[squeak-dev] Agreement among packages better than name-spaces from( Re: worst crash yet ... thread)

Jerome Peace peace_the_dreamer at yahoo.com
Wed Dec 17 03:35:40 UTC 2008

Agreement among packages better than name-spaces from( Re: worst crash yet ... thread)

[squeak-dev] Re: worst crash yet trying to load stuff from Packages Universe in 3.10.2....
Andreas Raab andreas.raab at gmx.de
>Tue Dec 16 05:00:30 UTC 2008
>Greg A. Woods; Planix, Inc. wrote:
>>> Puck would like to know exactly how to recreate this failure.
>>> (So he can try it out on some of his "friends" :).
>> Looks like all I had to do was try to load SmallDEVS from the default 
>> package universe.
>No surprise. Polymorph as a class called LabelMorph. SmallDEVS has a 
>class called LabelMorph. They don't agree. So much for not having name 

Um. I don't see this as a name spaces problem so much as an agreement problem.
If both packages live in the same universe then they have a need and responsiblity to agree on terms.

Namespaces would avoid this but create the dual of this problem. You then would have the clutter of needing two definitions in different namespaces to mean the exact same thing.

My take is that disagreements are inevitable. The repair is conversation, negotiation and agreement. That way you have one language not two or many.

The other opportunity here is this defines another Simple test for universes.

Invariant: No two packages living in the same universe shall define the same class.
One will have to differ to the other. Or they shall ever be galaxies apart.
The same is probably true for method overides.

The indicated repair is to repackage the conflicting code into its own package and have both other packages be dependent on the common package.

I say repackage because packages should have clearly defined responsibilities just like objects. The hierarchy of packages should make sense in that way also.

These things sort themselves out well if you look at objects (here packages) responsibilities and collaborations.

Gary is reinventing things that with more time and understanding would be created with reused classes and objects.

SmallDEVS is probably doing the same.

Its not like the code doesn't need another iteration of thought.
There are magnitudes of differences in speed between 3.10.2 Morphic UI and the Squeak-dev UI. Same with the MVC UI and Morphic UI in 3.10.2.

Yours in curiosity and service, --Jerome Peace


More information about the Squeak-dev mailing list