Hello Przemek
Przemyslaw Czerpak-2 wrote: > > I do not understand why we have WIN_AXRELEASEOBJECT(). For me > calling this function breaks autodestructors. Pritpal? > I had introduced this function before auto-release pointer and is kept by oversight. This function is not used in core as well as in GTWVG as well as in my application code. It is safe to remove it. > Release() method makes something strange for me. Mindaugas? > To implement such functionality it's necessary to change > the method of iSink allocation. It has to be allocated by > hb_xgrab() and then we should only allocate holder for > iSink pointer using hb_gcAlloc(). The only one reason this > code was working without GPF traps is this line: > pSink->count = 2; /* 1 for pCP->Advice() param and 1 for Harbour > collectible pointer [Mindaugas] */ > which effectively blocks executing hb_xfree() due to initialization > with 2 instead of 1. BTW I do not understand above comment. For me > it should be 1 and other code fixed. > This I had pointed to Mindaugus that this value should be 0 or 1 but not 2. Actually some active-x's may be gpffing without this hack, but am not sure. In GTWVG I place this value as 0. Regards Pritpal Bedi -- View this message in context: http://www.nabble.com/Possible-OleAuto-bug-tp25044738p25064961.html Sent from the Harbour - Dev mailing list archive at Nabble.com. _______________________________________________ Harbour mailing list Harbour@harbour-project.org http://lists.harbour-project.org/mailman/listinfo/harbour
