Weird behavior in formula cursor navigation

I have been using TeXmacs for about 2 weeks now and while I developed some muscle memory in inserting mathematical notation I find the cursor navigation often is very weird. I have read and understood the philosophy of graphical rather than structural movement.

The effect of the movement forces me to concentrate on the screen and pause essentially after every click of a cursor. I do not understand if I am not getting it, whether what I observe is a bug, or what. Please advise.

Let me give an example of one of the worst offenders: consider the displayed formula


and the cursor positioned as follows

I would expect a left-arrow movement to move the cursor behind the sigma. What actually happens is that the cursor is moves after the index 3 after the theta (not even after the F). So it does move inside the parenthesis but in a very strange location.

What does another left arrow from the above location do? It positions after the sigma:

This is very strange.

HOWEVER things are complicated by the following fact. If you get to the same position as before (it is actually the same, checked by looking at the status bar below) by any other means rather than by a left arrow from behind the parenthesis:

then another left arrow moves the cursor down into the 3 index of the theta:

This is natural.

As a result there is absolutely no way to know just from this picture:

where a left arrow will take you.

Please confirm whether this is a bug. If so, I will file it. I think that this should definitely not be the intended behavior, but please feel free to advise me on whether I am thinking about this incorrectly.

It may be a bug because if you apply left arrows from the same position in source code (at the right of the expression inside the equation environment), the cursor enters the around tag and is placed at the very right of the construct, and I think that is the expected behaviour.

Filed it on the bug tracker: GNU TeXmacs - Bugs: bug #60936, Wrong behavior in formula cursor… [Savannah]

Seems indeed a bug to me. I’ve just noted that if one uses the \around* tag (i.e. the resizable one), then the bug disappear. @guraltsev can you mention the bug number here so that people can navigate to it more easily?