Re: Mailutils new competitor

[Sam Roberts]

Quoting Jeff Bailey <address@hidden>, who wrote:
> On Thu, Feb 15, 2001 at 09:23:50PM -0500, Sam Roberts wrote:
> 
> > I'm a huge docbook fan, we converted all of our docs at the
> > last company I worked at into it (though not auto-generated)
> > and we could produce great ps,html,QNX helpviewer, and pdf,
> > which pretty much dwarfed the capabilities of 10,000 dollar
> > toolkits being used by all kinds of huge companies. Very
> > satisifying. My only gripe was the lack of .info format, and
> > then just before I left I found a reference to somebody
> > have written some kind of processor to generate that, but
> > I never got a chance to give it a try.
> 
> I've looked at docbook a couple of times.  I don't know if I'm just 
> looking in the wrong places, but there don't seem to be many tools to use 
> the files when you're done.  Like a Free docbook->postscript converter.

Sgml suffers a bunch from the unix open source toolset "problem",
there isn't one tool, its a bunch, and there's lots of different
ways to do things, and each piece is documented, but it can be
hard to find docs on how to actually write docs and get output,
because it involves lots of different tools, and people don't all
do things the same way. There is an open source toolchain, though. I
set the first version of a toolset running on Linux up, and then
our documentation guy (Bob McIlvride) ran with it, including
documenting the thing, here's his overview picture:
  
  http://24.112.178.91/cogent/prepdoc/pd-tools.html

Apparently doxygen still can't generate sgml, oh well! On the todo
is breaking it into a front-end that parses and generates xml, then
backends that convert the xml to various formats. That would be sweet.

> Doesn't GCC use yacc for it's parsing? It might be easiest to start with 
> what they have and remove the pieces that aren't needed.

Maybe, I spent a fair while looking for a grammer for C++, and didn't
find a good one, but C is way easier, and it never occurred to me to
look in the gcc source. Good luck!

Also, for a specific project, a perl script to rip documentation out
of the headers and into texinfo isn't out of the question.

I'm just musing, is there really a problem to be solved?

Sam

-- 
Sam Roberts <address@hidden> (http://www.emyr.net/Sam)