> Yes, and separate scripts for xdvi, dvips, kdvi, gdvi, texi2dvi? I > don't think that this is a good approach.
Need it be scripts? Can't you just make a system call with the TEXMF shell variable prepended, i.e. TEXMF=blah <program-and-args> > We explicitly put the version number in, to prevent lilypond from > accidentallyr reading files from old versions, esp. when people update > by hand. Fair enough. > > Perhaps best of all would simply be to install your stuff in TEXMF_LOCAL > > (if it exists) and otherwise TEXMF. I'm happy to help with getting this > > working if you think it's a good idea. > > Is there a standardized procedure for this? I don't know, and it doesn't look like it. tetex is pretty monolithic. In this particular case the problem was that the location (in fact the very existence) of TEXLOCAL changed as a result of going from Fedora packages to ATRPMs packages. But the general principle of having it work across lilypond and tetex upgrades still stands. I think the best solution for the RPM is to store files directly in /usr/share/texmf, and for user installs in a TEXLOCAL or TEXHOME tree. lilypond.sh could remain as an option for those who wanted to install everything under a lilypond directory. -- http://www.mupsych.org/~rrt/ | free, a. already paid for (Peyton Jones) _______________________________________________ Bug-lilypond mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-lilypond
