Open Stack

On Nov 19, 2012, at 4:28 PM, Brian Waldon wrote:

> 
> On Nov 19, 2012, at 4:16 PM, Chuck Short wrote:
> 
>> On 12-11-19 06:40 PM, Joe Gordon wrote:
>>> 
>>> 
>>> On Mon, Nov 19, 2012 at 3:06 PM, Monty Taylor <mordred at inaugust.com <mailto:mordred at inaugust.com>> wrote:
>>> 
>>>   I think these are orthogonal. In general I believe we're in
>>>   agreement that we should be targetting the same version of pep8,
>>>   and that we should pin that version early in the cycle. However we
>>>   install that version of pep8 is a different argument.
>>> 
>>> 
>>> Targeting the same version of pep8 is only part of the problem, we would also need to agree on what rules to ignore. For example:
>>> 
>>> glance (pep8 1.3.3) ignores: E125,E126,E711,E712
>>> nova (pep8 1.2) ignores: E12,E711,E721
>>> keystone (pep8 1.3.3) ignores: ---
>> 
>> Is there a reason why these were ignored?
> 
> I just ran glance's pep8 check without ignoring any errors, and it appears most of them are just arbitrary spacing. I'll attempt to minimize the number of pep8 violations we ignore.

After walking through the type of errors we're ignoring, I'm not actually comfortable removing any of the PEP8 violations (E125, E126, E711, E712) from Glance's ignore list. The first two are quite fragile and enforce a somewhat awkward coding style, while the second two are asking us to write code that is incompatible with sqlalchemy. Hopefully the maintainers will help improve this situation.

Waldon