Python documentation in DocBook

[Fernando Pérez]

Ken Starks wrote:

> So we need:
>   1. LaTeX for authors
>   2. XML for the machine
>   3. Variety for the readers
> 
> To me this boils down to a two-way conversion system, with
> a 'there-and-back' circular function that preserves all
> content, and turns rapidly into a constant loop.

[snip]

> If this works, then it doesn't matter a damn which of
> Latex and xml is the 'official' one.

Good luck. Writing such a tex<->xml converter is an insanely non-trivial task, 
and I know a few bright people who have tackled it. If you do succeed, 
_please_ let the Lyx team know about it. There's interest in using XML in 
Lyx, but one of the hurdles is that the Lyx format, while not exactly latex, 
is reasonably tied to it. Latex is very hard to parse correctly, from what 
I've seen. Considering it's basically a full-blown programming language, 
short of rewriting TeX itself, this won't be a simple task...

A nice starter project (which a friend of mine is working on) is a tex->MathML 
parser. Please drop a line if you get such a gizmo going, I'd love to see it.

Cheers,

f.