[petsc-dev] does next model mess up our histories

[Satish Balay]

On Thu, 2 Oct 2014, Jed Brown wrote:

> Satish Balay <balay at mcs.anl.gov> writes:
> 
> > On Thu, 2 Oct 2014, Jed Brown wrote:
> >
> >> Satish Balay <balay at mcs.anl.gov> writes:
> >> > I think is ok to be a bit more messy in next. Sure there is a tradeoff
> >> > - but you get better debugging in master :)
> >> 
> >> If bugs make it to 'master', the process has failed.
> >
> > If we have to do patch releases - the process has failed?
> 
> Of course.  Every bug is a failure of internal quality control.  It's
> not possible to avoid bugs entirely, but we should strive to catch them
> early.  I.e., saying that we won't worry about bugs in 'next' because we
> can fix them in 'master' defeats the purpose of the workflow.

I didn't advocate not debugging bugs in next. All I adovcated is: its
ok to have a few cases (not allways) reverts (of feature branches) in
next that does affect bisection a bit [but I don't think by much - as
this is usually recent history debugging] - in favor a cleaner history
that helps in long history debugging later on..

Satish