Open Stack

Thanks for the prompt response John!

On 02/14/2013 06:09 PM, John Dickinson wrote:
> 
> On Feb 14, 2013, at 1:44 PM, Sandy Walsh <sandy.walsh at RACKSPACE.COM>
> wrote:
> 
>> Hey!
>> 
>> I'm working on getting CM HACKING Compliant and ran into this
>> little gem: 
>> https://github.com/openstack/ceilometer/blob/master/ceilometer/objectstore/swift_middleware.py#L34
>>
>>
>> There's no easy way to have the correct order of imports and these tests
for specific Swift client versions.
>> 
>> I could disable Hacking checking for this file, but that feels
>> wrong. It could be easier to just not support Swift clients < 1.7.5
>> 
>> 
>> So, which option is best? 1. Drop support for Swift < 1.7.5 in CM?
> 
> Since CM hasn't been released as part of OpenStack (or ever?) yet,
> this seems like a viable alternative, IMO. Of course, this limits the
> existing clusters you can deploy CM in. Although we recommend that
> all deployers upgrade to each Swift release, I know it doesn't always
> happen.
> 
> FWIW, Swift 1.7.5 was released Tue Nov 13 (ie post-folsom). We have
> had once release since then and will have another before grizzly is
> final.

Cool ... are there any radical new changes in 1.7.5 that might
stall/prevent someone from upgrading?

>> 2. Solicit ideas for crafting an elegant way to make that code
>> Hacking Compliant? 3. Disable Hacking checks for this file?
> 
> I'm not familiar with what a hacking check does, but it seems to
> depend on the answer to 1.

It's the Nova pep8 style additions
https://github.com/openstack/nova/blob/master/HACKING.rst

> 
>> 
>> Thoughts? -S
>> 
>> 
>> 
>> _______________________________________________ OpenStack-dev
>> mailing list OpenStack-dev at lists.openstack.org 
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
> 
> 
> _______________________________________________ OpenStack-dev mailing
> list OpenStack-dev at lists.openstack.org 
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>