Re: New and improved development model?

Petter Reinholdtsen (pere@hungry.com)
Sun, 13 Dec 1998 23:44:18 +0100 (MET)

[David Mosberger-Tang]
> Are you volunteering to run and maintain this?

It neccesary, yes. It is not very hard to do. :-)

> If the regular developers express a strong interest in this, I'm
> willing to do this.

> However, I'd want to keep the list of people with write access very
> short. Not to prevent from people from contributing to SANE, but to
> ensure that all submissions are reviewed by someone who has
> demonstrated a longstanding committment and understanding of SANE.

I'd expect 2-3 peoples with write access. Enough to handle someones
holidays without stopping development. :-)

> Lack of necessity? Note that SANE is very different from other
> projects because it is primarily an API. If you work on
> backend/frontend X, the need for having the latest and greatest
> version of backend/frontend Y is virtually non-existent.

When I work on porting it to HP/UX and Irix, I would like to get the
latest changes without searching thu the mailinglist and webpages to
find the patches.

[David Huggins-Daines]
> Has anyone asked Debian or RedHat? (i.e. cvs.gnome.org and
> cvs.gimp.org) SANE is relatively small compared to GNOME or the
> GIMP, I doubt it would be much of a strain on their servers.

Maybe that is a good idea. I was thinking about either setting up my
own CVS tree, or include it at cvs.hungry.com.

> FWIW, I like the idea of at least having anonymous CVS access and
> regular snapshot releases / diffs (doesn't have to be nightly). That
> way, people would have more of a chance to try compiling and running
> the new stuff in a variety of different configurations (and a
> variety of different GCC versions :-) right up to the minute an
> official release is made.

I would like this too.

[David Mosberger-Tang]
> Huh? The pre1.0 release has been available for several weeks before
> we switched to 1.0. I don't see how would more frequent snapshots
> would get people to compile SANE more often. I think I strongly
> encouraged people to test out the pre1.0 releases and it's not my
> fault if people don't test it on platforms other than Linux until
> after 1.0 was released.

Well. Something changed from pre1.0 to 1.0 that made HP/UX stop
working. I tested pre1.0, but missed a new patch to check which
changes other made, and to verify that my patches was included into
the distribution.

> Maybe the real issue is that we may need/want a sane-testers mailing
> list so people can be alerted to impending new releases without
> having to deal with the volumes of mail on sane-devel. I'd be happy
> to maintain such a list.

I don't think there is need for another mailing-list. It would not
help me with my problems. :-)

-- 
##>  Petter Reinholdtsen <##    | pere@td.org.uit.no
 O-  <SCRIPT Language="Javascript">window.close()</SCRIPT>
http://www.hungry.com/~pere/    | Go Mozilla, go! Go!

--
Source code, list archive, and docs: http://www.mostang.com/sane/
To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com