Newer Debian versions of python on older Debian distros?

Rustom Mody rustompmody at gmail.com
Tue Sep 9 11:23:02 EDT 2014


On Tuesday, September 9, 2014 2:53:53 PM UTC+5:30, Chris Angelico wrote:
> On Tue, Sep 9, 2014 at 6:09 PM, Rustom Mody  wrote:
> >>  Does anyone have experience with using newer versions of python
> >> debian packages (in particular, python3 and python3-bson-ext from
> >> 'testing') on older stable versions ('wheezy' in this case)? If
> >> someone's figured out how to do this easily, I'd love to hear the
> >> recipe!
> > Wheezy appears to have a python3 (though not the latest)
> > https://packages.debian.org/wheezy/python3

> I think the point of "python3 from testing" is because the python3
> package in Wheezy is 3.2.3. (And if he hadn't explicitly told us he's
> using Wheezy, it could have been Squeeze, which went out of primary
> support just a few months ago, and is still in Long-Term Support for a
> couple of years. Squeeze ships Python 3.1.)

> All my Wheezy systems have a locally-compiled Python. But then, I've
> been installing quite a few Jessie (testing) systems, for various
> reasons (support for our network scanner being one of them), and the
> one really important Wheezy system here is my personal dev system
> where I worked on the PEP 463 branch, so compiling CPython from source
> was absolutely necessary. :)

> > Chris said:
> >> Alternatively, you could just run Debian Jessie. I have a few Jessie
> >> systems on the network, with a Python 3.4 IIRC, and there've been no
> >> stability problems lately. Both options are pretty easy.
> > I'm not so sure.
> > There's quite a brawl going on right now on debian users over
> > systemd.
> > [I am running testing myself]

> Sadly, yes. I wish these things could be resolved on technical grounds
> rather than political. I'm certain that systemd is superior to
> sysvinit; I'm fairly sure it's superior to Upstart, and others I don't
> have experience with. The technical downsides are few - it's
> Linux-only (or was last I checked - this stuff can change), and it's
> fairly invasive, needing kernel support. Most of the issues are
> political ("how much power will Red Hat have?") and I have no interest
> in arguing those.

> But frankly, that's not really much different from the OpenOffice vs
> Libre Office battles. Most people just don't care. I mean, let's face
> it, there are a lot of people who wouldn't care (and maybe wouldn't
> even notice) if you just go in and rename all the menu items to
> "Microsoft Excel" and "Microsoft Word" and so on, and claim it's a
> port of MS Office! If the average user doesn't care about an
> application that's right in the face, why will s/he care about an init
> system? "This one means the system boots faster." "Fine! Good enough
> for me."

I thought so too viz that the problems were teething troubles.
However the rants on debian-dev seem to be following from extensive breakage
from systemd.

Also there is this thread in which systemd broke the standard
kernel debugging options -- ok bugs happen.
And then the systemd devs refuse to admit to their bug. I find this alarming.



More information about the Python-list mailing list