On Monday 09 May 2005 11:14, Angus Leeming wrote:
[...]
> to:
> ...
> + QKeyEvent ev(QEvent::KeyPress, Qt::Key_AsciiCircum, '^', 0,
> + QString("^"));
> + keyPressEvent(&ev);
Not only that but if the patch is working there is no need for a special
handling for the caret accent. What happens if you a keyboard with dead
keys support is that certain keys need an additional step to show any
result.
Case in point, I need to type ^+Space to get a simple ^.
It would be nice if inside math mode typing a simple ^ would show a
subscript inset without the additional space, but that is clearly an
enhancement and not any bug we have now.
I will try to add the required changes to 1.4 and then we can backport the
changes as soon as the quirks are solved.
> There's no need to create ev on the heap.
>
> José, I seem to remember that you were having problems with accented
> characters in the past. Are you interested in getting this patch into a
> finished state? (Requires handling of different accents I guess.)
I am for sure. This is the most annoying inconvenience I ever found
working with LyX. People around here are not too concerned with the
slippery of 2003 Christmas timeframe (read 1.4.0 release) but this is
really, really obnoxious.
The only workaround is to use the xforms frontend, that implies to build
an rpm with both frontends, thus building the sources twice since the
building mechanism in 1.3.x does not support multiple frontends building.
Incidentally and without any scientific claim, I have used the cvs sources
to create the rpms and until the moment and don't have any complain other
than the dead keys issue. So it seems that although the source changed a
lot since 1.3.5 (mostly just for windows) the resulting source for linux is
working flawlessly. And no, I can not cite this sentence to claim the 1.3.6
has no bugs. ;-)
--
José Abílio
