On 11/29/2016 11:22 AM, Amos Jeffries wrote: > On 29/11/2016 4:40 p.m., Alex Rousskov wrote: >> On 11/28/2016 07:29 PM, Amos Jeffries wrote: >>> GCC 4.8 is not capable of building correctly operating Squid-4 >> >> You have said that before, but in all such cases that I remember, the >> reality was actually different. I do not know if something has changed >> in v4 within the past month or so [but any such change should have been >> discussed too].
> Coverity identifies 6 uninitialized objects and 4 memory leaks in the > STL templates when GCC 4.8 builds Squid-4. If Squid v4 built with GCC v4 leaks memory, then the leak should be reported and Squid should be fixed. However, a leak in itself is unlikely to be a sufficient criteria for "not capable of building" categorization. More importantly, given the lack of Project resources, we should continue using two versions (stable + development). While nobody has the energy or desire to fight your version increments and branch renaming, I doubt many have 20+% spare time to start working on three increasingly different branches/environments: "stable" v3, "current" v4, and "Amos special" v5. You should not introduce significant barriers between v4 and v5 until v4 truly replaces v3 as the go-to stable branch. At the very least, you should not introduce them without a discussion/consensus! When v4 is stable and v3 is obsolete, let's discuss whether it is time to require GCC v5 (and latest build environments) for master. Until then, let's conserve resources for more important matters. Alex. _______________________________________________ squid-dev mailing list squid-dev@lists.squid-cache.org http://lists.squid-cache.org/listinfo/squid-dev