Bug#61308: PROPOSAL] Initializing databases by using conffiles.
On 29-Mar-00, 08:40 (CST), Santiago Vila <sanvila@unex.es> wrote:
> On Wed, 29 Mar 2000, Julian Gilbey wrote:
> > In particular,
> > given the update-mime program, /etc/mailcap should obviously not be a
> > conffile. And as it is, the maintainer should change that -- but
> > changing policy isn't going to make that happen any more quickly.
>
> I also think this proposal should not be needed, but considering that
> nobody managed so far to convince the mime-support maintainer that
> /etc/mailcap being a conffile is really bad, it seems it is. What is
> obvious for you and me may not be obvious for everybody. This is one of
> the reasons policy exist in the first place.
It's already in violation of policy. From section 4.7.3:
"The other way to do it is to via the maintainer scripts. In this case,
the configuration file must not be listed as a conffile and must not be
part of the package distribution."
And from section 4.7.4, describing how to share configuration files:
"If it is desirable for two or more related packages to share a
configuration file and for all of the related packages to be able to
modify that configuration file, then the following should done:
1. have one of the related packages (the "core" package) manage
the configuration file with maintainer scripts as described in the
previous section.
[...]"
If it's to be shared and "multiply-modified" (as /etc/mailcap clearly
is), then it must be managed by maintainer scripts and not be a
conffile.
Also, at present, all the users of update-mime are in policy violation
(from 4.7.4):
"The maintainer scripts should not alter the conffile of any package,
including the one the scripts belong to."
(But submitting a bug against all the "violating" packages is not the
thing to do.)
> If it is obvious for you that this is the right thing to do, then please
> second the proposal. Having something like this written in policy will not
> only help us to convince the maintainer of mime-support that this is
> indeed a bug (currently he has it as a wishlist item), but also will
> help us to prevent similar cases in the future.
No, I oppose this proposal; It's already in policy. If it needs to be
clarified, then the right place to do it is in 4.7.
Steve
--
vmole@swbell.net
Reply to: