David Owens II
Joe Groff
@owensd We try to mitigate that a few ways. Protocol exts can't dynamically change behavior of already compiled code, unlike ObjC extensions
My biggest concern from POP talk: replacing unintended data sharing with unintended side-effects through extensions. Feels like code smell.
-
-
-
@jckarter what about frameworks in your workspace? Does an extension in proj A cause B and C to recompile? - View other replies
-
@owensd It conservatively has to now because we don't have a stable ABI. The visibility of extensions can be constrained though. -
@owensd A can extend types for its own purposes without exporting those extensions to B and C. -
@jckarter hmm... so prot P and type T defined in fmk A, extended in fmk B. Those extensions only called when T is used as signature, not P. -
@jckarter here's a gist with a proper example: https://gist.github.com/owensd/6ceb5b8ec2bd5b43d85b …. - View other replies
-
@owensd Alternately, if the tests module were imported into the main module, it would be able to see circleAt as well. -
@owensd We'll likely introduce warnings/errors to make these shadowing situations more apparent. - Show more
-
-
@owensd Method resolution favors more specific contexts, so if your methods must not be replaced, define them in your type and they can't. -
@owensd Trying to overlook the ageism in the POP talk. Being Crusty, I remember C++ templates and hope Swift will succeed where they failed. -
@owensd Now my class owns a whole bunch of methods that are defined outside its lexical scope! Awesome with caveats.
John Daniel
Steve Trewick
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.