Martin Sebor wrote: > I think the discussion has wound down so let's have a vote and > decide whether stdcxx committers should follow the Commit-Then > Review (CTR) or Review-Then-Commit (RTC) policy on stdcxx/trunk > by default. > > [+1] All committers follow Commit-Then-Review for safe changes, > Review-Then-Commit for potentially breaking changes. What > constitutes a breakng change is a judgment call to be made > by each committer. The rule of thumb is that changes which > involve advanced/sophisticated C++ features, C99 features, > or optional POSIX (and other) extensions may cause breakage > on some platforms. > > [+/-0] All committers follow Review-Then-Commit for all changes > with no exceptions. > > [-1] New committers and committers who haven't been active in > more than [ ] months follow the RTC policy for at least > [ ] weeks of sustained activity, all others CTR. > > Please check the box you're voting for (or put +1/-1 in any or > all of them). When voting for #3, also put numbers in the two > [ ] boxes to complete the rules.
I was concerned about the legitimacy of the last bullet, but the way in which you composed #3 leads me to merely disfavor it, not object to it as an invalid option. If the group accepted it, I don't think that a neutral, well-stated policy like this would be invalid to the ASF. Bill
