IIRC Germán told me that the extensions of the package in the repository should be refactors to make in Pharo, and there are things there related to enviroments and system organization. On Thu, Aug 19, 2010 at 5:49 PM, James Foster <Smalltalk@jgfoster.net>wrote:

Stef,

As part of his GSoC 2010 project, Germán Leiva has been doing some refactoring of this and has some sample code at http://www.squeaksource.com/Environments.html. While it hasn't come as far as I hoped, it does give an interesting insight into some of the relationships between environment, organization, and related things (the Monticello implications were somewhat nontrivial). Although the GSoC is officially over, I expect to be working on this going forward and would like to have a significant discussion with interested parties at ESUG next month. How quickly do you want to solve this? ;-)

James

On Aug 19, 2010, at 1:29 PM, stephane ducasse wrote:

...

so far we have

SystemDictionary>>organization "Return the organizer for the receiver" ^SystemOrganization

so it means that this is global in Smalltalk globals. But I'm skeptical that this can work for another instance of SystemDictionary and even if this works I find this code brittle since it relies on the compiler global look up.

I was thinking that

either defining an organization as an instance variable SystemDictionary would make sense or using an explicit lookup in the instance itself.

SystemDictionary>>organization "Return the organizer for the receiver" ^ self at: #SystemOrganization

now I'm dead so I need other eyes.

Stef

_______________________________________________ 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