[sc34wg3] New TMCL slides: at least 2 roles must be played

[Lars Marius Garshol]

* Steve Newcomb
> 
> Wow. I'm so surprised that I think I must misunderstand you, or maybe I've lost some context here. The two associations:
> 
> association-type(role-a: foo, role-b: bar)
> 
> and
> 
> association-type(role-a: foo, role-c: bar)
> 
> would be regarded as the same, despite the difference in their role types? Do you mean that they would merge?

They wouldn't merge, because the engine really does respect the TMDM, and if you used the API you'd see the difference clearly. But in Vizigator and the related-topics portlet the relationships would be treated as being the same. Omnigator would also show them as being the same (from the "foo" side, anyway).

It's been like this for almost a decade. Nobody's ever complained.

> I'm lost. What use case? Are you talking about associations in which there's only one role player?

No, I mean supporting association types with multiple signatures in TMCL.

--Lars M.
http://www.garshol.priv.no/tmphoto/
http://www.garshol.priv.no/blog/