(SystemNavigation default allMethodsSelect: [:meth |
��� | hasABadLiteral |
��� hasABadLiteral := false.
��� meth literalsDo: [:lit |
��� ��� (lit class = ScaledDecimal and: [lit numerator isNil or: [lit denominator isNil]])
��� ��� ��� ifTrue: [hasABadLiteral := true]].
��� hasABadLiteral])
��� do: [:m |
��� ��� m actualClass recompile: m methodSymbol from:� m actualClass]


2009/6/17 St�phane Ducasse <stephane.ducasse@inria.fr>
Ok nicolas
Sorry for the problem.

I guess that I will wait for the code snippet

stef writing should be writing a boring report for a month ago :)

On Jun 17, 2009, at 3:09 PM, Nicolas Cellier wrote:

> in update10307 (script385) related to
> http://code.google.com/p/pharo/issues/detail?id=816 :
>
> ScaledDecimal2-Part1-nice.1.cs had a postscript to mutate old
> instances to new instances and not losse existing ScaledDecimals (at
> least the literals in ScaledDecimalTests).
>
> This has not been applied, so some ScaledDecimal now have a nil
> numerator and denominator.
> This makes some tests fail.
>
> Cure: detect and recompile these methods having a broken
> ScaledDecimal literal.
> I don't have time to give a code snippet by now...
>
> Nicolas
>
> _______________________________________________
> 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