On Sun, Dec 12, 2021 at 12:32:35PM +0100, Julien Puydt wrote:
> Le dimanche 12 décembre 2021 à 12:15 +0100, Bill Allombert a écrit :
> > On Sun, Dec 12, 2021 at 11:39:19AM +0100, Julien Puydt wrote:
> > > Package: pari-doc
> > > Version: 2.13.3-1
> > > Severity: wishlist
> > > 
> > > The pari-doc package only ships documentation in tex, dvi and pdf.
> > > 
> > > For the giac package, I would need some html documentation too,
> > > like
> > > what can be found in:
> > > https://pari.math.u-bordeaux.fr/dochtml/html-stable/
> > 
> > Hello Julien,
> > 
> > The script used to generate them is not provided by the source
> > tarball and is
> > not particularly robust nor up-to-date with current web standard (it
> > use
> > html frames).
> > 
> > Could you elaborate why you need them ?
> 
> The upstream src:giac package ships a doc/pari with that documentation,
> and clicking in xcas' help menu for "PARI/GP documentation" opens it in
> a browser.

So fonctionaly, you only need a single URL, not the whole set of URL 
provided by https://pari.math.u-bordeaux.fr/dochtml/html-stable/ ?

This is important, because it is possible to convert PARI documentation
to HTML instead of PDF (with pod2html) but it will not look like
https://pari.math.u-bordeaux.fr/dochtml/html-stable/
at all... In particular it is missing an index.html page.

> I noticed that Debian's src:giac does not ship it (because it's an
> embedding), hence the link is dead.
> 
> My plan is to depend on pari-doc and link to what's there. In the mean
> time, I'm linking to pari's homepage (a privacy breach, but better than
> a broken link).

I suggest you write a simple HTML document which lists the links to the PDF 
files in
file:///usr/share/doc/pari-doc and then patches giac to open this URL.
(or use doc-base to do it for you, after all, all the pari documents are 
registered with
doc-base).

If you feel this to be inadequate for, then I can consider shipping the
pod2html-generated fils in the packages.

Cheers,
-- 
Bill. <ballo...@debian.org>

Imagine a large red swirl here. 

Reply via email to