> On Jun 14, 2016, at 1:22 AM, Haravikk via swift-evolution > <swift-evolution@swift.org> wrote: > > >> On 14 Jun 2016, at 07:54, Xiaodi Wu via swift-evolution >> <swift-evolution@swift.org> wrote: >> >> And a language feature being undocumented wouldn't explain why the entire >> stdlib uses it only three times :) > > I sometimes wonder whether the stdlib is always the best example of pure > Swift coding practices, it has after all been developed while the language > itself is in flux so chunks of it were being written while language features > were actively changing, and it has a lot of patterns in it that I just don’t > use in Swift. For example, I still haven’t figured out why the Indexable > protocol has an IndexableBase parent (other than to make figuring out why my > types don’t conform a lot harder),
AFAIK this is because associated types currently can't define a recursive protocol conformance constraint. For example, `SignedIntegerType` has a `Distance` associated type which should also be a `SignedIntegerType`, but has to be implemented using a shadow protocol (`_SignedIntegerType`) instead. > and there’s stuff going on in .gyb files that remains a mystery to me. > There’s also a lot of little hidden features that we don’t have access to in > our own code (I can't find a method that retrieves object identifiers, for > comparing two types without knowing what they are, outside of the stdlib for > example, yet you’ll find them used inside AnyIndex and similar types). Is the publicly exposed `ObjectIdentifier` type what you're looking for? > _______________________________________________ > swift-evolution mailing list > swift-evolution@swift.org > https://lists.swift.org/mailman/listinfo/swift-evolution _______________________________________________ swift-evolution mailing list swift-evolution@swift.org https://lists.swift.org/mailman/listinfo/swift-evolution
