PIVOT-738, workaround/safer fix

[Sandro Martini <sa...@gmail.com>]

Hi all,
in this issue https://issues.apache.org/jira/browse/PIVOT-738
I just put a candidate (or better, a working :-) ) patch for a
workaround where I add a style property that when enabled (like in
this case) all the viewport (visible) area is repainted (and not only
a portion of it, more optimized) ... because in some strange cases
like this the more optimized repaint sometimes doesn't calculate good
bounds.

Of course this is not an optimal solution, but we have to release very
soon 2.0.2, and I think that in any case a workaround like this could
be useful even in other places ... and we could fix the wrong
behaviour without changing existing calculations ... then maybe for
2.1 we can see a clean and always optimal solution.

What do you think ?


Note that in a comment there (
https://issues.apache.org/jira/browse/PIVOT-738?focusedCommentId=13269679&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13269679
) i wrote even to change a small block of code that is using a really
old and deprecated calls to AWT, and in my tests the change i have
made works ... but of course has to be tested more deeply (even with
transparencies) and an many platforms ...

Comments on this ?


Unless objections tomorrow I'll cleanup and commit this fix ...

Tell me.

Bye,
Sandro