Open Stack

Flavio Percoco wrote:
> On 02/05/13 14:57 -0700, Clint Byrum wrote:
>> The removal would be all pain for almost no gain. In fact, I would
>> think that the stable maintenance team would appreciate that the code
>> is more like trunk when they have to backport patches in areas
>> adjacent to experimental code paths. The inverse is true too (if the
>> experimental bit has been made the default, then they'll have more
>> burden). But I think we can expect this to be at worst a tie in terms
>> of maintenance burden.
> 
> My concern is about features that, for some reason, never get fully
> implemented.
> Is that something being considered here?
> 
> Wouldn't that require to remove the feature anyway?
> What's the maximum time allowed for a feature to be fully implemented
> before it is removed from the tree?
> 
> If we let incomplete features to land during a release cycle, I'm +1
> for removing them before the release if they haven't been fully
> implemented.

There are two cases: incomplete features (where time has run out and a
release is cut before the feature can be "enabled"), and abandoned
features (where nobody is working on that feature anymore, so it will
never be completed).

I suspect that Clint is advocating for not removing incomplete features
from release branches (to reduce last-minute disruption caused by
ripping out stuff, and minimize master-to-stable backport costs), but
that he would be OK to remove code for abandoned features from the
master branch once we are sure it's abandoned.

-- 
Thierry Carrez (ttx)
Release Manager, OpenStack