On Tue Sep 5 17:39:15 2023 +0000, Yuxuan Shui wrote:
Figuring out the exact behavior of native urlmon seems to be very involved. @jacek Do you think the solution proposed in this MR is acceptable? In which case I can create a test case with a couple of todos and call it a day...
This stuff is barely documented, so I don't know how it's supposed to work until I can see a test. It's fine to have some todo_wine for some implementation details, but it should at least show that we're doing things in the right place.
For example, I could imagine that we shouldn't change bindprot.c at all and that it's a responsibility of its caller to handle that correctly in its `GetBindInfo` implementation. That would be somewhere in binding.c for the most common case of binding using `CreateURLMoniker`. If that's the case, then adding a test to tests/url.c may be more interesting.