On Jul 21, 2005, at 9:10 AM, Paul Libbrecht wrote:
I am sorry I wasn't able to take this further yet.
The idea was to introduce something like a reset() method on tags
to ensure that nullity... but if you think we need to disable
caching then we'll have to need a form of method like
"doesNeedCaching" which would then cache in any cases. I know for
example such things for Swing or Define tags (but not per-class).
I still think Cache.reset() is better and Kristofer accept it as an
alternative. Would you ?
Since not caching tags was the default previously, I think we should
revert to that (The difference between revisions 136360 and 136390 in
svn).
It would be re-introducing JellyContext.isCacheTags(), but keeping
the memory leak fix (iirc, that was only removed because we thought
we could fix the memory leak by moving tag caching in the JellyContext?)
I just committed a unit test that illustrates the problem. Any
solution that doesn't require modification of the unit test is fine
with me (as I believe it represents the assumptions that Tag authors
have made, and it would be very onerous to make all users validate
their applications to determine where caching is *not* needed).
-pete
--
peter royal
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
