Steve Langasek <vorlon <at> debian.org> writes:
> But again, as Thomas points out, you can address this use case with much
> less per-package effort. A centralized ten-line shell script would be
> enough to locate all the installed packages on the system that ship files
> under /usr/share/doc not matching the pattern
> {copyright,NEWS*,changelog*,README*} and grab the short description of the
> responsible package. If you want to be really clever, some introspection of
> html documents could give you document titles. Then you only need to worry
> about the minority of cases where autodetection fails.
>
PDFs have titles too, and they can't be snarfed in any way I know of.
Also, doc-base has another benefit that you're throwing under the bus, namely
the subject classification hierarchy. Of course it is not perfect, there's
overlap among the topics, but so what. Still much better than plain linear
search which is what you're suggesting.
So what I'm gonna do, in chronological order:
1. Scratch my itch, which as I can now see will have to happen without help from
the distribution.
2. Look into debhelper and try to come up with a patch for auto-snarfing, then
file a debhelper bug as Zack suggests.
3. If by then the coming global energy crisis hasn't arrived yet and I have a
working computer, I'll fix the list of offending packages.
Back to my cave. OH and btw, given the state of affairs revealed by this
thread, should doc-base really be in policy?
> > PS.: I am not subscribed to the list, so please Cc me on replies. Is
> > Mail-Copies-To another good idea that has fallen by the wayside?
>
> I think you want Mail-Followup-To, which is actually implemented by a fair
> number of mailers.
>
Yes, thanks. Been in the cave too long :-)
--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/loom.20100810t194233-...@post.gmane.org
