would I be correct in assuming that sphinx's inability to find anything vaguely resembling the epub spec is well known and no one really wants a bug report about it?
brucem
I suspect that you could find a more constructive way to ask whatever your question is.
I use Sphinx to generate ePubs. I know they aren't great, but they're readable, apart from iBooks ignoring font settings in them. (Which I've previously reported a year or more ago.)
Hasimir
iBooks is in its own special world though
nah, this is just generating the Flask docs and getting nearly 400 errors from the validator
when the number of errors goes past the number of files, it's more than "a little bit broken"
brucem
I've never run the validator, I guess.
Hasimir
I run it all the time, but I usually don't generate epubs with sphinx, so it caught me by surprise (I use something else because of epub 3 support and I already knew sphinx didn't have that)
brucem
I'm sure patches would be welcome ... perhaps bug reports, but perhaps the issues are already known. (I have no idea.)
Hasimir
yeah, I might leave the question open here for a while first see if an official(-ish) "thanks, but we know already and maybe one day when we have time" or a "eep, really, lodge it" response - that's what idling in irc is for
obviously I'm not too fussed personally since for my epub generating I've got D4P with DITA-OT and it works almost entirely perfectly (it still has trouble getting XSL to always use UTC instead of localtime, but that's it and the validator won't pick it up anyway)
brucem
the project that I'm working on bailed from DITA and Docbook in favor of Sphinx some years ago for better or worse.
Hasimir
heh
they're the only things I've seen that take even a half-way reasonable crack at epub 3
brucem
Hasimir: Well, we had years of making no real headway ... and then in a matter of about 6 months, I converted thousands of pages of technical docs and api docs over to Sphinx.
Hasimir
ah
yeah, that'll get people's attention
ljosberinn joined the channel
personally what I'd like to see (from someone who actually understands xslt fluently; i.e. not me) is 2 xslt generic-ish transformations one for docutils xml to dita topic and one the reverse
palmergeraint joined the channel
between that and what pandoc already has, I think a lot of people would be very happy (the dita crowd would get markdown via the long way around and the foss world gets all those webhelp, epub and other transformations)
brucem
Hasimir: We have a Sphinx extension that allows us to mark up our API more completely. Never got anything like that to work well in any markdown system ... or via pandoc.
Hasimir
oh, I know reST is better, I prefer it too
it's just a common whinge on the dita-users list
and that path would be scriptable enough to keep them happy