On Apr 14, 2004, at 3:43 PM, Bert Freudenberg wrote:

> Am 15.04.2004 um 01:23 schrieb goran.krampe at bluefish.se:
>
>> 	(Smalltalk at: #Fruits::Apple) == ((Smalltalk at: #Fruits) at: 
>> #Apple)
>>
>> The left hand shows that SystemDictionary has some "compatibility
>> behaviour" making it seem as if the qualified names are all accessible
>> "flat" in Smalltalk, which they aren't. It will do a two step lookup
>> like the right hand shows.
>>
>> 	((Smalltalk at: #Fruits) at: #Apple) == Fruits::Apple
>>
>> Here the right hand shows that Scanner/Parser has been augmented to
>> properly handle qualified names (with a little help from Andreas).
>
> Does that mean I can not enumerate all classes in the system by 
> looking at the entries in Smalltalk? What about the invariant
>
> 	(Smalltalk at: aClass name) == aClass
>
> If this does not hold, many tools would have to be made 
> namespace-aware. This can be elegantly avoided when using the 
> fully-qualified name as actual class name. I thought your original 
> proposal did this?

Yeah, I'm with Bert.  Much better to have the "compatibility behavior" 
go the other way - that is, rather than "virtualizing" the 
SystemDictionary so that

Smalltalk at: #Fruits::Apple

is implemented with a two level lookup:

(Smalltalk at: #Fruits) at: #Apple,

you should instead virtualize the namespaces so that

(Smalltalk at: #Fruits) at: #Apple

does the fully qualified lookup:

Smalltalk at: #Fruits::Apple.

This requires *way* less modification of existing tools and code.

I, too, thought this was the idea all along.

Avi