On Tue, 2003-09-30 at 08:08, Michael Meeks wrote: > Well - I spent another chunk of time writing more regression tests to > try and catch this case - and they all work. I also spent some time > trying to unwind how it's possible for the case in Ettore's trace to > happen, and I found that it's possible we fumble the pobj use_cnt, a fix > to follow. > > Can people actually reliably reproduce this bug though ? I really want > to avoid having to build evo, and I want to know if this fixes it.
Yeah, at least I can. Thanks for looking into this, Michael! I know you are busy with other important stuff, and your help is much appreciated. Bye, -- Ettore Perazzoli <[EMAIL PROTECTED]> _______________________________________________ evolution-hackers maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/evolution-hackers
