Abdelrazak Younes wrote:
rgheck wrote:
Abdelrazak Younes wrote:
Pavel Sanda wrote:
That's the reason why I'd like us to standardize on "right-click"
== "context-menu".
either one left click to (un)collapsing or ok context menu, but
then lets go back
to old visual style of ert inset, where was possible to click on
the left label
to collapse job.
I don't have strong opinion on old style versus new visual style
for ERT. The new style looks better on screen of course.
in fact i dont have strong opinion about either context menu or
visual styles.
but what is wrong wrong and wrong is to loose the possibility of
one-click
(un)collapsing.
Two clicks is not a big deal for a not very frequent operation IMHO
but anyway, I mostly agree.
Would ctrl-left-click be OK as JMarc proposed? It'd be similar to
the zooming function in drawing software for example. That could be
implemented easily in InsetCollapdable I guess.
Whatever we do about clicking here, the more immediate problem is
that Ctrl-I ought to collapse the inset when you're inside it, but it
doesn't. This may have something to do with a problem mentioned in
another post of mine:
http://comments.gmane.org/gmane.editors.lyx.devel/104443
If you're in front of the inset, this function works; but if you're
in it, it does not, though it should. But it may be something else.
Namely, it seems that cursor().result().dispatched() is true when we
enter LFUN_NEXT_INSET_TOGGLE in BufferView.cpp (line ~1265), and then
if inset is false, nothing will happen; in particular, we won't go to
cur.dispatch(). Now we could solve this by just calling
cur.undispatched() here, but I don't know enough about this code to
know if that is the correct fix. I'm guessing it's not, and that the
problem here is more general.
For one LFUN_INSET_TOGGLE did not have an associated name, I fixed
this in r23884. Problem is that C-i is binded to
LFUN_NEXT_INSET_TOGGLE not LFUN_INSET_TOGGLE... maybe we should merge
the two LFUN?
Well, it works fine in 1.5.x, and that has much the same code. The big
difference is that, at the mentioned point in BufferView in 1.5.x,
cur.result().dispatched() is false rather than true.
I know you're on vacation, and hope you're having a good time, but if
you get a chance to look at the other problem, that'd be great.
rh
