K. K. Subramaniam wrote:
> On Friday 12 February 2010 08:53:36 pm Juan Vuletich wrote:
>   
>> - Support for the Unary numeral system, as suggested by Dan Ingalls at
>> http://lists.squeakfoundation.org/pipermail/squeak-dev/2000-March/013368.ht
>> ml
>>     
> 1r1111 = 4.
> Yay! I suppose this make Cuis the first Smalltalk to support tally counts. I 
> noticed you also fixed 1.0e+2 to return 100.0.
>   

I didn't :) . I just never loaded the updates that produce that behavior.

>> - Closure measurements (based on work by Eliot Miranda) are shown in the
>> annotation pane for any method
>>     
> Just to keep noise levels down, you may want to drop "No real (non-optimized) 
> closures". When there is nothing to report, report nothing.
>   

Mhhh. Sounds reasonable. Will try it.

>> - Removal of 43 isXXX methods, replaced by the general #is: method
>>     
> Could this be generalized in Object? e.g.
> is: aSymbol
>   ^self class = (Smalltalk at: aSymbol) or:  [ super is: aSymbol ]
>
> isA: aClass
>   ^self class = aClass or: [ super isA: aClass ].
>   

This, as you wrote it, won't work. Try it! Anyway, what you suggest is 
similar to #isKindOf:. The difference is that #isXXX and #is: are not 
tied to the hierarchy at all. I see them more as a test for a certain 
protocol. That's why #isKindOf: is so bad. It is tied to the hierarchy. 
See 
http://lists.squeakfoundation.org/pipermail/squeak-dev/2009-December/141694.html 
.

> Use of ALT+arrow to cycle through windows is a nice addition for keyboarders 
> like me. I would have preferred to use ALT+SPACE (or one of special character 
> keys) to do the same because arrow keys are placed differently on different 
> keyboards.
>   

I see. Will take a look. Thanks for the suggestion.

> A nicely thought out update. Thanks .. Subbu

Thanks.
Cheers,
Juan Vuletich