Ok, I got back too late yesterday and now I was examining my patch. And I was surprised when I noticed that I didn't remembered my patch as it was out of my head.
My patch never changed any (class is TClass) handling, all inside is your code. I only changed after else (it's not relaxed and won't be as it was said n mail about relaxed mode), and I all along thought that I forgot to check is_related. How it could do too much, did you even inspect the code??? I should probably ask now: "Can I now really get answer what would the problem be?? Interface has no code and thus "is" should be flexible, class has code and it should not be flexible." but I won't, it's getting tiresome, and as always I won't get answer. class "is" (some not related class) - bad interface "is" (some not related interface) - good, that is how it is supposed to work, and you can't make MI interfaces without this option. I'm on about 90% of working MI interfaces and after that is inclass implementation and foreach (inclasses are much better solution than completely missed "implementing interfaces by delegation" or "naming resolution clauses". Whoever needs those two, well it can implement them, I wont because they are stupid and delphi compatible is not on my needed list"). Don't worry, I won't bother you to include MI, inclasses or foreach (in fact I don't even think about it), I suspect it would just lead to another tiresome thread. I'll just notify you, where you can download the patch and documentation, and what you do with it is your decision (somehow history on this mailing list made me think that I would be better off even not knowing your decision). As I said I only want to contribute things incommon and not force anybody. I can't fix the problem if I don't know what the problem is. You can either: - say where the problem lies (example) - correct it as you wan't (I can't, I don't know where problem is) - implement it on your own - use my patch - send my patch in trash and live without (I won't) Personaly, I stopped carring (its been a week or two now, and I still haven't got the answer even where the problem lies) ml ---------------------------------------------------------------------- This mail sent through Horde-Toaster (http://qmailtoaster.clikka.com/) _______________________________________________ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
