Re: [proaudio] Question for proaudio devs & team |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/proaudio Archives
]
Hello Nedko,
How can we contribute ?
To upgrade the pro-audio overlay, we need at least :
- enumerate all packages and versions in the dead overlay
- find the latest versions
- elaborate a priority list (by the needs of the contributors, in my
case : rt-sources, jack, ...)
- upgrade each ebuild
At my side, I will locally bump all my used packages, then propose
them for the overlay (which one ?)
If the pro-audio dev teal don't have the time, can we ask to "get the
keys" and work on the tuxfamily hosted site?
Or ask Gentoo to incorporate pro-audio in a more maintained overlay?
Kind regards,
Xavier Miller.
Quoting Nedko Arnaudov <nedko@xxxxxxxxxxxxx>:
Content-Transfer-Encoding: quoted-printable
Xavier Miller <xavier.miller@xxxxxxxxx> writes:
Hello, is there no more maintainer ?
To me, proaudio overlay appears to be inactive. I recently switched to
gentoo and plan to invest some time in linux audio ebuilds. As wmladi
(from laditools) was unusable because wmdocklibg is missing, I
git-svn-ed and created a git branch where i'll push ladi related
stuff. The repo can be seen here:
http://nedko.arnaudov.name/git/cgit.cgi/proaudio-overlay/
and cloned from here:
http://nedko.arnaudov.name/git/proaudio-overlay.git
my stuff is in the ladi branch, the master branch follows the svn trunk
If proaudio overlay switches to git, more distributed development model
will be possible. Maintaining master branch probably should remain but
at least ebuild contributors/maintainers could merge important stuff
without overhead of a central authority. Don't get me wrong, I'm not
against proaudio central authoritiy. It is useful entity to provide
proaudio overlay for users with less or missing packaging skills. It is
useful to provide somewhat consistent policy. OTOH, I, as an upstream
developer have some vision on proaudio stuff and it may collide with
expectations or requirements of some other respectful folks (see
jackdbus flamewar from May as an example). I prefer the packaging
approach where user has the choice and packager provides it. This
matches the gentoo philosophy but in reality ebuilds may give suboptimal
set of choices.
So there are two good things about distributed development model:
* split overlay policy enfororcement from ebuild contribution/maintainership
* allow merges between ebuild contributors without participation of the
central authority.
As bad things I could imagine:
* it is not what is currently used and people may feel pain from
svn -> git migration
* it reduces the push to central authority to maintain something usable
by the users without packager skills.
I have to admit that as a gentoo newbie I may have missed
something. Especially branching overlays may not be good idea because it
is alternative to the overlay mechanism itself. Please, tell me if it is
so.
--
Nedko Arnaudov <GnuPG KeyID: DE1716B0>
----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.