Django 1.6 release timeline

[Jacob Kaplan-Moss]

Hi folks -

Unless there are strong objections, here's what I'm thinking for the
Django 1.6 release timeline:

Alpha: May 16
Beta: June 20
RC: Aug 1
Final: as early as Aug 8, or later if more RCs are needed.

Remember that Alpha is the deadline for "big" features, so if you've
got something sizable that you want to get into 1.6 you'll need to
have it done in a couple of weeks. If it's really awesome but you need
more than a couple weeks, I *might* be convinced to push the deadline
back just a little bit, but please speak up now.

1.6 won't be a huge release -- the new transaction management is the
biggest feature -- but I'd like to get a release out this summer. That
paves the way for work being done over the summer -- Summer of Code
and Andrew's migration work -- to land shortly after 1.6 and ship as
part of 1.7 in early in 2014.

So yeah, objections if you've got 'em, otherwise we'll roll with this.

Jacob

[Donald Stufft]

On Apr 30, 2013, at 3:31 PM, Jacob Kaplan-Moss <ja...@jacobian.org> wrote:

> Hi folks -
>
> Unless there are strong objections, here's what I'm thinking for the
> Django 1.6 release timeline:
>
> Alpha: May 16
> Beta: June 20
> RC: Aug 1
> Final: as early as Aug 8, or later if more RCs are needed.

Sounds reasonable.

-----------------
Donald Stufft
PGP: 0x6E3CBCE93372DCFA // 7C6B 7C5D 5E2B 6356 A926 F04F 6E3C BCE9 3372 DCFA

[Shai Berger]

On Tuesday 30 April 2013, Jacob Kaplan-Moss wrote:
> Hi folks -
>
> Unless there are strong objections, here's what I'm thinking for the
> Django 1.6 release timeline:
>
> Alpha: May 16
> Beta: June 20
> RC: Aug 1
> Final: as early as Aug 8, or later if more RCs are needed.
>

I see one issue with this: According to current procedures, if this timeline
is followed, support for 1.4 will be dropped less than 6 months after the
release of 1.5. At least for some of us (which, as I mentioned earlier on the
list, only moved to 1.4 when the 1.5 release forced us to), this may be a bit
of a problem.

Shai.

[Andrew Ingram]

On 30 Apr 2013, at 23:38, Shai Berger <sh...@platonix.com> wrote:
> I see one issue with this: According to current procedures, if this timeline
> is followed, support for 1.4 will be dropped less than 6 months after the
> release of 1.5. At least for some of us (which, as I mentioned earlier on the
> list, only moved to 1.4 when the 1.5 release forced us to), this may be a bit
> of a problem.
>
> Shai.

It seems like 1.4 support might need to be extended. I'm assuming that given the success of the kickstarter campaign, Andrew's schema migration functionality will be made available to 1.4, and the functionality seems to be a prerequisite for migrating to the new way of handling user models.

Andy

[Donald Stufft]

There is no requirement to migrate for the new way to handle user models. The only time you'd need to migrate is if you want to swap out your existing user models that Django provides with new ones. If you don't do that then you don't need to migrate.

[Andrew Ingram]

Absolutely, there's no requirement to migrate, but it does have the feel of an weird kind of deprecation. Obviously there are workarounds, like you said, I can upgrade to 1.5 without changing user models (though that does involve incurring a bit of technical debt). Anyway, this whole line of reasoning was based on the faulty belief that Andrew's work will be added to 1.4, rather than as an external tool, so never mind!

Andy

[Carl Meyer]

On 04/30/2013 05:27 PM, Andrew Ingram wrote:
> Absolutely, there's no requirement to migrate, but it does have the
> feel of an weird kind of deprecation. Obviously there are
> workarounds, like you said, I can upgrade to 1.5 without changing
> user models (though that does involve incurring a bit of technical
> debt).

I think it's worth being extra-clear about this to avoid scaring people
off of 1.5 upgrades with FUD: there is no deprecation of the built-in
User model in 1.5, and I doubt that there will be in any upcoming Django
version. Neither is there any technical debt incurred by upgrading to
1.5 without switching to a custom User model. Custom user models are a
100% optional feature; continuing to use the built-in User model in 1.5
is not a "workaround" for anything, it's fully supported and reasonable.
If you already have working 1.4 code with the existing User model
there's little reason you'd want to switch over when upgrading to 1.5
(unless you are also adding new User-related features to your codebase).

Carl

[Carl Meyer]

Hi Shai,

Yes, thanks for bringing this up. The core team discussed this at PyCon
and I believe we had general agreement that if we achieve faster
releases we'll need to extend support longer, at least for certain
"long-term support" releases (a category which would probably be
retroactively applied to 1.4). We haven't hammered out the details of
the new policy yet, but I think it's safe to say that if we do release
1.6 on the proposed schedule, we won't drop support for 1.4 at that time.

