I've been going through and modifying TreeNode from dynacore, and I have
a few questions about why things are done the way they are (or rather,
why doing them any other way seems to make things break inexplicably).
1. Why is precreate needed to prepare the listeners and content;
why can't this be done in the constructor? This could probably
be asked of just about any DynAPI code.
2. Is there any way to avoid making a layer specifically for
catching mouse events? I gather that if you don't do this, then
the canvas layer gets in the way, so is there a simpler way to
make all events automatically bubble so that the widget itself
can catch them?
3. When you add a subnode, it is not added immediately to the
children of the node; rather, this is done in a loop later.
Doing otherwise doesn't work. Why?
4. It seems like TreeNode goes a bit overboard with the events,
using them where simple methods would be more appropriate. Are
there any strong reasons to use events instead of methods, other
than simplified extensibility?
Can anyone recommend any /well-commented/ widgets that might clear these
and other questions up? Thanks.
-Adrian
_______________________________________________
Dynapi-Help mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dynapi-help
