[tex-live] [LONG] Improving TeX package classification and the associated documentaion

[George N. White III]

On 7/2/07, Joachim Schrod <jschrod at acm.org> wrote:

> Florent Rougon wrote:
> >
> > Well, probably not, but I don't think it's a big problem:
> >   - if package maintainers write their own metadata, well, that's great
>
> Sadly, Your assumption does not fit reality. Even the few authors
> who tried to submit TeX catalogue files got it mostly wrong. For
> many submissions, one even has to ask for READMEs and PDF
> documentation files.

Isn't a README the one file that the instructions require (and where
you meniton the license)?  Actually, having had some experience
with collections of scientific data, I'm not surprised.  It is far better that
stuff get on CTAN in disarray than having people searching net to find
a copy.  If you ask for anything more than the minimum effort, some
potential contributors won't be bothered.

> >   - the most important packages (geometry, graphicx, etc.) will get
> >     tagged in a reasonable timeframe by either their maintainer or by
> >     volonteers (yeah, could be me, but preferably some LaTeX guru :).
>
> I am *very* sceptical that you will find gurus that will help in
> tagging packages. Otherwise the Catalogue would not be maintained
> by just a handful of persons (and Robin does 80% of the work there
> singlehandedly), such gurus would give a hand there, too.

Maintaining the Catalogue requires special insights.  I often provide
advice to LaTeX users -- 90% of the time my advice is "forget what
you found using google and read TLC" (another 9.9% is "that file
you have with the .eps extension isn't an EPS file").  Of course
nobody comes to me after advice they found using google solved
their problem.

I do like the idea of using tags better than the texdoctk fixed structure,
but to succeed tags should be user maintained.  Suppose the user's
initial search fails.  They are passed to a Catalogue search, and
the initial tag becomes a candidate for the results.  The tool gives the
user the option of saving the new tag locally and/or passing their
updated tag database on to an engine that collects suggestions into
an updated database.  Users can also benefit by downloading updated
databases.

Eventually, if this succeeds it can be made part of CTAN, but if
successfu it might spread to other places (R?, IDL) that can benefit
from such tools.

-- 
George N. White III <aa056 at chebucto.ns.ca>
Head of St. Margarets Bay, Nova Scotia