On Sat, Dec 28, 2013 at 04:01:16PM -0800, Robin Lee Powell wrote: > > Given that major web-browsers nowadays open pdf just like any html > > file through additional plugins and hyperlinks can be used in pdfs > > as well, what exactly are the reasons for generating HTML in the > > first place? > > PDF viewers are, almost without exception, steaming piles of crap? So are browsers. > Mostly it's just that a book-format PDF has completely different > user needs from a web page. It's broken up by physical pages, which > is unnatural for the web, it doesn't wrap to fit the browser, etc. > > It's certainly a possible fallback, but it's pretty weak IMO. IMO that's nitpicking in this context. Yes, there are differences between HTML and PDF. But the CLL is ba'e the official document for lojban and I do not see why it has to be available in different formats to begin with. With multiple output formats you will always end up with situations like "hm, I could need some additional space in the pdf here" or "Argh, I hit a special case using this environment and now it looks awkward in this format". As far as I can see this is quite a problem at the moment. If you then introduce lots of this-format-only tags into your sources, you're basically back to your two-fork mess imo. > > The main problem you're trying to solve the whole time is > > generating HTML and PDF from one source. Granted, that's > > non-trivial, though possible. > > Quite. The question is, how would I need to be doing it to get your help? docbook and xslt is a pretty confusing mess imho. If the CLL really uses only few features of this combination, it would really help quite a bit to have the whole conversion process in one ``simple'' module written for this specific translation task exactly. However, we might again end up with something very confusing. v4hn
Attachment:
pgppGYZc26l6t.pgp
Description: PGP signature