staikos at kde.org
Wed Mar 16 01:39:20 CET 2005
On Tuesday 15 March 2005 19:24, Barth Netterfield wrote:
> In principle, one could shrink the window to zero (perhaps make it small,
> then turn on the toolbar/statusbar). While it is hard to imagine when this
> could be useful, it is similarly hard to argue that we should prevent it,
> or that kst should die if it happens....
> Baring recent changes, kst has a minimum size for the window. With the
> statusbar and toolbar off, a usable plot gets drawn. With them on, the
> draw space is not zero, but is too small to draw a plot. It is possible (I
> haven't verified this) that with an appropriately huge menu font, the
> minimum size would allow the pixmap size to shrink to zero.
> I think that the correct behavior should be:
> -no preset minimum size.
> -kst should happily sit with a zero sized window until it gets made larger
> But I don't feel strongly enough to suggest that anyone do anything about
I'd still like to know how to see this. That's why the assert() is there.
I haven't found a way to trigger it yet, though I'm imagining that
hypothetically one could have >n plots in a column in a window n pixels high.
That would make at least one plot be 0 pixels high. I think we should guard
against that instead. It's an absurd use case, and having a possibility of a
null buffer/pixmap seems worse (ugly) to me. The whole (relevant) idea
behind the view objects design in the first place was that this case
could/should not happen.
KDE Developer http://www.kde.org/
Staikos Computing Services Inc. http://www.staikos.net/
More information about the Kst