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

Re: Bug#65320: gettext: fails on postinst



Rob Browning <rlb@defaultvalue.org> writes:

> These are the two best options, but I'm *really* hesitant to try and
> guarantee that emacsen-common will work properly before it has been
> configured.  This means second-guessing the whole package
> installation/dependency process and committing to some guarantees that
> might make development (or redesign of emacsen policy) a lot harder
> later on.
>
> How hard would it be for gettext to just provide a gettext-el package
> and have that depend on emacsen as per policy?
>
> I may be over-concerned here, but trying to guarantee that a package
> is functional before it's pre/postinst scripts have run seems like
> asking for trouble.

However... that said, I went back and looked at the logs for bug 48681
(http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=48681&repeatmerged=yes),
and now I'm wondering if the fix suggested there might be sufficient.

The suggested fix is to try and make sure that the
emacsen-package-install script fully configures/initializes
emacsen-common -- i.e. does the same things as the pre/postinst.  Now
that I've thought about it again, I'm beginning to feel like that
might be OK -- any dissenting opinions?

If this were in fact "safe", then we could think about changing
debian-emacs-policy to fully eliminate the requirement that packages
that have emacs components *depend* on emacsen, but instead they just
check for /usr/lib/emacsen-common/emacs-package-install, and if found,
run it.  This is certainly more flexible, but I'm wondering if there
are nasties that I'm overlooking.

Thoughts?

-- 
Rob Browning
rlb @defaultvalue.org, @linuxdevel.com, and @debian.org
Previously @cs.utexas.edu
GPG=1C58 8B2C FB5E 3F64 EA5C  64AE 78FE E5FE F0CB A0AD



Reply to: