David Carlisle skrev:
> 
>> On 11/11/2009, at 2:17 AM, Chris Rowley wrote:
>> 
>>> (Trying to get a team going to implement (AFAP) Presentation
>>> MathML3 via LaTeX -- Standard or expl3 or anything!! -- to pdf
>>> for screen and paper rendering.)

I'm glad I don't have to do that. MathML is scary (and I mean that as a 
programmer, not as an author; all those attributes and stuff) ...

>> I don't know what AFAP means (oh, ASAP I guess) but considering Hans  
>> has had MathML2 rendering in ConTeXt for years now I figure he'd  
>> probably have MML3 done before we decide what prefix to give our  
>> internal variables in the module. Or do you need a LaTeX intermediary  
>> for some other reason as well?
> 
> 
> hmm, but perhaps some of us have a head start:-) see 2nd column of
> 
> http://www.w3.org/Math/testsuite/results/tests.html#results

(Wow! Googling for pmml2tex finds that page *only*.)

I'm a bit curious, though: The goals for pmml2tex is stated as 
"rendering via conversion via OpenMath to Presentation mathml and TeX".

Having dabbled a bit with implementing OpenMath->LaTeX conversion 
myself last year (on the reasoning that it'll be faster in the long run 
than hand-converting a large collection of computer-derived formulae; 
in order to write a paper reporting the results, they somehow need to 
be turned into LaTeX), I did consider going via MathML, but eventually 
decided not to; MathML felt too messy to be a convenient intermediate 
format. (Still, I'll probably also want to do OM->PMML at some point in 
the future.) So is PMML in your mind an optimal intermediate when going 
OM->TeX, or is it more that you get it for free from composing OM->PMML 
and PMML->TeX conversions?

(I should perhaps add that I considered sensible formatting of the 
generated LaTeX code to be important; since it's going into a document 
that I'll do editing of, I don't want it to be a compact block of 
commands without any hint of the logical structure of the formula.)

Lars Hellström