On 31 Okt., 02:14, Francois Bissey
wrote:
> > On 30 Okt., 21:13, Nils Bruin wrote:
> > > On Oct 30, 10:12 am, Francois Bissey
>
> > > wrote:
> > > > There is also the fact that a number of packages check:
> > > > sys.platform == 'linux2'
>
> > > One quick data point: on FC 15 they are shipping l
> On 30 Okt., 21:13, Nils Bruin wrote:
> > On Oct 30, 10:12 am, Francois Bissey
> >
> > wrote:
> > > There is also the fact that a number of packages check:
> > > sys.platform == 'linux2'
> >
> > One quick data point: on FC 15 they are shipping linux 3.* kernels
> > with a fake 2.* version numb
On 30 Okt., 21:13, Nils Bruin wrote:
> On Oct 30, 10:12 am, Francois Bissey
>
> wrote:
> > There is also the fact that a number of packages check:
> > sys.platform == 'linux2'
>
> One quick data point: on FC 15 they are shipping linux 3.* kernels
> with a fake 2.* version number exactly for compa
> On Oct 30, 10:12 am, Francois Bissey
>
> wrote:
> > There is also the fact that a number of packages check:
> > sys.platform == 'linux2'
>
> One quick data point: on FC 15 they are shipping linux 3.* kernels
> with a fake 2.* version number exactly for compatibility reasons (I
> don't think it
On Oct 30, 10:12 am, Francois Bissey
wrote:
> There is also the fact that a number of packages check:
> sys.platform == 'linux2'
One quick data point: on FC 15 they are shipping linux 3.* kernels
with a fake 2.* version number exactly for compatibility reasons (I
don't think it's just python). I
The "good" thing is that sys.platform reports the [major] version
Python was *built* on, so our binary distributions (if built on a 2.x
kernel) won't break on 3.x Linux kernels, at least not because someone
uses sys.platform=="linux2" to detect Linux.
The suggested idiom to use instead is sys.plat
