I'm currently coming close to releasing an alpha version of a
pair of new module I have written.

TeX::Kpathsea will provide a Perl interface to Karl Berry & co's
Kpathsea(rch) library, which is used for file searching in the
Web2c/texk implementation of TeX and friends.  

TeX::Kpsewhich will provide a Perl version of the kpsewhich using
the TeX::Kpathsea library.  This will be far more efficient than
calling kpsewhich itself multiple times, as it will not require
loading the ls-R database into memory multiple times.

I have also reimplemented the mktex* scripts in Perl using these
modules.  They run 2-3 times as fast as the shell script versions;
not a huge improvement, but ample.

My intention is to enhance the Perl versions of the mktex* scripts
to support an improved security version (where the scripts will be
setuid 'tex', and the font trees will be owned by TeX).

Some questions.  Is this appropriate for CPAN in addition to CTAN
(Comprehensive TeX Archive Network)?  Is my choice of module names
acceptable?  Is the world round?  Anything else I should know which
is not included in the module writer's guide on CPAN?

Thanks for any advice!

   Julian

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

  Julian Gilbey, Dept of Maths, QMW, Univ. of London.

        Debian GNU/Linux Developer,  see http://www.*-*-*.com/ ~jdg