Bug#533287: debian-policy: please clarify 10.7.4

[Sune Vuorela <reportbug@pusling.com>]

On Tuesday 16 June 2009 20:05:40 Russ Allbery wrote:
> severity 533287 wishlist
> thanks
>
> Sune Vuorela <reportbug@pusling.com> writes:
> > There has recently on #debian-devel been a few discussions about
> > wether it was allowed to edit other packages configuration files (not
> > 'conffiles') in maintainer scripts.
> >
> > My interpretation of policy is that you are only allowed to edit other
> > packages configure files thru a specific provided interface
> > (e.g. update-inetd), and if a package don't offer such a interface,
> > you aren't allowed.
> >
> > Some people seems to claim that you are allowed to do
> > sed/cat/echo/perl magic on other packages configuration files in
> > maintainerscripts if you really need it and there is no provided
> > interface.
> >
> > I would like to get the wording clarified so that it is either fully
> > permitted or, in my preference, clearly not allowed.
>
> What isn't clear about Policy right now?

To be honest, I also have a hard time seeing the issues, but after 
participating two days in a row debating this with different people in 
#debian-devel, I just thought that something must be unclear. I asked a bit, 
and a specific "should" the primary blame.

>     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 be done:
That should over there ------------------------------->   ^^^^

>     * One of the related packages (the "owning" package) will manage the
>       configuration file with maintainer scripts as described in the
>       previous section.
>
>     * The owning package should also provide a program that the other
>       packages may use to modify the configuration file.
>
>     * The related packages must use the provided program to make any
>       desired modifications to the configuration file. They should
>       either depend on the core package to guarantee that the
>       configuration modifier program is available or accept gracefully
>       that they cannot modify the configuration file if it is not. (This
>       is in addition to the fact that the configuration file may not
>       even be present in the latter scenario.)

so it seems that the "alternative" interpretation, is that "if there is a 
interface, then it must be used", but all that is wrapped in a "should", which 
is not as binding as a "must".

> I'm not seeing how that could be made less ambiguous.  It seems to me
> that your position is clearly correct and Policy spells that out rather
> explicitly.

I will try get some of the others to add their comments here.

/Sune