RFU: tmexpand

[Rafael Laboissiere]

* Jörg Sommer <joerg at alea.gnuu.de> [2007-05-28 14:38]:

> Rafael Laboissiere schrieb am Mon 28. May, 13:47 (+0200):
> > 1) The package depends on jed | xjed but the future jed will build-depend on
> >    it.  Won't this cause a chicken-and-egg problem?
> 
> Good point. It's possible to build the packages, but bootstrapping isn't
> possible. But rebuilding the hlp files if tmexpand is installed and
> otherwise leave them as they are, wasn't what I want and leads to
> different build results.

It looks like tmexpand should be part of the jed source package.  In that
case, the docs could be built using the freshly built jed-script which is in
the build tree.

Another possibility would be to split the jed documentation from the jed
tarball and creat a jed-doc package which would build-depend on jed and
tmexpand.

Anyway, the chicken-and-egg problem is pretty unacceptable in Debina.

> I revert the tmexpand change, add a dpatch and check for changes in
> check_package.

This is something I forgot to ask you last time:  What are the scripts
check_package (in jed) and check_installation (in tmexpand)?  Shouldn't they
be in the debian/ directory?

> > 2) We should address the following Lintian warning:
> > 
> >    W: tmexpand: unusual-interpreter ./usr/bin/tmexpand #!jed-script
> 
> That's only possible with an override file. Mmm, yes it is the best. It's
> better than leave this message or adding jed-script to lintian.

What is the problem in asking the Lintian maintainers to include jed-script
into the list of known interpreters?  There are already lots of interpreters
listed in /usr/share/lintian/checks/scripts and one more won't hurt.

-- 
Rafael