Jeff Shell wrote:
On 9/28/06, Martijn Faassen <[EMAIL PROTECTED]> wrote:
Jim Fulton wrote:
Baiju M wrote:
Hi, What is the target Python version for Zope 3.4, is it
Python 2.5?
That's a good question. I fear it will take a fair bit of work
to get to it and, frankly for me there are higher priorities.
I think we'd be okay in not supporting Python 2.5 yet for this
release. Python 2.5 has just been released, and the 9 months or so
will give it time to gell out, Python extensions a bit of time to
catch up with the newer version, and for us to do some
experimenting. Linux distributions will likely still be supporting
Python 2.4 (along with 2.5) at that stage as well.
Are you talking not-supporting Python 2.5 at all? Or just not wanting
to require it?
I'm talking about not supporting it at all. As I just posted elsethread,
I do not think we should spare the development resources to make this
work. It's a lots of added overhead to the process.
If Zope 3 can't at least run on Python 2.5 within three months of
Python 2.5's release, I don't think that looks good for us. I would
hope that by being more "Pythonic" and free from the old
ExtensionClass magic - especially in Zope 3 - that Zope could keep
better pace with Python. But nine months? A year? Can I expect the
developers of toolkits we use, in some cases more heavily than Zope
itself, like SQLAlchemy to hold off on requiring Python 2.5 for that
long?
I don't know. All I know is we can't spare the resources to support
both. Either we go for Python 2.5 completely or we don't at all this
release cycle. I must also note that many Linux distributions also have
half-year or longer packaging cycles. The new Ubuntu is surely not going
to have much Python 2.5 in it.
So unless a volunteer steps up to do lots of hard work between now
and march next year, let's stick with Python 2.4. Otherwise let's
plan it in for Zope 3.5 and Zope 2.12
I really really really hope it doesn't take that long to be able to
at least run on Python 2.5: even if it has to be with some caveats or
mild warnings.
If it works that way without the core developers having to do any
additional work, then your wish may become true. That's a far cry from
official support, though.
If security and restricted python / security proxies are the main
issue, what about if one is running Zope sites with absolutely ZERO
through the web code - no page templates, nothing - can't there be a
lighter weight security implementation that wouldn't take half a year
of "lots of hard work"?
It's going to be a significant investment of work to make sure
everything works with Python 2.5. Either someone credible steps up,
convinces us, and does the work, and we all move over to Python 2.5. If
they don't, we do it the next cycle.
Regards,
Martijn
_______________________________________________
Zope3-dev mailing list
Zope3-dev@zope.org
Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com