but the GPL waffle and an #include. (presumably source4/utils/setntacl.c was the shortest file containing this). So the list above is from `git log lib/talloc/pytalloc* | grep Author: | sort -u` combined with manually doing the follow with `git log 0f043c197c473c801fc32c727194b5a2d6ae232f -- source4/scripting/python/pytalloc* | grep Author | sort -u`. That last one tells you it is Jelmer all the way down, which does not expand the set of 14.
This bug was referenced in samba master: