Re: [Bug-gnulib] module maintainers

[Paul Eggert]

Bruno Haible <address@hidden> writes:

> The idea is that you cannot make a change to a [gnulib] module if
> one of its maintainers doesn't agree.

I'd rather not make this change to gnulib's informal rules, as it adds
an extra level of bureaucracy with little benefit and some harm.

If I need a minor change installed because (say) Bison doesn't port to
Solaris 10 without some minor tweaking, I don't want to have to wait
for a module maintainer's official approval in order to install the
obvious tweak.  Sure, I can wait and fork versions in the meantime;
but it's one more thing to worry about and we already have too many
things to worry about.

Part of the charm of gnulib is that it's just us: it's not a bureaucracy.
Let's keep it that way.


> I would like to avoid incidents like the one
> in http://mail.gnu.org/archive/html/bison-patches/2003-01/msg00042.html.

Bison is maintained differently from gnulib, and I don't think you
need to worry about incidents like that happening here.  Unnecessary
and even controversial changes occur relatively often with Bison; they
are installed by both Akim and myself, and sometimes countermand each
other.  Despite what may seem to be chaos to outsiders, Akim and I
continue to think quite highly of each other, and we both see each
other as essential contributors.  (He's more essential than I am,
truth be told.)


> I would like to ask that Paul doesn't make unnecessary changes

That's fine with me.  We both have better things to do than worry
about unnecessary changes to gnulib.