This is the mail archive of the guile@cygnus.com mailing list for the guile project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: SCWM's embedded docs/text proc benchmarks/perl's 10x faster than guile.


"Perry E. Metzger" <perry@piermont.com> writes:

 > Harvey J. Stein writes:
 > > "Perry E. Metzger" <perry@piermont.com> writes:
 > > 
 > >  > Harvey J. Stein writes:
 > >  > >  1) There's been some talk of using a similar documentation system for
 > >  > >     guile, but I'd assume you'd need a guile version of the
 > >  > >     extractor.  Jim's also said that they need to generate info pages,
 > >  > >     not sgml.  It should be pretty easy to modify my version to also
 > >  > >     output texinfo or info.
 > >  > 
 > >  > Better to stick to the Docbook DTD in my opinion, and produce a more
 > >  > general tool to create TeXinfo from the SGML. There is a reason lots
 > >  > of people are moving towards using the Docbook based tools.
 > > 
 > > So you're basically suggesting that we do without info & tex output
 > > until someone writes docbook->info & docbook->tex drivers, even though
 > > it'd be pretty trivial to output texinfo directly from the extractor.
 > 
 > Yes. That is because the issue isn't the extractor -- it is the pain
 > and suffering associated with converting the doc strings themselves,
 > which should be in docbook over the long haul. Best to do the right
 > thing now, which will (at the very least) pressure the rest of the
 > correct solution.

Firstly, I dissagree with the pain & suffering part.  I think that
people won't (and they shouldn't) put much markup into the comments
themselves.  And, I think that converting the fairly minimal comment
markup from one markup language to another would actually be fairly
minimal work, be it by hand or automatically.

Secondly, there's input & there's output.  I never said that the
comment strings should have texi markup in them.  Your argument
applies to the comments in the source, not to whether the extractor
outputs texi or only sgml.  I think that whether the extraction &
conversion to texi/tex/sgml/... is done by the extractor or some other
combination of tools is irrelevant.

-- 
Harvey J. Stein
BFM Financial Research
hjstein@bfr.co.il