Re: supertyping

2006-12-14 Thread TSa
HaloO, Larry Wall wrote: Or we could say that you can't reopen the Num role; you can only reopen the Num class and mix in the Complex role. That's where it stands at the moment. This is not too bad an outcome. One question though: is the augmentation of the class lexically scoped? Or does a m

Re: supertyping

2006-12-14 Thread Smylers
TSa writes: > Larry Wall wrote: > > >role Num is also does Complex { > > method im {...} > >} > > Is that the actual syntax? Larry's words that you snipped introducing that code fragment were: We might *possibly* get away with reopening roles like we can reopen a class: It see

Re: supertyping

2006-12-14 Thread TSa
HaloO, Larry Wall wrote: role Num is also does Complex { method im {...} } Is that the actual syntax? I mean is it the keyword pair 'is also' or does 'also' by itself have a meaning? With a more natural 'role Num also does Complex'. but roles are really supposed to be fairly

Re: supertyping

2006-12-14 Thread TSa
HaloO, Luke Palmer wrote: For now (because of this example, in fact), I'm inclined to change the proposal to "please don't design the language to prevent a module from implementing supertyping". I think disallowing reopening of roles will prevent that. I might not have formulated it this way