Re: dh_python and python policy analysis
On Thu, Aug 24, 2006 at 11:46:25PM +0200, Josselin Mouette wrote:
> Let me rephrase it: the internals of python-support, and how it helps
> implementing the python policy, are developed in the python-support
> documentation. They don't need to be part of the policy
Yes they do.
> and they have nothing to do with debhelper either.
That's good, but not enough.
> > This has now been going on for long enough that I conclude that the
> > Python policy pushers really do intend to make using debhelper a Policy
> > requirement for any package containing any Python code.
> I can't speak for others, but python-support provides
> pysupport-movemodules and pysupport-parseversions to separate the
> debhelper snippet from the actual abstraction code.
That is still not what is required. Unless these tools become part of
the dpkg-dev package, it should be documented in policy how they (are
supposed to) do their job.
Sure, you're free to seriously discourage people not to use these tools;
but the assumption that it is possible to create a Debian package by
just copying the right files to the right place and calling 'dpkg -b' on
that has always been true. It would be wrong to change this assumption.
> (BTW, for a similar problematic that involves more than a hundred
> packages, nobody ever asked me how to make a package using GConf without
> using dh_gconf. Which means the GConf policy has never been written out
> but is currently defined by the dh_gconf behavior.)
The mere fact that a given bug exists somewhere else, too, does not make
it less of a bug.
<Lo-lan-do> Home is where you have to wash the dishes.
-- #debian-devel, Freenode, 2004-09-22