Re: How to use update-updmap safely
frank@kuesterei.ch (Frank Küster) wrote:
> By the way: Why do we not run updmap automatically from update-upmap? I
> doubt there's any sensible use of update-upmap without a subsequent call
> of updmap, is there?
Good question. I think there would be if dpkg supported to queue actions
for later, unique execution (i.e., it would be able to run
update-updmap, mktexlsr, and updmap only once after all packages
installing or updating maps are (almost-)installed...), but this is not
the case at the moment.
However, if you make update-upmap call updmap, ideally, you want
update-upmap to support all updmap options (such as --nohash). This
would make update-upmap a bit longer than it is now.
Another nice solution would be that the upstream updmap directly uses
/etc/texmd/updmap.d and uses updmap.cfg or its equivalent in /tmp only
as a temporary, internal file. There would be no need for update-upmap
anymore. I am not sure TE would appreciate this solution, though (but it
would allow tetex users to separate their locally added maps from the
default tetex maps, which is not so bad from my point of view).
The last solution I see would be to have a Debian teTeX Policy that
states the right way to use Debian teTeX tools. Or as a workaround, a
low-volume mailing-list to which the maintainers of all TeX-related
packages would have to subscribe. _And_ listen, _and_ act accordingly. I
think the Policy idea is better, 'cause then, you have a real document
to point them to when their package does things wrong.
--
Florent
Reply to: