On Thu, 29 May 2008, Mindaugas Kavaliauskas wrote: Hi Mindaugas,
> P.S. do we have a final decision about PCOUNT()? At the begining I wanted to change only PCOUNT() to return Clipper compatible results without touching other things to return to this problem after final 1.0. But after your message I changed my mind and I suggest to now only modify hb_vmFrame() to work just like hb_vmVFrame() It fully resolves the problem and HB_APRAMS()/HB_PVALUE() will operate on exactly the same parameters as user passed to function. IMHO even the modified order of destructors execution for additional parameters will be more natural for users (sth what they expect) and (the most important) it will be the same as for .c functions. Please note that now it's different and when user execute function written in C all parameters are cleared on function exit but when it's .prg function additional parameters are cleared at startup and rest on exit. So if no one will vote against I plan to change hb_vmFrame() ASAP and leave any other modifications/optimization after 1.0. best regards, Przemek _______________________________________________ Harbour mailing list [email protected] http://lists.harbour-project.org/mailman/listinfo/harbour
