Lars Gullik Bjønnes wrote:
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| This needsredraw() function is right now necessary because the
| BufferView::update() method is used all other the place. But the plan
| is to remove the needRedraw() method.
Ok... so the need to redraw of not will eventually be done through a
function return value?
Yes, the correct protocol still needs to be decided.
| Eventually, when the cleanup is done the command flow will be like this:
|
| User press a key.
|
| WorkArea::processKey(...) {
| BufferView::processKey(...) {
| ...
| BufferView::update(); // _no_ call to any LyXView method in there.
| }
| WorkArea::redraw();
How does the WorkArea::redraw() know whether to redraw anything or
not?
Maybe something like this:
bool BufferView::processKey(...) {
...
return BufferView::update();
}
WorkArea::processKey(...) {
if (BufferView::processKey(...))
WorkArea::redraw();
}
| > My plan was to do this merge yeasterday, but real life intervened.
|
| You should have started with this instead of merging the most recent patch.
Why?
The recent path was ortogonal to the first.
Because I have to do a lot of manual merge with trunk. You have also
committed partially the patch we are discussing right now.
Abdel.