Carl

[Andrew Ingram]

Don't worry, I'm not misunderstanding you.

Andy

(Pointless aside: My use of technical debt comes from the idea that pre-1.5 I had a lot of technical debt from hacking user models that couldn't be repaid, so in essence it was written off. Then 1.5 suddenly made it possible to repay it. In general, whenever a new preferred idiom comes along, I consider legacy code to have gained technical debt.)

[Andrew Godwin]

I'm happy with this - I doubt schema alteration will be mergeable by the 15th (though I'm trying to get something together for DjangoCon EU) and I want to encourage smaller release cycles. And the new transaction stuff is great, damnit, we should ship that.

Andrew

--
You received this message because you are subscribed to the Google Groups "Django developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to django-develop...@googlegroups.com.
To post to this group, send email to django-d...@googlegroups.com.
Visit this group at http://groups.google.com/group/django-developers?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

[Aymeric Augustin]

2013/5/2 Andrew Godwin <and...@aeracode.org>

get something together for DjangoCon EU

That's a good point. DjangoCon EU sprints are on May 18-19th.

Shouldn't we push the alpha to May 23rd, all the more since we plan to fork the stable/1.6.x branch at the beta?

--
Aymeric.

[Jacob Kaplan-Moss]

On Thu, May 2, 2013 at 4:42 AM, Aymeric Augustin
<aymeric....@polytechnique.org> wrote:
> Shouldn't we push the alpha to May 23rd, all the more since we plan to fork
> the stable/1.6.x branch at the beta?

That sounds reasonable; give us a chance to get any work done at
DjangoCon Europe into the tree before we roll the alpha.

So, revised schedule:

Alpha: May 23
Beta: June 27
RC: Aug 8
Final: as early as Aug 15, or later if more RCs are needed.

(Same as the original, just plus an extra week).

Jacob

[Tim Graham]

Obviously, we've missed our target for the release candidate. We're currently targeting the week of August 26 for that, pending no new release blockers.

You can help us make our goal by triaging unreviewed tickets. Thanks!

[to...@exablox.com]

Tim,

Do you have an update on the schedule for the the Final? I guess it would be any week now, but I have some internal factors that require me to ask for your updated forecast.

Thanks,

Tony

[Aymeric Augustin]

2013/9/11 <to...@exablox.com>

Do you have an update on the schedule for the the Final? I guess it would be any week now, but I have some internal factors that require me to ask for your updated forecast.

Hi Tony,

Since we released 1.6 beta 3 yesterday, the next step is to release 1.6 RC 1. We haven't chosen a date yet, but I would guess next week.

At that point, we'll have to wait two weeks for the translators to take into account new translated strings. Hopefully, after three betas, we won't have a RC 2. This puts 1.6 final in the first days of October.

As usual, take this estimate with a grain of salt. To my great embarrassment, we're very bad at following a release schedule, and security issues can get in the way.

 

--
Aymeric.

[seocam]

Hi,

How are we going with the release process? Do we have a new date for the RC?

Cheers,

--
Sergio Oliveira

[Aymeric Augustin]

2013/9/26 seocam <seo...@gmail.com>

How are we going with the release process? Do we have a new date for the RC?

Well -- security issues got in the way :( and I'm still pondering over a release

blocker: https://code.djangoproject.com/ticket/21134

Hopefully I'll deal with it during the week-end. Then the release managers will

be able to propose a new date for 1.6 RC1.

 

--
Aymeric.

[Aymeric Augustin]

Several release blockers were discovered over the last two weeks.

Most were fixed, two are still open:
https://code.djangoproject.com/ticket/21197
https://code.djangoproject.com/ticket/21164

At this point the ETA for 1.6 RC1 is really "once these tickets are fixed".

Thanks for your understanding,

--
Aymeric.

[Russell Keith-Magee]

On Sun, Oct 6, 2013 at 4:31 AM, Aymeric Augustin <aymeric....@polytechnique.org> wrote:

Several release blockers were discovered over the last two weeks.

Most were fixed, two are still open:
https://code.djangoproject.com/ticket/21197
https://code.djangoproject.com/ticket/21164

At this point the ETA for 1.6 RC1 is really "once these tickets are fixed".

I've committed a fix for #21164, and on reflection, #21197 isn't a release blocker for 1.6. It has some implications for 1.7, but they'll come up in the process of me working on the GSoC merge.

So… looks like we're ready for an RC!

Russ %-)

[seocam]

Hey guys! Any updates?