The action concept, of course, applies: execute a method based on some triggering criteria defined in a subclass of action. However, the specific API does not provide additional value in this case as behavior is based off component listeners and needs to subclass decorator to pick up the paint event (there are ways around this as well of course).
A named action does not help because I would introduce a leak by putting it into a static bag. The "inheriting decorator" needs to have the same lifetime of the component it is decorating. There are some deeper thoughts on separating the action execution request from the action execution via handlers... Yes, could work, but I thought it was not a natural mechanical fit. -----Original Message----- From: Greg Brown [mailto:gkbr...@mac.com] Sent: Monday, June 14, 2010 6:26 PM To: dev@pivot.apache.org Subject: Re: [jira] Updated: (PIVOT-514) >> Could you use the Action API here? > >> > Will check. Don't think so for these use-cases. Could you explain why? Based on what you have described, it seems like it might be a natural fit. >>> I am putting a >>> cross-cutting concern into the Component's scope and allowing it to >>> initialize itself on the component. I'm pushing code up into the class >>> inheritance tree but not using a dynamic language to do it. > >> I assume you mean "component hierarchy" (or "object hiearchy") vs. "class >> inheritance tree", since you mentioned that you were trying to do this >> without subclassing - correct? >> >> > Yup...actually both into the inheritance tree and up the logical tree > kind-of. I won't continue to pursue this since it's probably a bit off-topic anyways, but I'm not sure how it could be both. You can't have inheritance if you don't subclass (in OO, anyways). >>>> - Describe how you have tried to apply existing Pivot constructs to solve >>>> the problem, including source code as needed. >>>> >>> >>> By scripting and subclassing the serializer. >> >> That's a pretty big hammer. >> > I actually don't consider it a big hammer. I suspect that you are unique in that respect. ;-) >> Maybe you could solve this problem via decorators? > it looks like it works. The only error I received was that the Decorator > cannot have 0,0,0,0 bounds and null affine transform's returned The empty bounds issue sounds like a bug. Null transforms are not supported - you should return an identity transform. >>> To be honest, I have spent more time trying to get a >>> split pane (horizontal) to show another split pane (right side, but >>> vertical) than inherited styles. >> >> This is supported and is demonstrated in the Kitchen Sink tutorial. What is >> the problem you are running into? >> >>> > I'll check it out. Never been able to get any of the demos to work across > 3 browsers. Not sure where kitchen sink code is at...could not find in > subversion. http://svn.apache.org/repos/asf/pivot/trunk/tutorials/src/org/apache/pivot/t utorials/KitchenSink.java http://svn.apache.org/repos/asf/pivot/trunk/tutorials/src/org/apache/pivot/t utorials/kitchen_sink.wtkx > My toplevel split pane shows everything fine. But when I put a grid or > boxpane into a split view (right side), its disappearing completely. No > divider or anything. If I just put a label or a grid directly into the top > level split pane, it shows up. I spent more time playing around with this > and looking at the demos (did not see KitchenSink in the demo sub-project). > I'm not even sure how to debug it. Can you provide some sample code (WTKX) that demonstrates the problem?
