Jose' Matos wrote:
On Saturday 10 September 2005 19:23, Peter Flynn wrote:

2. yum install lyx.

Aaaarrrrggghhh! *Never*, never, never do this if you already have TeX.
Especially not if you have already carefully removed the outdated mess
that is the RH kludge of tetex, and replaced it with the real tetex from
the TeX Collection DVD.

If you do that you are on your own.

No, RH is on its own. Posters to c.t.t have consistently told users
of the RH tetex RPMs to trash them and replace them with the TUG CDs.

One other possibility is to redo the tetex rpm and then yum will work.

The TeX community has been trying for years to get whoever is
responsible for the RH tetex RPMs to update them properly. But they insist on meddling with the directories and the subset of features
apparently deliberately to make it inconsistent with the TUG CDs.  I
have no idea why they insist on doing this.

What are the problems you have with FC tetex package?

It was out of date last time I looked. I have consistently told my users never to install it but always to use the TUG CDs instead. For FC4 I didn't even bother looking at it, just ripped it out immediately the OS
was installed (http://silmaril.ie/cgi-bin/blog#fc4).

If it has been updated, then the foregoing does not apply, and I owe the
maintainer an apology.

Have you reported it to bugzilla.redhat.com?

I believe people have tried, but BugZilla is virtually useless: all it
does is provide a talking-shop for the packagers to explain why they
won't change. I have reports and requests in for various pieces of s/w
pending for years, and all the authors do is talk.

One other possibility would be to package that version and replace the require in lyx rpm from tetex to tex...

All that's required is for the maintainer of the tetex RPMs to use up-to-date versions from CTAN, and for the author of the embedded install script in the LyX RPM to test for a working kpsewhich instead
of assuming it's in the location the RH tetex RPMs install it.

Sorry for the OT flak, but I've been supporting TeX for 20 years, and
the inconsistencies of the RH tetex RPMs are the biggest headache we
have.

I suggest we don't pursue this here but move it offline.

///Peter

Reply via email to