Hi,

Why should it be in Object?� Can't it be placed in ClassDefinition maybe?�
Ok, i should look what the compiler does, it's just a suggestion.

I think we should avoid putting more weird isXXX methods in Object :).

On Tue, Jun 29, 2010 at 11:02 AM, Alexandre Bergel <alexandre@bergel.eu> wrote:
Hi!


SLICE-Issue-2598-LoadingTraitCompositionFromMonticello-Alexandre_Bergel.1 in the inbox
-=-=-=-=-=-=
Monticello can now load a trait composition with exclusion from Monticello. A test has been added for this.
-=-=-=-=-=-=

Something else, I would like to introduce Object>>isTraitComposition. This will greatly simplifify MCClassDefinition>>createClass. The following code is really ugly:

�composition := Compiler evaluate: self classTraitComposition.
� � � � � � � � �((((composition isCollection and: [ ( composition includes: nil ) not])
� � � � � � � � � � � �or: [composition isKindOf: TraitComposition ])
� � � � � � � � � � � � � � � �or: [composition isKindOf: Trait ])
� � � � � � � � � � � � � � � � � � � �or: [composition isKindOf: TraitTransformation ])
� � � � � � � � � � � �ifTrue: [ class class setTraitComposition: composition asTraitComposition ]

Is that ok if I had Object>>isTraitComposition in the Trait package?

Cheers,
Alexandre



On 29 Jun 2010, at 13:04, St�phane Ducasse wrote:

> alex can you have a look at it?
> http://code.google.com/p/pharo/issues/detail?id=2598
>
>>>
>>> Stef
>>>
>>> On Jun 28, 2010, at 10:45 PM, Simon Denier wrote:
>>>
>>>>
>>>> When loading a package with class definitions containing a trait composition like ClassX uses: TraitY - {#selector}, the composition is lost in the image.
>>>>
>>>> It seems closely related to previous issues 2477 and 2148
>>>> But I am not sure if this is the same bug reappearing or a variant.
>>>>
>>>> To reproduce:
>>>> Gofer new squeaksource: 'MooseAlgos'; package: 'Moose-Algos-Graph'; package: 'Moose-Tests-Algos-Graph'; load
>>>>
>>>> Run tests in Moose-Tests-Algos-Graph: you should have 3 failures and 4 errors, all due to the missing trait composition in some classes.
>>>>
>>>>
>>>> After loading, you can make a diff changes in monticello, it will highlight the package as dirty and show you the lost trait compositions in two classes:
>>>>
>>>> MOGraphNode subclass: #MABfsNode
>>>> � �uses: MATEdgeNode - {#previousEdges. #from:edge:}
>>>> � �instanceVariableNames: 'nextEdges firstDepth'
>>>> � �classVariableNames: ''
>>>> � �poolDictionaries: ''
>>>> � �category: 'Moose-Algos-Graph'
>>>>
>>>> MODisjointSetNode subclass: #MANextNode
>>>> � �uses: MATNodeNode - {#from:. #previousNodes}
>>>> � �instanceVariableNames: 'nextNodes'
>>>> � �classVariableNames: ''
>>>> � �poolDictionaries: ''
>>>> � �category: 'Moose-Algos-Graph'
>>>>
>>>>
>>>>
>>>> The problem has been reproduced in 11383 and 11400 RC2.
>>>>
>>>>
>>>> --
>>>> Simon
>>>>
>>>> _______________________________________________
>>>> Pharo-project mailing list
>>>> Pharo-project@lists.gforge.inria.fr
>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>
>>>
>>> _______________________________________________
>>> Pharo-project mailing list
>>> Pharo-project@lists.gforge.inria.fr
>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>
>> --
>> Simon
>>
>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>> Pharo-project@lists.gforge.inria.fr
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
>
> _______________________________________________
> Pharo-project mailing list
> Pharo-project@lists.gforge.inria.fr
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel �http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.






_______________________________________________
Pharo-project mailing list
Pharo-project@lists.gforge.inria.fr
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project