Quality is of course important but not to the point of having development waiting for months.
In the 5.2 milestone, some PRs are ready since June, without unaddressed grounded concerns expressed on them, and they sit idle. And we have opened PRs dating back to 2013. This does not sound reasonable. A trade-of has to be made. Allowing an author to merge without an approved review does of course increase the risk of quality issues. This is why I propose to only allow that on PRs ready since one week. If no one has time to review it and express grounded concerns during that week, I think the quality risk should be taken. In case of an actual trouble, at best this would cause some noise in the source code history, being fixed before release. At worst a patch will have to be done. This sounds reasonable to me. I agree that on heavy changes, giving more time or explicitly asking for other reviews is a good idea, and should be done. But I let this to contributors own judgement, as it looks hard to define what is a big change. (By example, some changes can impact a lot of files while being trivial, while some other can heavily impact the software while being limited to few files.) Le lundi 22 octobre 2018 11:01:10 UTC+2, Ricardo Peres a écrit : > > But what is most important: the pace or the quality? Is anything in 5.2 > that is so highly demanded that it needs to get it out ASAP? > Jut my 5 cents... > > RP > -- --- You received this message because you are subscribed to the Google Groups "nhibernate-development" group. To unsubscribe from this group and stop receiving emails from it, send an email to nhibernate-development+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
