I checked the code of bert and the order is just important for
instance variables
now I want to see if this fix the problem is see often.
Now I checked and there was no change in the load semantics of MC
since 3.9
in squeak as well as in pharo so I do not see why we could not load
closure fixes.
probably it needs some care but not a change from MC
Stef
> Hi Stef,
> does that changes implies that the ordering of class variables
> declaration is now important?
> The apparent problem here is, that class vars held by dict, which is
> inherently unordered.
>
> 2009/9/10 Stéphane Ducasse <
[hidden email]>:
>> I'm currently intergating changes to fix that. Apparently bert
>> already
>> fixed the missing part.
>>
>> Stef
>> On Sep 10, 2009, at 8:32 PM, Stéphane Ducasse wrote:
>>
>>>
>>>
>>> Hi
>>>
>>> even if we integrated the fixes of eliot to fix the ordering class
>>> var
>>> by MC I still see this behavior: from time to time
>>> I get more packages touched just because the order of classVar
>>> changes.
>>> And now this is a problem when trying to integrate the last fixes
>>> issued for the block closure.
>>> So I would really appreciate if someone would give a try to fix
>>> that.
>>> I just have the time to integrate some pending fixes.
>>>
>>> Stef
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Pharo-project mailing list
>>>
[hidden email]
>>>
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>>
[hidden email]
>>
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project>>
>
>
>
> --
> Best regards,
> Igor Stasenko AKA sig.
>
> _______________________________________________
> Pharo-project mailing list
>
[hidden email]
>
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Locked
