[V3dot9] Re: About condenseChanges and Traits

[Stéphane Ducasse]

> What we did in 7053 is *condenseSources*.
in 7054 condenseSources not 7053
because in 7053 we just removed the FFI and the condenseSources was  
done in 7054


> A condenseChanges
> was *never* done, as far as I know, so i wonder why why talk about  
> that?

Thanks
Oops I was wrong because I read the tests.

By the way I did a condenseChanges in 7053 and it breaks my tests.
I'm redoing it with the tests of philippe.


> So what needs to be done is
>
> 1) Verify that condenseChanges is correct.
> For that
> 	- take 7052, do condenseChanges
>        - take Philippe's code, see if problem.

I did that for 7053 and my test does not work.
Now redoing it with the tests of philippe (of course each time from a  
fresh unzipped image/changes couple)

> I think the result will be that * condenseChanges*
> is working correctly.
>
> 2) Verify that CondenseSources is broken
> 	- take 7052, do condenseSources
>        - take Philippe's code, see if problem.
>
> Then we are sure what we need to discuss... is
> condense*Changes* or  condense*Sources* broken
> or both?
>
>> On 24 août 06, at 07:16, Philippe Marschall wrote:
>>
>>> problems := Dictionary new.
>>> Smalltalk allClassesAndTraitsDo: [:class |
>>> {class. class class} do: [:behavior |
>>> behavior selectorsDo: [:selector |
>>> (behavior sourceCodeAt: selector) ifNil: [
>>> problems add: behavior -> selector ] ] ] ].
>>> ^problems
>>
>> What I'm looking for is a way to avoid to do a condenseChanges to  
>> be able to test
>> what is breaking the condenseChanges.
>>
>> Stef_______________________________________________
>> V3dot9 mailing list
>> V3dot9 at lists.squeakfoundation.org
>> http://lists.squeakfoundation.org/mailman/listinfo/v3dot9
>