On Wed, Jun 04, 2014 at 09:27:03AM +0100, Garth N. Wells wrote: > On Wed, 4 Jun, 2014 at 9:22 AM, Anders Logg <[email protected]> wrote: > >How do we handle small bug fixes, like for example a small change > >proposed on the list yesterday about removing a debug print statement > >when extrapolating functions. > > > >It would take me less than a minute to fix if I fixed it directly in > >`master`, but 1-2 days if I first need to create an issue + a branch > >for the issue, then fix it, test it, merge into next, wait one day and > >then merge into master. > > > >The overhead for following our guidelines is quite substantial in this > >case, so in my opinion it's better to risk breaking `master` for > >quick fixes like this one, especially since such small changes can be > >easily reverted. > > Since we started with the current work flow, we've made small > changes in master. For me, small changes that are unlikely to break > anything or generate any controversy should be made in master.
Great. -- Anders _______________________________________________ fenics mailing list [email protected] http://fenicsproject.org/mailman/listinfo/fenics
