Hi Viktor, > >> I like it better still, because these are "plugins" in a sense. > > > > But this fact, "in a sense", has *zero* impact on a technical > > problems > > In my first mail regarding this, I clearly wrote that > while I see no TECHNICAL problems here, yet I (IMHO) > find it better to have them separate. When deciding about > something there is not just always technical possibilities > to consider (aka: 'if I _can_ do it, doesn't make it always > the best, or right').
It's *clear* to me, but *I don't see* this argument usable to GT/RDD subsystems. See below. > GT/RDDs are _very_ similar to plugins > (but not equal to them, but at least we don't used to use > this word for it, hence the word "in a sense"). This open, > extendable, "plugin"-like implementation is their biggest > strength and one the very nice and unique features of > Clipper/Harbour. That's why I wouldn't hide it. So we must have a different point of view of what plugin *is*. To me it is a feature which *can* be loaded/unloaded during runtime. Usualy this means loading/unloading dlls. So to me *if* we could load/unload GT/RDD subsystems during runtime then it could be described as plugins. Otherwise they are simply *drivers*. AFAIK we're not able to load/unload GT/RDD at runtime, are we ? BTW, I am not sure you can even switch GT driver dynamicly during runtime - if you can't then GT is even not comparable to RDD, which can be switched during runtime. > If we hide it, we're basically _masking_ this feature IMO, > so that a new developer could not even find out that Harbour > has this possibility and consider the whole "rddSetDefault()" > a mere annoyance without seeing the benefit. Sorry, I don't take this "argument" seriously. We're not hiding anything IMO. We're simply "simplifing" the whole thing. That's all. > I hope we'll hear more opinions from others. Me too. -- Marek Wyjatkowe pioro Andrzeja Mleczki! Licytuj! >>> http://link.interia.pl/f1cf8 _______________________________________________ Harbour mailing list Harbour@harbour-project.org http://lists.harbour-project.org/mailman/listinfo/harbour
