Re: Controlling what's in the archive (Re: MIA, Incompetent and holiday-loving maintainers)
ti, 2004-01-06 kello 01:34, Benj. Mako Hill kirjoitti:
> As do I. Co-maintainers seem like a pretty easy way to specify the
> backup in the current system so that when someone goes MIA or offline
> for whatever reason, anyone with the package will know where to direct
On the other hand, we have many packages such as my liwc, which
rarely changes. I first created the package in 1996, and it has since
had a total of 13 changelog entries in the Debian changelog: three in
1996, one each in 1997, 1998, and 1999, two each in 2000 and 2001, none
in 2002, and a surprising three in 2003 (one NMU, one entry by another
developer while he maintained the package). Some of those changes were
prompted by changes in Debian policy, the C library, or the C language.
I suspect that a co-maintainer would have been bored dead for that
package. Requiring a co-maintainer for each package would almost
certainly be more work than benefit for a number of such very simple
packages. There is some bureaucracy and setting things up for
co-maintainership, and keeping them set up as co-maintainers come and
On the other hand, I'm perfectly happy with a fairly low NMU threshold,
especially for such simple packages. Less effort and the same end
result: if the primary maintainer doesn't do his job, for whatever
reason, the package gets fixed by someone else.
For more complicated packages co-maintainership and such has already
proven its value.
In fact, I wouldn't mind a somewhat lower level of "ownership" of
packages, except for packages that are considered difficult or vital
(say, libc or XFree86).
 I'm on a self-advertising drive tonight. This is the second of my
own packages I've mentioned on the list. Let's assume, for the sake of
the discussion, that the package is actually useful for Debian.