mailing-list for TeXmacs Users

[Phil Mendelsohn <address@hidden>]

On Fri, Mar 29, 2002 at 11:59:57PM +0100, Joris van der Hoeven wrote:

> What is really going on is that cursor movement is purely graphical.
> If you type \hat{S} at the end of a mathematical formula,
> then the middle of cursor is graphically outside the formula,
> because of the right slope of the S. Consequently, you will go
> out of the formula when moving the cursor right.
> Nevertheless, you may move the cursor back left again and
> you will be able to position the cursor as you wish.

Yes, this is consistent with what I see; you understand the situation.

> Maybe the current scheme should be optimized somehow.

Yes, indeed!

> One might also choose the graphical reference point for
> the cursor at the base line, but this might cause some
> other problems. Ideally speaking, I might try to find
> a graphical cursor position which rematches the logical
> position after conversion (by moving left or right
> for instance). 

Hrm.  There are two things cursor position is representing.  First, it
represents the physical position, and you deal with that.  But, it
also represents where the cursor is in the structure of a document.
Moving it physically out of an environment should not move it out of a
logical environment until stated, either explicitly or by a consistent
algorithm; indeed it does not for letters that stay within
graphical bounds (as you have indicated.)

Deciding what mode the cursor is in needs to be a logical AND
condition; when it outside the formula physically and logically, only
then should the cursor be considered outside the environment.  I
haven't looked at the source, but I'd think you'd have a "mode
flag" of some sort.  The fix would be to adjust that so that first you
hit the end of the environment, and only from there can you exit the
environment by CTRL-f.

> Please put this on the wish list
> as not so urgent (this problem is not that frequent,
> although it occurred to me in other situations too).

Thanks, I will.  I tend to disagree, however, about the urgency when
it is viewed as fundamental inconsistency in displaying/editing
structured documents.  Inconsistent behavior is a real turn-off to
potential new users, and I can tell you that anyone using it to type
set notation or graph theory in particular wants overbar-S or
overbar-G pretty frequently.  But, that's just my two cents worth.

Cheers,
Phil

-- 
www.rephil.org

"Trying to do something with your life is like
sitting down to eat a moose." --Douglas Wood