Re: RFC: Guide for maintaining software in the Zope repository

On 24 Aug 2007, at 15:25 , Stephan Richter wrote:
> On Thursday 23 August 2007 20:37, Philipp von Weitershausen wrote:
>> I would like to get your comments on it. No matter what this  
>> evolves to,
>> I wouldn't mind eventually seeing it set in stone with your  
>> blessings,
>> so that the checkin police can use it as the highway code to issue
>> tickets to anyone who's speeding on the repository lane.
>
> I don't like the section on coding style. A while back we agreed  
> that people
> can choose it freely as long as every package in the *namespace*  
> has the same
> style.

Hmm, I recall having discussed this on the list at some point. I  
don't recall having reached an agreement. That could be me, though...

> So for example, ``zope`` and ``z3c`` use the original Zope 3  
> styleguide, while ``zc`` uses PEP8 compliance.

So does this mean I'll can't put my stuff in the 'z3c' namespace if I  
want to use PEP8 (which I do, not because of personal preference but  
because of consistency)?

> This is much easier to keep track of than having to remember every  
> package's
> style.
>
> I personally do not like underscore-style method naming, so I would  
> never use
> it for packages that I am starting from scratch.

This and other aspects are things I don't particularly love about  
PEP8 either, but I value consistency over my personal preferences.

> I do honor other people's
> decisions though, and would always follow the original author's style.
> Consistency is better than correctness in this case. (I usually  
> tend to value
> correctness higher than consistency.)

Well, this may sound harsh, but I see some appeal in actually forcing  
a particular coding-style on everybody. It's soo late for anything  
that has been started already, but I don't see a reason why we simply  
can't say:

   If you start a new project on svn.zope.org, it'll have to be in  
PEP8 styling.

The rule being behind this (as already mentioned above), that  
consistency values higher than personal preferences.