Open Stack

On 04/28/2015 01:17 PM, Neil Jerram wrote:
> Apologies for commenting so late, but I'm not clear on the concept of
> bringing all possible backend projects back inside Neutron.
> 
> 
> I think my question is similar to what Henry and Mathieu are getting at
> below - viz:
> 
> 
> We just recently decided to move a lot of vendor-specific ML2 mechanism
> driver code _out_ of the Neutron tree; and I thought that the main
> motivation for that was that it wasn't reasonably possible for most
> Neutron developers to understand, review and maintain that code to the
> same level as they can with the Neutron core code.
> 
> 
> How then does it now make sense to bring a load of vendor-specific code
> back into the Neutron fold?  Has some important factor changed?  Or have
> I misunderstood what is now being proposed?

The suggestion is to recognize that these are all part of the larger
Neutron effort.  It is *not* to suggest that the same group of
developers needs to be reviewing all of it.  It's mostly an
organizational thing.  The way teams operate should not change too much.

-- 
Russell Bryant