Noosfero 1.1 delayed, and thoughts for the 1.2 cycle

[Antonio Terceiro]

As you may have noticed by now, The Noosfero 1.1 release is now
officially delayed. There are still failign tests, including a few
regressions in the last days:

http://ci.noosfero.org/

Until we are green, we won't have a release. Let me rehash one of the
items in our development policy:

  * The release schedule may include a period of feature freeze, during which
    no new features or any other changes that are not pre-approved by the
    release manager must be committed to the repository.

We had too many changes that fall out of this category during this
cycle, so let's us all please stop. If you commit something that _I_
didn't say it's OK to commit, you will get yelled at.

And next week I will be travelling and will have little to no time for
Noosfero. Unless we all work together to make the tests green the
release will be delayed even more.

I am thinking about how to mitigate this for the next cycle, and
considering changing the rules to include stuff like:

- no MR can be merged without a passing CI run
  - this will involve a certain effort to setup proper CI runners that
    integrate with gitlab (yes I am aware of the tests the Braulio
    recently did but still didn't have the chance of setting up a
    runner)
- commits that break the tests get imediately reverted, no matter what.
- people who break the tests twice in less than N days lose their commit
  rights

If we implement some of the above, maybe Noosfero 1.2 can have a very
short freeze, or even none at all.

-- 
Antonio Terceiro <terceiro at colivre.coop.br>
Colivre - Cooperativa de Tecnologias Livres
http://www.colivre.coop.br/


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://listas.softwarelivre.org/pipermail/noosfero-dev/attachments/20150402/5695901c/attachment.pgp>