I sent the following earlier today but, with the network misbehaving it may
not have gotten through. I am doing a Leak test build now and will report
results when they are available:
---------------------------------------
Yours may still be the culprit. We've been having some problems running
the
leaks tests. We got results on the 16th, missed the 17th, and then got
results again on the 18th and 19th. Both 18th and 19th numbers were up but
I waited until the 19th to send notice because I wanted confirmation that
the jump wasn't just noise. So, leaky code checked in on the 17th would be
consistant with our test results.
In any case, once we get the network back so we can do a build, today's
tests should tell us if your check-in did the trick.
Thanks,
Curt
Bernd Mielke wrote:
> Hmmm, sorry for the leak, which is now fixed. But I am afraid that this
> is not the leak we are looking for. The CVS log for
> BasicTableLayoutStrategy.cpp shows that my checkin was on Apr 16 21:32.
> The problem should have been visible in the tests from Apr 17 and Apr
> 18. I haven't seen any notice on these days about a leak increase.
>
> Bernd
