I've noticed that MoinMoin seems to make some effort to be backward
compatible with Python 1.5.2. Or maybe I should say: it does a pretty
good job of running on 1.5.2, whether that's purposeful or accidental
I'm not 100% sure. But I did run across one hiccough so far in my use
of the system, in userform.py where a couple of the newer features
from the cgi module are being used.
Is there interest in a patch for this, or has 1.5.2 been officially
abandoned? (Has someone already reported this and my problem would
have vanished where I using a CVS version?)
Geoff Gerrietts <geoff at gerrietts net> http://www.gerrietts.net/
"Now, now my good man, this is no time for making enemies."
--Voltaire, on his deathbed, when asked to renounce Satan
From: Thomas Waldmann <tw@wa...> - 2003-06-11 15:13:23
> I've noticed that MoinMoin seems to make some effort to be backward
> compatible with Python 1.5.2. Or maybe I should say: it does a pretty
> good job of running on 1.5.2, whether that's purposeful or accidental
> I'm not 100% sure. But I did run across one hiccough so far in my use
> of the system, in userform.py where a couple of the newer features
> from the cgi module are being used.
> Is there interest in a patch for this, or has 1.5.2 been officially
Juergen plans to maybe even require Python 2.2 soon, see the plans for
version 1.1 and 1.2 in the MoinMoin wiki.
Why do you use such an old version of python? The 2.2.x versions run
Quoting Thomas Waldmann (tw@...):
> Juergen plans to maybe even require Python 2.2 soon, see the plans for
> version 1.1 and 1.2 in the MoinMoin wiki.
Okay, I'll keep my patches to myself then. I can understand why you
might decide to require a later version; they're much more fun.
> Why do you use such an old version of python? The 2.2.x versions run
> pretty well!
At home, I do use 2.2.x (unfortunately across a few values for "x" at
the moment); at work, our entire production site (some 250k lines of
code) is written against Python 1.5.2 and an old version of Zope.
Migration would cost about 2-3 weeks of an engineer's time, some
project and release management resources, and 2-3 weeks of the entire
QA team's time, to a total bill of around $10,000 - $15,000.
Engineering has successfully argued that this cost must be absorbed,
but engineering has not been successful in demonstrating that this
cost must be absorbed /soon/. I'm reasonably confident we'll be on a
more recent release within 18 months, because we're getting to the
point where pieces our current platform won't run on new hardware.
Now, there's always the option of installing a separate python2, and
maintaining a suite of modules against that. For a long time, when I
was on RedHat 7.x at home, that was the approach I took. I didn't mind
because the 1.5.2 was really only around to keep RedHat happy, and I
was using 2.x for myself. But, at the office, it's usually easier,
less complicated, and harder to explain if I simply back-port new code
to old interpreters. Most python code is less than a couple thousand
lines, and most of the new features that people are using are simply
convenience constructs: the concepts express fine in another 4-5 lines
of code, but recent versions have delivered one-liners.
P.S.: I'm leaving this list after sending this mail; if you intend a
response to get to me, please CC me.
Geoff Gerrietts "Me and my homies, we tag O.D.."
<geoff at gerrietts net> --Unknown grafitti artist at a party