[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Bug-AUCTeX] Re: preview oddity when "merging" inline formulas...

From: Evil Boris
Subject: [Bug-AUCTeX] Re: preview oddity when "merging" inline formulas...
Date: Mon, 11 Jul 2005 19:47:40 -0400
User-agent: Gnus/5.11 (Gnus v5.11) Emacs/22.0.50 (windows-nt)

David Kastrup <address@hidden> writes:

> Evil Boris <address@hidden> writes:

>>    and this depends on $x$ and $y$ MORE TEXT MORE TEXT MORE TEXT MORE

>> [...]

>>    and this depends on $x, y$ MORE TEXT MORE TEXT MORE TEXT MORE

>> with the cursor someplace around the comma.  Notice that there are two
>> preview-generated pictures around [can't really tell what they are
>> supposed to represent, too small for my eye sight], which is annoying
>> but understandable.  If I type C-c C-p C-p now, everything after the
>> $x,y$ disappears from the display!  Very disconcerting, especially the
>> first time it happens.  C-c C-p C-d makes everything peachy again.

> "Merging" previews is not really a sensible operation.  What you end
> up with is regenerating the only one, crippled preview.  The results
> are that the math environment is incomplete, and TeX looks beyond the
> actually cut region for fixing it.

I do not know if I explained myself properly.  What I do is have "$x$
and $y$" in my text, previewing fine.  Then I edit it to say "$x,y$"
and try to force preview to update the picture at the point.  At this
point all hell breaks loose---all or some part of the containing math
expression gets a proper preview image, but the rest of the file (and
I had this experience with a 20 page paper) **appears** to disappear.
I thought that it was unnecessarily brutal for an unsuspecting
user. :)   

A secondary issue is that there are those slightly annoying pictures
attached to "opened" math expressions --- in this case the second one,
that was attached to $y$ refuses to go away, even though it looks to
me like I have have deleted character (namely the $ in front of "y")
to which it seems to be attached.

> Now we _do_ have the problem that an error occuring outside of the
> actual region will not get correlated correctly.  This is pretty
> involved and a result of AUCTeX's region processing.  We won't fix
> this before the next release.

I am not sure how preview picks a region when doing a
preview-at-point, so I am not sure what you specifically you are
referring to here.

> However, it might be a useful idea to regenerate butting edited
> previews together.  Does the following patch lead to more satisfactory
> behavior for you?

Thanks.  Will take a look a bit later.  


reply via email to

[Prev in Thread] Current Thread [Next in Thread]