Hi All, IMO, these new functions should be moved away from hbwin, until we clear up these issues. These functions are nowhere used inside Harbour (no in any contribs) so it's very difficult for anyone on the list to decide about their usefulness or to even try / test them.
Adding such complicated function just for the sake of having them is IMO pointless. Until we find a fitting place for them and clear them up, we can move them to GTWVG with slightly different names. My question is whether to rename__AXGETCONTROL() to WIN_AXATLGETCONTROL() ? Is this such a function which users will want to (and can safely) use in their final apps? If the return value is an obscure one, useful only for AX extensions dealing with internals, I think we should keep the internal name. Brgds, Viktor On Tue, May 26, 2009 at 5:35 PM, Mindaugas Kavaliauskas <dbto...@dbtopas.lt> wrote: > Hi, > >> * contrib/hbwin/axcore.c >> + __XAXREGISTERHANDLER() >> An extension of __AXREGISTERHANDLER() where connection points are >> enumerated instead of finding - many OLE's have more than one interface - >> a practical discovery. Mindaugas please check if it can be the >> default. > > Sorry, perhaps my OLE knowledge is too poor, but I totally do not understand > that you want to do with this. OK, OLE can have more than one connection > interface. But you do not even look at that kind of interface you've > obtained. You return the first interface as IDispatch. This function can not > be used even enumerate interfaces or obtain any interface except for the > first. > > In file system terminology, it sounds for me like... You need harbour.exe to > compile .prg code. So, you do file enumeration of the folder and then try to > use the first file (regardless filetype: .exe., .txt or .doc) as harbour > compiler. > > >> + WIN_AXGETUNKNOWN() - Useful in many situations. > > Where you can use IUnknown in Harbour? > > >> + WIN_AXSETVERB() - ---^--- > > Why SetVerb if method is DoVerb()? How can you pass MSG parameter to DoVerb > method (it's input parameter, it should not be stuffed with 0 in general > case)? > > >> + WIN_AXRELEASEOBJECT() - Experimental >> Note that these function are primarily aimed at making experiments >> via gtwvg/winsink.c interface and after testings can be moved or >> removed. > > I would include it into gtwvg, if this is the reason. > > >> /* Mindaugas, I receive a GPF when terminating the appln. >> Also I cannot see that ISink is ever subject to garbase >> collection. >> And "count" never returns to 0. I must be missing something, >> but... >> */ > > Any self containing GPF sample? > > > Regards, > Mindaugas > _______________________________________________ > Harbour mailing list > Harbour@harbour-project.org > http://lists.harbour-project.org/mailman/listinfo/harbour > _______________________________________________ Harbour mailing list Harbour@harbour-project.org http://lists.harbour-project.org/mailman/listinfo/harbour
