19 Dec
2007
19 Dec
'07
8:21 p.m.
Hi adrian
I like your smart sets better than my uniqueOrderedCollection.
The problem with UniqueOrderedCollection is that it is not clear what
to do with add:before: and the like.
Since we cannot garantee the order anyway.
Stef
19 Dec
19 Dec
8:52 p.m.
In fact the protocol of OrderedCollection is not conceptually the
same as
GrowingArray. This is fun to see that for simple usage we do not
really care of the fat interface
of OrderedCollection and would benefit of a simpler one. So may be
traits are the way to go.
may be we should use SmartSet in Moose instead of OrderedCollection?
Is there any plans for that. I would be in favor.
Stef
On 19 déc. 07, at 20:21, stephane ducasse wrote:
Hi adrian
I like your smart sets better than my uniqueOrderedCollection.
The problem with UniqueOrderedCollection is that it is not clear what
to do with add:before: and the like.
Since we cannot garantee the order anyway.
Stef
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
9:17 p.m.
New subject: Thoughts after starting to implement UniqueOrderedCollection
may be we should use SmartSet in Moose instead of
OrderedCollection?
Is there any plans for that. I would be in favor.
You mean for all multivalued things (which are not keyed)? I guess this
would not work for all arbitrary metamodels; which might depend on the
location of elements; and the amount of occurrences?
Btw, I put an implementation of my opposites in the VW implementation of
Fame. The implementation can be found under Fame-Slots, and is used in
the definition of the FM3 classes. I'm working on porting all my
Python-Fame stuff to VW now.
cheers
20 Dec
20 Dec
8:49 a.m.
New subject: Thoughts after starting to implement UniqueOrderedCollection
may be we
should use SmartSet in Moose instead of OrderedCollection?
Is there any plans for that. I would be in favor.
You mean for all multivalued things (which are not keyed)? I guess
this
would not work for all arbitrary metamodels; which might depend on the
location of elements; and the amount of occurrences?
Btw, I put an implementation of my opposites in the VW
implementation of
Fame. The implementation can be found under Fame-Slots, and is used in
the definition of the FM3 classes. I'm working on porting all my
Python-Fame stuff to VW now.
cheers
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
5969
days inactive
5970
days old
3 comments
3 participants
participants (3)
-
stephane ducasse -
Stéphane Ducasse -
Toon Verwaest