4.5.3

Dirk Hohndel dirk at hohndel.org
Mon Dec 7 09:27:05 PST 2015


On Mon, Dec 07, 2015 at 06:38:21PM +0200, Lubomir I. Ivanov wrote:
> On 7 December 2015 at 18:01, Miika Turkia <miika.turkia at gmail.com> wrote:
> >
> >
> >> On 07 Dec 2015, at 17:55, Lubomir I. Ivanov <neolit123 at gmail.com> wrote:
> >>
> >>> On 7 December 2015 at 17:45, Dirk Hohndel <dirk at hohndel.org> wrote:
> >>> No comments from anyone? No nominations for additional commits?
> >>
> >> hmm, is cherry-picking really required - e.g. are there features which
> >> are WIP ATM?
> >> why not include all commits between the last version and this one?
> >
> > filtering is currently broken
> 
> *scratches head*
> well, if there are no regressions in master and if there are no WIP

Yes, filtering is broken. That's a regression.

> features - then cherry picking is redundant and all patches between
> ver X and ver Y can be included.
> too optimistic.

My usual approach for the dot releases is to pick things that are clearly
fixes plus small, self contained things that are clearly improvements but
appear very localized (that's why I picked you NSIS change - clearly
missing that it depended on another patch which I'll now pick as well).

But I try to shy away from picking random "warning fixes" or "restructure
cmake" or "move things around" patches that don't add anything of value
for a dot release but increase the risk of something going wrong.

That's why I picked what I picked. I'll add the one missing commit that
you pointed out, plus the fix for the heartbeat.

/D


More information about the subsurface mailing list