[Idle-dev] IDLE Unittest framework: Commencing Initial Design

[Terry Jan Reedy]

On 5/13/2013 4:04 PM, francis wrote:
> On 05/13/2013 08:19 AM, Terry Jan Reedy wrote:
>> I still prefer idlelib/itest (better than 'test' to avoid masking
>> Lib/test when in idlelib) for multiple reasons and have a draft
>> explanation of them, which will accompany the patch.

> That's a bit weird Terry,

I think it a bit weird that you say this without waiting for and reading 
the promised explanation.

 > why not to follow "Special cases aren't
> special enough to break the rules."?

PEP 434 recognizes Idle as a special case that *is* special enough to 
break the rule for changes in bugfix releases. The same specialness is 
part of my reason and explanation.

tjr