>>>>> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> But let me bring up another matter. The reported bug is
Martin> apparently, that the code in cursorRight:
Martin> if (cur.boundary()) return setCursor(cur, cur.pit(),
Martin> cur.pos(), true, false);
Martin> if boundary == true, carries the cursor, not to the next
Martin> *character*, but to the next *table cell*. It is as if the
Martin> positioning is using a cursor slice on the wrong level of the
Martin> cursor. Do you get what I mean?
I tried to follow this code this morning, and failed to see why
setCursor would change the cell (so presumably the idx() of the cursor
slice).
Martin> (BTW this bug occurs *only* in tabular. I couldn't find a
Martin> similar bug in any other inset.)
Martin> Anyway, what to try next? I'm still groping in the blind. Some
Martin> more explanation please :-)
I think the code that handles LFUN_RIGHT in insettabular::dispatch
sees that the cursor did not move and decides that it should jump to
next cell.
JMarc
