[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Darkplaces commits

At Tue, 21 May 2013 10:31:32 +0100,
Simon McVittie wrote:
> Thanks, I've had some time to look at Darkplaces now, and committed some
> more changes:

Thank you

> * when importing new upstream stuff, please at least skim-read/search
>   the diff, and watch out for new copyright holders/licenses/etc.

Here I have made a mistake - I think I was too naive.  On the
upstream branch,

git diff HEAD HEAD^ | grep -i copyright

returns nothing.  With hindsight, knowing that the year changed, it is
clear that there should be a change in the copyright file and so I
should have done something deeper.  Should I have checked the commit
logs to update the copyright file?

> * in this particular package, also please watch out for new (ab)uses
>   of dlopen() to load shared libraries - for Debian we want them all
>   to be either a "real" dependency or never used, I think

Will do.

> I was going to do an upload with your updates + my additional changes,
> but while testing it I found a nasty bug in libsdl1.2 which prevents
> darkplaces from exiting when run on my system
> (<http://bugs.debian.org/708760>). It doesn't seem to be a regression in
> darkplaces, but it's also pretty problematic - if you're running in
> full-screen or the mouse is captured, you can only recover by logging in
> on a text console and killing darkplaces. So, I'm not uploading just
> yet, in the hope that we can sort that out first.
> If that bug turns out to only affect pulseaudio/experimental or
> something, then I might upload anyway.

I have not encountered this with pulseaudio/sid, however I too think it is good
to postpone the upload until this problem is fixed rather than panic
whenever pulseaudio migrates.

> Please let me know when d0-blind-id is ready for review. I'll probably
> although if you don't mind me joining the Uploaders,

Of course.

I have only one remaining issue with d0-blind-id.  It is regarding
hardening (fortified functions of d0-blind-id.so), but I believe that
it is a false positive (using dh 9, autotools and the other hardening
options work correctly) but I will check that the flags are correctly
getting passed.

> Upstream have made actual releases since then

Release    - commit
0~20121222 - 11867
0~20130301 - 11920
0~20130304 - 11924

I do not know of the quality of these releases except that, for
0~20121222, the release note contains mostly fixes.  The other releases
look like they may be more intrusive.  I will investigate.

> If you (or the Xonotic developers) can encourage LordHavoc to tag
> releases in svn, I'm sure that would make life easier too.

I will try...


Reply to: