Re: Major change proposal
"Mark Hahn" <[email protected]> Thu, 5 Aug 2004 10:38:35 -0700
| Newsgroups | gmane.comp.lang.prothon.user |
|---|---|
| Message-ID | <002401c47b13$08e46f40$da01a8c0@walker> |
Paul Prescod wrote: > Inheriting of interfaces is an interesting issue though. It probably > makes those classes more useful to languages in the rest of the CLR. I'm > not as sure how to handle that in the case where you don't have classes. > > It almost seems to suggest that you SHOULD have a class concept and that > concept should map exactly to the CLR strengths and weaknesses. So you > use prototypes within your Prothon code for all kinds of funky runtime > things and then classes for statically making code available to other > CLR languages. Yes, I've been thinking the exact same thing. I'm back to thinking of having the class keyword (and now interfaces) that only does a contract in Prothon, but in .Net gives you the real .Net classes and interfaces. BTW: I think my old use of the term design-by-contract disagreed with your meaning (and probably everyone else's). All I meant by a design-by-contract feature was something enforced at the language level only for the purpose of guaranteeing something like a variable's type, or the adherance to an interface. I think you were reading something more into the term than that.