[Qgis-psc] QGIS 2.18 bug fixing effort

[Nyall Dawson]

On 23 May 2017 at 18:11, Jürgen E. Fischer <jef at norbit.de> wrote:
> Hi Paolo,
>
> On Tue, 23. May 2017 at 09:13:23 +0200, Paolo Cavallini wrote:

>> Is there a serious lack of tests in this area to prevent these problems to go
>> unnoticed?
>
> People should test the nightlies - esp. before releases - and not wait for the
> new release.

+1

> Anyway, that wouldn't have helped in this case either as the offensive fix was
> backported on release day.
>

On this note, I'd like to float the idea of some "thaw" time (say a
week) from when a release is tagged before we actually finalise the
release. I realise this adds administrative burden (mostly to Jürgen,
sorry) and I'd welcome suggestions on how we could avoid this. But in
general I think some short delay between when we consider a release is
ready and when it actually becomes packaged and available would give
time for testing the final pre-release candidate before the release is
made.

Regardless of how extensive our test suite becomes, there'll always be
the opportunity for bugs like this to slip in unnoticed. I can't think
of any other alternative approach besides this proposed "thaw" which
could avoid this situation in future.

Nyall