[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#753620: wishlist: idl/gdl-written software



* Оlе Ѕtrеісhеr [2014-07-04 14:45:14 +0200]:
> Sylwester Arabas <sarabas@igf.fuw.edu.pl> writes:
> > - how to inform GDL package about their location?
> 
> No idea.

One would want to augment the !PATH variable within GDL. This can be done
in a startup script or by the user setting an environment variable. I'm
not sure it's desirable for the packages to unconditionally add themselves
to the !PATH (some of them may be too specialised and/or interfere with
one another or with users' own code), but that could be under debconf
control. I don't think GDL currently looks for startup file snippets
under /etc but that should be an easy feature to add, if desired.

> > - how to make these packages usable for IDL / PV-WAVE users?
> 
> Since IDL is non-free, I would consider this not as a primary goal. I
> would concentrate on GDL.

Still, I for one would like them to be usable with proprietary IDL as well.

I certainly do not expect the Debian maintainers to develop workarounds for
bugs in IDL that are not present in GDL; that sort of thing can be left to
the upstream developers. "Concentrate on GDL" sounds right in that sense.

> > - should idlastro become a dependency of gdl?
> 
> GDL should be a requirement of the package(s). 

I'd vote for a simple Recommends:, to cater for those who want to use the
packages in some other way (e.g., in conjunction with proprietary IDL).

> Again: priority should be
> the use with GDL -- and only packages that work with GDL can go into
> Debian (main); everything else would have to go into contrib.

And can be moved to main as GDL's feature coverage catches up. Yet one
could argue, given GDL's goal of compatibility with IDL, that any
unsupported feature is to be treated as a bug (usually in GDL) rather than
as an argument for relegating the package to contrib. Should a test suite
for a package in main have to be in contrib just because some tests
are still failing?


Reply to: