https://bz.apache.org/bugzilla/show_bug.cgi?id=69333
--- Comment #14 from Remy Maucherat <[email protected]> --- (In reply to John Engebretson from comment #12) > > but is a problem with some custom instance manager where it can cause a > > memory leak. > > Superficially, that sounds like an extreme edge case. I would hate to lose > a massively-multiplicative optimization for such a small impact... although > I acknowledge the cost at an individual method is not high. (multiplicative > means hundreds/thousands of JSPs, each with 1 to n instances of the change) > > Is there a written contract? If so, we obviously follow that; otherwise we > may want to evaluate frequency of issue and alternative solutions to the > edge case. This change was basically a revert of a previous fix that Mark did himself (to add the try/finally). So it's simply a mistake. Sorry to see this optimization go, but that's really the way it is ... -- You are receiving this mail because: You are the assignee for the bug. --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
