Re: Patch-needs_work vs. others

[Graham Percival]

On Tue, Oct 25, 2011 at 01:20:33PM +0200, David Kastrup wrote:
> Graham Percival <address@hidden> writes:
> 
> > convenient).  It should only contain patches that have completed a
> > countdown, and/or patches that the author wishes to skip the
> > review process.
> 
> Shrug.  That means to me that this patch is dead.  There is no
> conceivable reason that anybody should change its status to anything
> else.

I already changed the status back to Patch-new when I removed my
objection.

> We have its current state "Patch-new" -> Patch has received no obvious
> checks Of course, I checked the patch.  But even if I decide to put the
> "Patch-review" state on myself, this will merely mean: "Patch has passed
> obvious checks, and needs review".  Well, it is under review.  Who
> should decide to change its need of review, and why?

- if somebody reviews the patch and finds problems (ideally solid
  technical problems), they change it to patch-needs_work.
- if it's still patch-review when Colin makes the next countdown,
  it becomes patch-countdown.

> And in any case, it is _impossible_ to let the patch series get checked
> before having a plan for which version the convertrules.py needs to be.
> After applying the reviewed patch, one needs to autoconvert before a
> check can be made.

Right.  This is a special case of having a collective convert-ly
clustermao.  I started a new email thread for that.  Short answer:
just make a patch that combines all your rules, we'll ram that
through (quite possibly avoiding the review process), and get on
with life because this has become a farce.

- Graham