Are primary associated types supposed to be transitive?

[CTMacUser (Daryle Walker)]

Have a function that returns "some Sequence<MyElement>." The sequence type is also a lazy sequence, but "some LazySequenceProtocol<MyElement>" doesn't work. So:

• Transitivity is supposed to work, but there's a (major) problem with the compiler.
• Transitivity isn't supported, and the Swift team forgot to tag a primary associative type for LazySequenceProtocol.
• The Swift team deliberately doesn't want to give LazySequenceProtocol a primary type.

I can see transitivity not (initially?) being supported a valid decision, so I'm guessing the answer is #2. Unless someone can supply a reason why it's actually #3.

[Quedlinbug]

I don't know the answer. I've just used composition to get around it so far.

some Sequence<MyElement> & LazySequenceProtocol

[xwu (Xiaodi Wu)]

@CTMacUser

Unless someone can supply a reason why it's actually #3.

It's actually #3, because there was doubt about the right primary associated type, which would notably not have been Element.

@Douglas_Gregor

@lorentey

LazySequenceProtocol and LazyCollectionProtocol use Elements as their primary associated type, not Element . The idea here is that if we spell out LazyCollectionProtocol , we probably want to also say something about the wrapped collection, not just its element type -- for example, by constraining it to random access collections using some LazyCollectionProtocol<some RandomAccessCollection<String>> .

Personally, I'd rather we not add any primary associated types to LazySequenceProtocol and LazyCollectionProtocol. They aren't likely to be used often, and it feels odd to have them be based on Elements rather than Element like the other sequence/collection protocols. We could always add them later if compelling use cases show up in sufficient quantity.

@lorentey

The document no longer proposes primary associated types for LazySequenceProtocol and LazyCollectionProtocol. I still think Elements is the right choice for these, but this is mostly based on speculation rather than actual use cases. I would like to add these back, but I'd love to see some actual example cases to confirm/reject this choice.