Alex Rousskov
Alex Rousskov
> Confirmed latest commit does not show reuse of outbound TCP source ports Please see my earlier [review](https://github.com/squid-cache/squid/pull/1115#discussion_r943464151) for that bug triage and suggested fixes.
> Thanks for the last commit [d96d055](https://github.com/squid-cache/squid/commit/d96d055e52a5b91900540941a2d172c830105b23) to the PR. If we need to wait for further approvals, can you please let me know so that I can plan for...
> Hopefully this PR enhances the scaling aspect w/o breaking other existing functionality. :) Yes, that is my hope or understanding as well.
I adjusted PR description, primarily to avoid leaving autoconf version hanging on its own line.
> I am interested in what you consider a "final" solution I hope to post my suggestions this week. > Each script contains the rules to perform a different type...
> clang-tidy has a quite effective modernize-use-nullptr module for this purpose It was also a part of my squid-dev [RFC](http://lists.squid-cache.org/pipermail/squid-dev/2020-April/009569.html). > This is effectively just a s/NULL/nullptr/ with better detection...
> Alex: I see no significant problems with removing all NULLs from C++ code > Amos: Removal of NULL ... is really only worthwhile doing when the result is clean...
> Can we unstick this PR? ... Can we converge on a technical solution? If nobody beats me to it, I will post a comprehensive solution based on using clang-tidy....
> Alex: If nobody beats me to it, I will post a comprehensive solution based on using clang-tidy. Hopefully, that will clearly show the already known advantages of that approach...