Eric Armstrong wrote on 2003-04-10:
> Hello. I'm not (yet) a member of this list, so please
> reply to all.
> The markup standard looks amazingly complete, but I
> do have few questions:
> 1. Is the preferred abbreviation rst or reST?
This I wanted to know too :-). David?
> 2. Is there an rst to HTML filter program out there
Sure, ``docutils/tools/html.py``. The sandbox also has all kinds of
other writers - to LaTeX, PDF, docbook and what not - in various
degrees of completion.
> 3. I'm inclined to think that Mozilla got text formatting
> right, with *foo bar* in bold, /foo bar/ in italics (but
> /foo/bar is recognized as a path), and _foo bar_
> for underlining. Is there some overwhelming reason why
> ** was chosen for bold?
Mozilla went with WYSIWYG approach here. re?ST tries to be a logical
format, so it just has *emphasis* and **strong emphasis**. It so
happens that italics are the preferred emphasis style according to
typographic wisdom, bold coming second. Underlining is especially
frowned upon - you'll almost never see it in a book...
These mappings are writer-dependent and can be tweaked in most cases,
for example you can provide your own stylesheet to the html writer...
> (One thing I liked about my intitial introduction to the
> standard was the way it codified existing practice. But
> I've never see ** used in an email anywhere.)
Me neither - but emails don't have any accepted practice of normal vs.
strong emphasis so there was nothing to choose from. Every one uses
other ranking of the possible "decorations" of text, some creative and
amusing but only appropriate for ascii-only mediums like email...
> 4. The "quick reference" at
> contains a huge number of markyp devices (which leads me
> to believe it must be complete).
No markup standard can ever be complete ;-). It's pretty close
however and has some extension hooks: directives are very easy to add
and easy role extensibility is now being discussed.
> But the "primer" at
> has a much more approachable subset that is easily grasped.
> Is there perhaps a "subset standard" like the one described
> in the tutorial, and perhaps a filter for that?
There was some disccussion on this recently; the problem is that any
such subset will be quickly outgrown by one's need - and what does he
do then? All re?ST constructs were added when someone needed them
(and some needs were left unanswered still). What do you want to gain
by defining a subset?
Beni Cherniavsky <cben@...>