Georg Baum wrote:
Abdelrazak Younes wrote:
For someone who knows the intricacy of GUII perhaps. I've done a lot of
GUI programming and I don't share at all your opinion. Requiring someone
to learn Qt and also to learn LyX GUII architecture is too much IMO and
that is the very reason why we don't have so much GUI developers.
I agree in general, but my point was that (in comparison to other existing
problems) there is not much to learn (not for people doing a redesign like
you did, but for people who e.g. simply want to implement a new inset). I
did not invent this stuff, but I had no problems using it.
Well, Andre's question was about creating new dialogs (which the GUII is
about) not new insets but never mind, I got your point.
Far more time is consumed by half finished half documented redesigns of
half finished half documented redesigns. For example, metrics/draw: The
first unfinished redesign I know of happened before 1.4.0, and created
among others bug 1814. The next redesign happened before 1.5.0 and was
not finished either.
I think I left the metrics in a better state, mind you.
The last time I tried to help to fix bug 1814 and understand this code
during 1.5 freeze (you may remember because I asked some questions and you
created several longer answers) it was not possible to understand it, and
there was very little (if any) documentation. I remember especially several
methods called metrics() doing quite different things and two ways of
passing some update flags that were very confusing.
Year... the update flags mechanism is also a mess. But I did not change
much that state in the 1.5 development.
I can tell you why bug 1814 is still not solved, although it is a major
issue, makes e.g. branches with more than a few lines unusable and
several people already invested a considerable amount of time trying to
solve it: Because fixing this requires a full understanding of the
current state of the metrics/draw machinery (not some idealized, but
unimplemented theoretical idea), and gaining that is simply too
difficult.
Not anymore. And I fixed some parts of this bug in trunk already.
I am glad to learn that this has progressed since then. Does this mean that
this bug can now be fixed completely?
I think so yes, by calculating and storing real paragraph height of the
whole document when it is first loaded. Then each new full update
metrics call will update the stored height of the recalculated paragraph
metrics. I'd like to finish correcting the remaining metrics bug with
selection before tackling that.
Note that scrolling with PageDown/Up key is now fixed in trunk. So is
scrolling by clicking in the gray area of the scrollbar.
Abdel.