j
: Next unread message k
: Previous unread message j a
: Jump to all threads
j l
: Jump to MailingList overview
I'm playing around with traits, making a field (the abstract algebra kind) by glomming together two groups. A TGroup defines #*, #identity and #inverse. Obviously one has to resolve name clashes, and a TraitAlias is just the thing:
Trait named: #TField uses: TGroup @ {#identity->#zero. #* -> #+. #inverse->#negated} + TGroup @ {#identity->#one. #inverse->#reciprocal} category: 'Group-Theory'.
Great. Except that I'd expected to see stubs for the aliased methods, like
negated self requirement.
(I'd also expect, if one altered the alias, to see the stub vanish and be replaced by a stub on the new name.)
Because of the lack of stubs, TField organization allMethodSelectors is much emptier than I'd like.
Am I being unreasonable?
frank
squeak-dev@lists.squeakfoundation.org