[tex-live] Suggestion for future releases

[Frank Küster]

Robin Fairbairns <Robin.Fairbairns at cl.cam.ac.uk> wrote:

> Karl Berry <karl at freefriends.org> wrote:
>
>>     Would it be possible to put in some verification in the Catalogue
>> 
>> Well, the CTAN folks maintain the Catalogue.  I am not sure what the
>> story is on the file: url's, how they are determined, who might be using
>> them, etc.
>> 
>>     Since the catalogue is quite often what you get as output from
>>     "texdoc"
>> 
>> Interesting.
>> 
>> As it happens, I removed the catalogue package from TL development a few
>> days ago, since it seemed to serve no purpose and is always out of date.
>> It seems like overkill to have the entire Catalogue just for these
>> references, especially when you have found that they are typically
>> wrong.

The reason why they are wrong is that they contain links like 

title="Read documentation for this package from ctan:/macros/latex/contrib/a0poster/a0_eng.pdf."></a>
  <a href="../../../macros/latex/contrib/a0poster/a0.dvi"><img

with relative links, but of course the directory structure in a TEXMF
tree is different than on CTAN.  On the other hand, it shouldn't be too
hard to automatically fix the link.

I didn't object when Karl removed the catalogue, but I didn't really
agree, and I would be in favor of keeping/adding it.

> i agree the revised script is the way to go, with script exceptions
> rather than texmf exceptions.
>
> how would such a change interact with texdoctk (which i hardly ever
> use)?  i started thinking about a replacement a while back, and i could
> be interested in a small development project (but i know neither tcl nor
> tk).

I don't think it would need to interact with texdoctk at all, nor have
unwanted side effects.  texdoctk.dat contains relative paths starting
in $TEXMF/doc/, and looks in all available TEXMF trees.

Regards, Frank
-- 
Dr. Frank Küster
Single Molecule Spectroscopy, Protein Folding @ Inst. f. Biochemie, Univ. Zürich
Debian Developer (teTeX/TeXLive)