On 02/28/2014 11:07 AM, stefano franchi wrote:
On Fri, Feb 28, 2014 at 9:56 AM, Jürgen Spitzmüller <sp...@lyx.org> wrote:
Richard Heck wrote:
Here, I think it's especially helpful to distinguish "round trip"
conversion, which would be used during collaboration, from final export
for publication. I take it that in the former case we just need to make
sure not to lose bibliographic information. Only in the latter case do
we need to be able to use or mimic biblatex, or whatever, to get the
bibliographic information into some final form.
Yes, maybe. But isn't the real and more serious problem, at least in the
humanities, that you are forced to submit papers and chapters in Word format,
notwithstanding if this is collaboration work or an individual's work?
The possibility to export and re-import strikes me rather secondary compared to
that. But this is of course again just my own experience.
I agree with Jürgen here. Moreover, after having spent a few hours
looking at the details of mmd and pandoc formats yesterday night, I've
become very skeptical about my previous suggestion to use conversion
to either as an intermediate step that would allow both export to
Word and round-trip conversions.
So it's back to Richard's suggestions of keeping the two projects
(export to word vs. roundtrip) separate. The issues I see, however
are:
- htlatex is in a state of suspended animation. I am not clear for how
long if will be maintained, if at all. On the other hand, replicating
htlatex's approach (compile with (xe)latex to a special dvi format
than parse the dvi output) seems a massive undertaking.
- for the roundtrip conversion: is it better to settle on an
intermediate format (XML|pandoc|etc) or attempting a direct (and
limited) translation?
Anything we did for one could of course be reused for the other. And I
would try to do all of this directly. The last thing we need is another
format.
Perhaps the initial focus of the project should be on export, for the
reasons Jurgen gives. The main issue here, which comes up on the list
all the time, is the need to send Word-compatible files to journals. So
I'd suggest we (someone, not me) implement a native export routine that
writes ODT and/or DOCX. We can figure out what to do about biblatex
along the way.
Richard
