On Tue, 2008-10-28 at 09:23 +0100, Hans Dockter wrote: [ . . . ] > I don't feel pain with the current way we generate the website. But > if it grows more complex things might change.
Complex and difficult for LaTeX is not the same set as for DocBook/XML and both sets are far from empty :-( > One unique selling point of LaTeX is that it is a true type setting > system. It calculates the position for every letter it writes, based > on algorithms containing many of the rules of the typesetting craft. > You can publish a book with LaTeX. I doubt that the apache FO can > compete here. This is the reason why most of the modern mark up > languages have a transform to LaTeX code. But there is a docbook to > latex converter (as well as latex to docbook converter). Apache FOP is no match for LaTeX hence why so many XML toolchains convert to LaTeX and use it. Some publishers even go this route, though one or two have XML-FO processing toolchains that avoid LaTeX -- but they cost large amounts of money. Of course the real question is which is the primary documentation set and how is that most easily produced and maintained. If PDF is the main thing then LaTeX is the winner. If website is the main thing then an XML/XSTL/XML-FO based system might be viable, though I would probably still go with LaTeX because of the PDF aspects. > > More importantly, the documentation describes how to do the > > customisations, which cannot be said for the atrocious tex4ht > > documentation. Was latex2html ever considered? The tex4t website doesn't do well on the documentation front does it. > If it weren't for you, we would still have the crappy html user's > guide, because it was not obvious to me how to configure tex4ht. So I > see your point ;) Of course now that the main trick to configuring tex4ht is sorted out, the rest is CSS programming, so now even more can be done to make the site look cool. Even with LaTeX. :-) > I'm not wedded to LaTeX. For me it was an obvious choice as I knew it > and it did the job (except easy html transformation). But your points > are very valid. > > The question is which alternative to choose. I'm not an expert here > at all. One thing I have come across lately is, that the Python guys > have switched from LaTeX recently to Sphinx (they still use a LaTeX > transform to produce the PDF). See: http://sphinx.pocoo.org/. But is > has the disadvantage of not having a Java tool chain (unless it works > with Jython). This is yet another text markup language. I guess Python got Perl envy and wanted to copy the POD way of working. I agree there are benefits to something simple -- if what you are working with is a wiki or simple website content management system. However for high quality documentation with partial code inclusion etc. aren't you left with LaTeX as the only viable tool, and having to hack with DocBook/XML to get anywhere close. I suspect Sphinx will run out of capability and become a pain, even compared with DocBook/XML and especially compared to LaTeX. -- Russel. ==================================================== Dr Russel Winder Partner Concertant LLP t: +44 20 7585 2200, +44 20 7193 9203 41 Buckmaster Road, f: +44 8700 516 084 London SW11 1EN, UK. m: +44 7770 465 077
signature.asc
Description: This is a digitally signed message part
