User: Password:
Subscribe / Log in / New account

epub shortcomings, mathematics

epub shortcomings, mathematics

Posted Sep 2, 2011 9:00 UTC (Fri) by pjm (subscriber, #2080)
In reply to: epub shortcomings, mathematics by n8willis
Parent article: Ebook editing with Sigil

I think you'd agree that "[MathML] is designed as a presentation markup language" at least needs qualifying, given the content/semantics part of MathML and Mathematica's authors' substantial involvement in its development.

To anyone looking at adding MathML support to an editor like Sigil, I'd first suggest having a look at to see what the options are. There are a few approaches one could take for an initial implementation, from using human-editable ASCII representations à la TeX for simple uses (mentioning x² and the like in body text) to launching an external application for more complex uses. The external application might even be something that can only export to MathML rather than editing the MathML directly (so long as Sigil keeps track of what the input file is). Granted, these simple approaches are nothing if not "unoptimized"; but they're a start, and using TeX-like input is no worse than what academics have been doing for a few decades.

(Log in to post comments)

epub shortcomings, mathematics

Posted Sep 2, 2011 14:29 UTC (Fri) by n8willis (subscriber, #43041) [Link]

Yes, I mean Presentation MathML (I don't think Content MathML is applicable in the ebook context, since it's designed as a medium for human reading). We did a piece about the whole can of beans back in April: .... Unfortunately (though probably predictably) the comment thread eventually devolved into an argument over the abstract virtues of different approaches.

If you ask me, ultimately writing mathematics is always going to involve ambiguity -- just like writing language -- thanks to context and the brain, so debates about devising the "perfect" form of content markup are similarly moot.

Another question entirely is whether a text-centric app like Sigil is the right application to do something as layout-driven as a scientific work. I'd probably use Scribus for that; it already supports "render frames" for content in other formats (TeX included, as is gnuplot, lilypad, and a few others).


Copyright © 2018, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds