[Qgis-psc] Working on pull requests

[Sandro Santilli]

On Thu, Sep 10, 2015 at 12:16:27PM +0200, Matthias Kuhn wrote:
> On 09/10/2015 11:11 AM, Sandro Santilli wrote:

> > Something simple for a start could be:
> >
> >  "incomplete" for PR that are are missing a testcase or otherwise
> >               known to be a work in progress
> >  "ready" for PR that are formally ready to be merged
> >
> > What do you think ?
> >
> > --strk; 
> I am in favor of using label, they are a great way to classify things.
> It's just that the last time I introduced labels they were deleted and I
> am not aware of the reason.
> 
> I think it's good to be able to assign "incomplete" but I am not sure of
> "ready". Labels can only be assigned by core-committers and if one of
> them thinks it's ready he can as well just merge it.

I was thinking "ready" as in "ready for review" by someone with
experience in the field the patch is touching. A kind of second level
review:

 0. PR is submitted
 1. PR is formally valid (does not contain a zillion commits or
    back-meges, includes testcases, makes travis happy)
 2. PR is reviewed (confirmed to be good to merge)

That way a "ticket manager" with no need to know all the code details
might only deal with setting the "incomplete" or "ready" (or maybe
renamed to be less confusing) labels to either get the contributor to
deal with incompleteness or get a reviewer assigned for further work.

--strk;