Re: dh_installtex and Policy
Ralf Stubner <ralf.stubner@web.de> wrote:
> That doesn't hurt for sure. However, as you noted kpsewhich is used in
> user-specific mode only, while update-updmap should always run in
> system-wide mode when called from maintainer scripts.
I agree. I was a bit too lazy to really check that the kpse* tools were
used *only* in user-specific mode. Howvever, I don't think it makes a
big difference whether to move update-updmap in or out of the kpsewhich
check. The only difference is when kpsewhich reports an error. In that
case, updmap-sys won't be run, so I don't think there is a big advantage
in running update-updmap at this time (sure, it is possible, but is it
useful? It will eventually be run just before updmap-sys, when kpsewhich
is in order).
However, I don't oppose to moving update-updmap again out of the
kpsewhich check if you think it is better (I agree it provides a more
up-to-date updmap.cfg until the "base TeX system is configured", as you
put it).
> It looks as if we are reaching the point where a font package no
> longer needs to depend on tex-common ...
Yes. It seems the only thing missing would be to check in postinst-tex
whether the update-* programs are available before running them.
--
Florent
Reply to: