Re: [Axiom-developer] Re: Pamphlet format discussion

[Andrey G. Grozin]

On Sat, 21 Jul 2007, Stephen Wilson wrote:
> address@hidden writes:
> > Would a correct summary of your concern be:
> > Axiom pamphlet format should not be restricted to latex syntax?
>
> No, I dont think so.  The syntax does not bother me in any way.
>
> I cannot see how we can get past the the need for a weave stage in
> general, especially when we consider wide range of possibilities open
> to us in the future.  This is my main concern.
I think it is best not to introduce an additional syntax, when the 
familiar LaTeX syntax is good for describing any structured documents (in 
this sense, I think that xml is a re-invention of the LaTeX wheel). An 
additional advantage is that in the simplest case (which means 100% of 
cases now, but may decrease to 99% in the future) we can directly run 
LaTeX on pamphlet files, thus saving build time. I don't see why 
references to chunks from other files can be a problem - just 
\usepackage{hyperref} and get nice hyperlinks in dvi or pdf. It should be 
also quite easy (in principle) to generate, say, hyperdoc pages from LaTeX 
pamphlets - we can invent LaTeX structures for hyperref-specific purposes, 
such as uses/used-by lists or running axiom commands from hyperdoc. Of 
course, this requires some programming, but quite trivial - LaTeX syntax 
is very simple and regular (unless we play dirty low-level TeX games with 
re-defining character classes, etc.; such tricks should be certainly 
banned in pamphlet files). If some more complicated processing of 
pamphlets is needed, such as extracting some special kind of chunks from 
many pamphlets (for some other kind of documentation in The Crystall :-), 
it can be done by a program. In this sense, weaving will be available when 
we need it (of course, it should be programmed, but this is certainly not 
difficult). But in the majority of cases, weaving will by just the 
identity transformation. I think this is a Good Thing - saving build time 
and the necessity to learn one more syntax.

I'll have a look how \begin{verbatim} ... \end{verbatim} is handled in 
auctex. I am sure this can be generalized in such a way that chunks 
written in some programming language will be syntax-highlighted by an 
appropriate mode. Ideally, we should have the possibliity to have chunks 
written in several languages in the same pamphlet, and to work with them 
comfortably from emacs. Currently, this is not possible; so, I don't think 
that the current setup is purfect and should not be improved.

Andrey