[rfc][patch]

[Robey Pointer]

On 2 Jan 2006, at 16:06, Robert Collins wrote:

> On Mon, 2006-01-02 at 15:28 -0800, Robey Pointer wrote:
>>
>> Sadly, it turned out that python reverses the argument order from c
>> ++/
>> java, so the first version (actual, expected) is correct.  (Sigh.) :(
>
> *our* tests are not pythons tests.
>
> I think its fine for us to use (expected, actual) everywhere.

Sorry, what I meant was that the python unittest library assumes  
you're using (actual, expected) format, and uses this assumption in  
its error reporting.

I agree that it doesn't really matter.  It's just lame that they  
decided to go their own way -- no doubt this will cause me to second- 
guess myself every time I write unit tests now, since I still  
regularly use C and java. :)

robey