On Jul 22, 2009, at 23:59 , Nicolas Cellier wrote:

> This is what I get during the update:
>
> ThirtyTwoBitRegister doesNotUnderstand: #load:
>
> This is triggered by:
>
> MD5NonPrimitive>>initialize
>       "MD5 initialize"
>
>       "Obscure fact: those magic hex numbers that are hard to type in  
> correctly are
>       actually the result of a simple trigonometric function and are  
> therefore
>       easier to compute than proofread.  Laziness is sometimes a virtue."
>
>       | c |
>       c _ 2 raisedTo: 32.
>       SinTable _ Array new: 64.
>       1 to: 64 do: [ :i |
>               SinTable
>                       at: i
>                       put: (ThirtyTwoBitRegister new load: (c * i sin abs) 
> truncated)].
> SNIP...
>
> Is this related to load order or something?

Probably. It's strange that I've not noticed this because I tested the  
loading several times. MD5NonPrimitive is a new class. Note, you mean  
the *class side* initialize. If this initialize is run before the new  
version of the ThirtyTwoBitRegister class is loaded, it will break.  
But looking at MCPackageLoader>>basicLoad, the initialize methods  
should be run at the very end. Not sure how to fix this.

>
> Also mind the previously banned underscores...

Yep. I missed them wile integrating. Will fix it.

Adrian

>
> Nicolas
>
> _______________________________________________
> Pharo-project mailing list
> [email protected]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


_______________________________________________
Pharo-project mailing list
[email protected]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Reply via email to