[Zope-dev] Zope 2.12 features

[Tres Seaver]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Andreas Jung wrote:
> On 30.10.2008 21:41 Uhr, Tres Seaver wrote:
> 
> What are our current needs for having a Python 2.5 compatible Zope 2 
> version (except it would be nice for having one). Being in sync with the 
> latest Python 2.6 version is much more important than jumping on a 
> half-dead horse like Python 2.5. Introducing another major Zope release 
> with very little new exciting feature does not make sense to me at the 
> time right now. It just adds another major release we have to support 
> (we already support 2.9-2.11 and a bit 2.8) + the complexity for 
> supporting two different Python versions for Zope 2.12. I am open to 
> arguments but I really want to see why we need Zope 2.12 with Python 2.5 
> support this year (or at least very soon).
> 
>>> In that case Plone will neither use 2.11 nor 2.12 but go straight for a
>>> Zope 2.13 including Python 2.6. A major release every six month would be
>>> desirable for us in that case. Right now I don't see anyone, who would
>>> be using those releases. If those people exist, please speak up.
>> Plone isn't the only consumer of Zope2, although it is clearly the
>> biggest one.
> 
> Tres, what are you current needs and requirements (properly based on 
> your Repoze project)? I agree that Plone isn't the only consumer but I 
> wonder if all other consumers really have the need jumping on every 
> train passing the train station. We have the luxury with four supported 
> Zope 2 major release. I don't want a fith right now unless we are having 
> very good reasons.

I think we need to move toward 2.6 compatibility, but we need to give
people a migration path, largely because 2.6 will break 3rd party apps
in ways that 2.4 doesn't warn about.  So, I would like to see a 2.12
which is explicity about bridging first to 2.5 support:  that way,
people get a chance to clean up the new deprecation warnings (e.g., for
the 'with' keyword, etc.).

If ZODB 3.9 lands in time, then a near-term release of Zope 2.12 could
be this consolidation release (2.5 support, new ZODB, including maybe
RelStorage, other work done to date).

We could the focus trunk development on 2.6 compatibility, with the goal
of releasing a 2.13 no later than Q3 next year.

In the meantime, we can acknowledge that 2.8 and 2.9 are "retired" (no
future work except maybe important security fixes), and announce that
2.10 will be retired after the 2.12 release:  that may be incentive
enough for Plone 3.3 to ship on 2.12, for instance.

WRT repoze:  except for the 'repoze.zope2' and 'repoze.plone' packages,
we already have 2.5 / 2.6 support in hand, and in fact are deploying
non-Zope2 customer applications using both versions.

>  >Keeping an orderly succession of releases with good
>> compatibility is important for the whole ecosystem (frankly, Plone
>> should be willing to move to newer Zope versions even in a second dot
>> release, but that is another debate).
> 
> This is already the case (more or less). In my experience strategic 
> consumers are possibly more interested in slower release cycles instead 
> of getting major releases very often. I doubt that much people care 
> about using Python 2.4 or Python 2.5 (I personally don't care much about 
> Python 2.5)..other feature likes e.g. a new ZODB version with some cool 
> new feature is more important for justifying a  new major release.
> 
> If Sidnei should be successful with making Plone 2.11 compatible with 
> Python 2.5 then we might add "inofficial" support for Python 2.5 to the 
> current Zope 2.11 release...but as stated earlier I would like to see 
> some arguments why Python 2.5 compatiblity is necessary now and why 
> Python 2.6 support at some time next year would not be good enough.

I don't see a win there, myself:  I'd rather make an easier transition
for 2.12 than spend resources on a potentially-destabilizing backport of
2.5 compatibility.


Tres.
- --
===================================================================
Tres Seaver          +1 540-429-0999          tseaver at palladion.com
Palladion Software   "Excellence by Design"    http://palladion.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFJC0Ia+gerLs4ltQ4RAv8MAKDJAgpzus+Oh86aH0RgbGEXh26EDgCfQUL/
cnb/9SyKbWUg/JwqmC7tv5w=
=qvj8
-----END PGP SIGNATURE-----