Wily >I think you need to continue your "binary search debugging" (== "commenting >out stuff)." You seem to be very lucky in this case, at least inferring from >your statement that running a 2nd thread gets you to "kerpow" quite >reliably. That's a blessing with threads...
You'd think so but if you look at the bit of code left there's nothing left to AV >Just continue knocking out strategic (large, if possible) sections of your >code (making sure that instantiations are matched--included or excluded-- >with their associated releases--ie. if you comment out an object creation, >make sure its destruction is commented out too), till you find what triggers >the problem and what doesn't... BUT, sometimes you need to follow the >problem areas INTO calls, as you narrow things down, to identify the true >issues. Using FastMM4 with the advanced debugging options on, can often >help too... Very good advice and all done. >If you already narrowed it down to TMimeMess, it may be time to single-step >through those calls, or some other debugging techniques... But then when out walking the dog I realised I that my main & news reader I remembered I'd been using TMimeMess in background threads for ages. It has to be some horrible complex interaction of something. The last one like this (several years ago now) took months to track down. At least this time its not intermittent. Roy Lambert ------------------------------------------------------------------------------ The Planet: dedicated and managed hosting, cloud storage, colocation Stay online with enterprise data centers and the best network in the business Choose flexible plans and management services without long-term contracts Personal 24x7 support from experience hosting pros just a phone call away. http://p.sf.net/sfu/theplanet-com _______________________________________________ synalist-public mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/synalist-public
