[prev in list] [next in list] [prev in thread] [next in thread] 

List:       lyx-devel
Subject:    Re: HTML as a New Output Format
From:       Pavel Sanda <sanda () lyx ! org>
Date:       2009-04-30 15:17:32
Message-ID: 20090430151732.GF28912 () atrey ! karlin ! mff ! cuni ! cz
[Download RAW message or body]

Richard Heck wrote:
>
> I'm reposting this separately.
>
> Comments?

..ok i'm reposting too..
-----------------------------
this looks like something you can't get completely right. now what i mean
'right' here - right means to get the same version of the document in pdf or
html - in the sense of content not appearance...
i can't imagine how you would care about various ERTs - imagine reseting some
counters etc. now this is the advantage of using latex->conveters, they are
prepared for tex, math is done externaly via latex itself etc.

if one wants good typography, choose pdf/ps if one wants only structure and
semantical part one can choose html, which was never designed for
fingerpaintings.

i accept that some people didn't like aesthetically the output of the current
latex->html tools so i didn't comment on just next convertor to html and in a
way i'm not concerned whether it supports all the lyx features - like the list
you sen't previously. just a small tool for people not wanting too much of
features and a little bit better whistles and bells :)

when i commented thats better to use internal lyx code for output i meant that
it is better then parsing .lyx file (not .tex file) - i didn't mean we should
do that coding! :)

if there are problems with lyx->html conversion lets help fix latex->html tools
i would say on contrary.

to sum it up - i think this project would spend lot of energy, the output would
never be total of things you see in pdf output (in semantical sense) and if one
doesn't aspire on this totality, the current solutions seems to be quite good now.

also i dont think that having such proper and total html output is so much
valuable for lyx - i may enter flame here - but html is very bad output format - 
each few years new version, each browser different interpretation, neither
cleanly semantical structure like xml nor fingerpaintings like dvi, but some
madhouse inbetween - thats why no need to spend long month for html development.

where we are really failing is the sematical part - lyx claims to have support for
docbook/sgml output, but last working manual for sgml is targeted on lyx
1.2 and the recurent questions on users list there would interest...
the later manuals are doing some ugly sed-like operations on the generated outpu
output to get proper xml and i suspect to get this right would be not hard
for somebody who knows something about xml. because we dont have this working
people tend not to support this formerly (complete?) output format in new features
and we dont get feedback from users to fix them either...

pavel

[prev in list] [next in list] [prev in thread] [next in thread]