# [NTG-context] preview-context

Mon Dec 24 01:29:00 CET 2007

```On Sun, 23 Dec 2007, Hans Hagen wrote:

> Gour wrote:
>> Hi!
>>
>> I've asked on auctex ml about the prospects of getting preview-context, similar
>> to preview-latex and here is the reply:
>>
>> "There are no ConTeXt users among the active developers (actually, there
>> is not even much activity in the LaTeX area, but the status quo is
>> rather more advanced).  As long as that does not change, I don't see
>> much of a point to talk vapour."
>>
>> (see  http://article.gmane.org/gmane.emacs.auctex.general/2387)
>>
>> Since I plan to move to ConTeXt using AUCTeX, I'm interested to hear if
>> there is some interest here (amongst emacs users) to have preview-context
>> withing AUCTeX package?
>>
>> otoh, I'm interested if there is newer version of  etexshow
>> (Browser for ConTeXt commands) ?
>
> in my opinion previewing contradicts symbolic markup (apart maybe from
> special subconstructs); anyhow, because context sometimes relies on
> multipass info, previewing is no option unless on fast machines

I disagree. I was grately impressed by preview-latex when I tried it
(around 2 years ago), so much so that I even considered moving to emacs
from vim. It is extremely useful for mathematics and images: you can see
the typeset result rather than a bunch of code. This means that you can

What is needed for an improved support of preview-latex in context? Last
time I looked into preview.sty, I could not understand how it works. If I
were to do the same thing, this is how I would do it.

Run context with a custom module, which writes the content of each
\startformula <formula> \stopformula into a temporary file as
\startTEXpage \startformula <formula> \stopformula \stopTEXpage. Lilypond
and gnuplot modules already do something similar. Then process the
temporary tex file. You will get a pdf file with one formula on each page.
Convert each page into png, and use some lisp magic to insert the image at
appropriate places in the buffer. Obviously I am missing a lot of things.
But, am I correct in assuming that from the context end, the problem is
simple?