+++ Aristotle Pagaltzis [Jul 24 10 20:14 ]: > * John MacFarlane <j...@berkeley.edu> [2010-07-23 17:50]: > > But using HTML isn't a great solution for pandoc, which > > supports many output formats. > > IMO separating Markdown from HTML is the wrong approach. Likewise > everyone who filters out `<em>` but allows asterisks for emphasis > is doing it wrong. > > Markdown is an alternative syntax for HTML. The right thing to > do with a Markdown document is convert it to regular HTML, *then* > optionally apply a tag white list, *then* optionally convert this > HTML (subset) document to other formats. If there are HTML tags > you cannot (easily) support in your target formats, just omit > them from your supported subset of HTML. That way you support > real Markdown, not a crippled Markdown-lookalike syntax.
There are two issue here, really. 1) Should `<em>hi</em>` and `*hi*` be parsed to the same intermediate representation? 2) Should this representation be HTML? I think the answer to 1 is probably yes. But I don't see any strong reason to answer yes to 2. Personally, I'd rather have a nice Haskell data structure I can manipulate simply and convert to any format I like. And I want to represent some things that don't have a real HTML counterpart (footnotes, mathematical formulas, inline quotations). Anyway, my original point was not that pandoc *couldn't* have used the syntax `<s>hi</s>` for strikeout, no matter what the output format. It could, of course. The point was that it might seem odd and unnatural for people who are primarily writing for other output formats to have to use HTML in their documents. John _______________________________________________ Markdown-Discuss mailing list Markdown-Discuss@six.pairlist.net http://six.pairlist.net/mailman/listinfo/markdown-discuss