[tex-live] teTeX: no next release

Robin Fairbairns Robin.Fairbairns at cl.cam.ac.uk
Sat Jun 3 11:49:00 CEST 2006


> Robin Fairbairns <Robin.Fairbairns at cl.cam.ac.uk> writes:
> 
> >> >>>>> "Karl" == Karl Berry <karl at freefriends.org> writes:
> >
> > the problem is precisely that such a script has to have built-in
> > knowledge about the package and how its docs are supposed to be
> > processed.  there are a significant number of packages on ctan that
> > just won't work in (what i think of as) the standard way -- process
> > .ins, process .dtx -- enough that every so often i'm surprised by
> > another twist.
> 
> That's not the standard way.  The standard way is process .ins,
> process .drv.  Only if there is no .drv generated, is .dtx worth a
> try.

ah.  the .drv style is in *so* much of a minority in latex2e packages,
it never occurred to me to suggest it.  it's presumable a pre-2e
notation?  (i never used the pre-2e doc stuff.)

every 2e package i've created follows the style used in the latex
kernel, which i had simply assumed was in some way "standard".

however, your alternative technique has the advantage of being
programmable.  dealing with packages with no .ins file (you're
expected to run docstrip direct and guess what options to type) isn't
(afaict) programmable.


More information about the tex-live mailing list