[OE-core] [PATCH 2/2] python-six: put python2/3 variant together

richard.purdie at linuxfoundation.org richard.purdie at linuxfoundation.org
Tue May 21 18:38:43 UTC 2019


On Tue, 2019-05-21 at 20:00 +0200, Andreas Müller wrote:
> On Tue, May 21, 2019 at 6:41 PM Richard Purdie
> <richard.purdie at linuxfoundation.org> wrote:
> > On Tue, 2019-05-21 at 11:00 +0200, Andreas Müller wrote:
> > > Upcoming 60.2.3 mozjs (meta-oe) requires python(2)-six-native for
> > > its
> > > configuration. Currently python(2)-six lives in meta-python and
> > > it
> > > could be
> > > moved to mozjs's home meta-oe but that would be just further
> > > clutter.
> > > 
> > > Once applied here patch to meta-python will be send.
> > > 
> > > Signed-off-by: Andreas Müller <schnitzeltony at gmail.com>
> > > ---
> > >  meta/recipes-devtools/python/python-six_1.12.0.bb | 2 ++
> > >  1 file changed, 2 insertions(+)
> > >  create mode 100644 meta/recipes-devtools/python/python-
> > > six_1.12.0.bb
> > > 
> > > diff --git a/meta/recipes-devtools/python/python-six_1.12.0.bb
> > > b/meta/recipes-devtools/python/python-six_1.12.0.bb
> > > new file mode 100644
> > > index 0000000000..9075745d74
> > > --- /dev/null
> > > +++ b/meta/recipes-devtools/python/python-six_1.12.0.bb
> > > @@ -0,0 +1,2 @@
> > > +inherit pypi setuptools
> > > +require python-six.inc
> > 
> > Can it not use python3? Python2 is rapidly running out of time at
> > this
> > point and I'd not expect new dependencies on python2 to be
> > appearing...
> > 
> > Cheers,
> > 
> No sorry mozjs uses old crap in their configurations scripts nobody
> else does (python which) and ask for python2 explicitly.

I've been giving this a bit more thought and its our intent to try and
remove python2 from OE-Core entirely so adding py2 modules back seems
like a backwards step.

I think this will therefore need to go to one of the other layers,
painful as that might be from an include perspective.

Upstream are only going to keep py2 compatibility to the end of the
year, if that so the recipes will have to fork sooner or later anyway.

Cheers,

Richard





More information about the Openembedded-core mailing list