The old versions will suffer the same issues. This is not something newly
introduced, but rather something fixed in 3.4 and above. Anyway, --vanilla
should always work.
On Tuesday, August 30, 2016 at 11:14:15 AM UTC+9, Jonas Obrist wrote:
>
> Hi everyone,
>
> Over the weekend at PyCon Malaysia I
Il 02/09/2016 10:45, Jonas Obrist ha scritto:
> Managed to get the tests to work reliably (on Py 3.4 and higher).
> See https://github.com/ojii/django-better-test/pull/9
>
> But had to remove Django 1.6 support (don't think too many people will
> get upset about this) and Py 2.7/3.3 are now "unsta
Potentially randomly failing tests on low performance machines due to
unreliable multiprocessing. Though these issues seem to primarily manifest
when doing bad things (leaking tests) our running in parallel mode with
more processes than cores.
I had to give up trying to get the internal tests to p
Hey Jonas,
Thanks for the update.
Just to clarify, when you say "unstable", do you mean random failing tests?
or just unsupported versions?
On Friday, September 2, 2016 at 4:45:47 AM UTC-4, Jonas Obrist wrote:
>
> Managed to get the tests to work reliably (on Py 3.4 and higher). See
> https://g
Hello Tim,
Sorry for trouble.
Following Django practices, we've started to only do *one* commit per PR
which means that we no longer have a change with commits that work and
others that don't.
The big "Bumped version to 3.4.0rc1" commit contains release *only* changes
like PO file changes from
Managed to get the tests to work reliably (on Py 3.4 and higher).
See https://github.com/ojii/django-better-test/pull/9
But had to remove Django 1.6 support (don't think too many people will get
upset about this) and Py 2.7/3.3 are now "unstable". Without going to great
lengths, I don't think I
