On Sat Aug 3 20:23:16 2024 +0000, Elizabeth Figura wrote:
The caller is just IShellFolder::ParseDisplayName() on the desktop folder, which wraps this function pretty much directly. It seems very easy to move the tests to shell32 and convert them to call IShellFolder::ParseDisplayName() instead of IEParseDisplayNameWithBC() [which would allow the tests to actually work on any of our testbot machines], and it seems very easy to move the implementation to shell32 and stop calling into shdocvw [which would apparently match the architecture of modern Windows]. Is there any reason not to do this?
By working on this project and submitting patches we generously give away the most valuable thing that we have - our time. It's pretty much disappointing that the project maintainers take that for granted. This particular patch is old, already took several rounds of suggestions and changes before the assigned for the review person even looked at it. After that the reviewer asked for a code change, and once it was done asked for a complete rewrite of the completely different part of the code from scratch, and after that dropped the ball. Isn't that a little bit not fair? And that's just for this MR with a single patch, there are plenty of other MRs without any attempt for a review or a comment.