Alex Rousskov

I have resolved change requests addressed as of commit https://github.com/squid-cache/squid/pull/335/commits/f31f3fae0a920fbf8687d74d5a9867c73540b56f. There are approximately four change requests that remain unresolved, and I have commented on most of them. All of these...

> I think this is all the requests handled now. Please check. I found three unaddressed old change requests: [1](https://github.com/squid-cache/squid/pull/335#discussion_r361541057), [2](https://github.com/squid-cache/squid/pull/335#discussion_r361539120), and [3](https://github.com/squid-cache/squid/pull/335#discussion_r361541502).

> note: mswindows.h changes are from the mingw-cross-1 branch . I'm not entirely sure how to stack PR on top of each other so that github only shows the current...

> we use phabricator and it supports PR stacks, it's quite convenient. To avoid misunderstanding: GitHub supports a flavor of stacked PRs for _same-repository_ branches. We use that feature internally...

> since there have been changed since I'd like to have @rousskov 's opinion My [opinions](https://github.com/squid-cache/squid/pull/1085#pullrequestreview-1048568272) have not changed.

> This work ... doesn't regress the general status and is a step forward to the intended goal. I hope that is true, but I cannot validate that claim. >...

> the callsites use bind(int..). The macro in compat/os/mswindows.h is for them to maintain the illusion that this is what the OS provides. That macro then calls Squid::bind(int..) which does...

In other words, what will break if we remove our macros _and_ remove our Squid namespace?

> It might be that mingw is doing the emulation job for us? I can try to remove most of mswindows.h if that is the case and see what happens...

@hfisiteimprove, thank you very much for working on this useful improvement! We do want to reject configurations with conflicting port configurations. I share the _primary_ concern [voiced](https://github.com/squid-cache/squid/pull/792#pullrequestreview-625363563) by Amos --...