I'd like to take a shot at writing such a proposal. At some point improved existential support will need to go through review, so we might as well.
Austin On Tue, May 17, 2016 at 11:02 AM, Douglas Gregor via swift-evolution < [email protected]> wrote: > > On May 17, 2016, at 10:40 AM, Adrian Zubarev < > [email protected]> wrote: > > You don’t seem to be tackling the case of “A Collection whose Element type > is String”. If we’re generalizing the current “protocol<>” notion, why not > make it as powerful as a generic signature, with the ability to specify > same-type constraints and conformances on associated types? > > - Doug > > > Which part of the manifesto did I left out? ^^ Could you provide a quick > pseudo code example? > > Do you mean something like `Any<Collection where Element == String>`? I’m > not sure where I should consider such a scenario, maybe at future > directions? > > > That’s the part I’m referring to, yes. I just realized that your proposal > isn’t lifting the restrictions on protocols with Self types or associated > types, so my suggestion doesn’t make sense for your proposal without a > significant increase in scope. > > - Doug > > > _______________________________________________ > swift-evolution mailing list > [email protected] > https://lists.swift.org/mailman/listinfo/swift-evolution > >
_______________________________________________ swift-evolution mailing list [email protected] https://lists.swift.org/mailman/listinfo/swift-evolution
