On 11/30/2015 04:59 AM, Amos Jeffries wrote: > As far as I am aware all the major features and functionality that were > roadmapped for Squid-4 have made it into the 4.0.3 beta. > > So I am planning to branch trunk on Thurs/Fri this week. The sponsored > additions can then gain some testing in Squid-5 without delaying Squid-4.
Your call. Given the current poor state of v3.5 and v4, I doubt trunk will receive significant testing if you start committing major "sponsored additions" there: We have more than enough technical debt in v3.5 and v4 to keep us busy for a while, with no working regression testing setup. If major changes start appearing in trunk, then I would have to stop recommending running trunk for getting access to the state-of-the-art features like I normally do. Instead, we would be forced to add those state-of-the-art features and fixes to a v4-based branch from where they will eventually trickle to trunk, with even more delays and overhead than today. Thus, I recommend against branching v4 and for committing "sponsored additions" with high regressions risk to [sponsor-dedicated] branches [for testing by sponsors and other interested parties] until we have a decent regression testing in place. My recommendation would have been different if we had not wasted so much time on regressions recently and if we had a working regression testing setup today. FWIW, I am renewing my efforts to launch that project and hope to post the second proposal for the Board by the end of this week. Cheers, Alex. _______________________________________________ squid-dev mailing list [email protected] http://lists.squid-cache.org/listinfo/squid-dev
