[Zope3-dev] Ranking adapters (2nd try)

[Phillip J. Eby]

At 07:52 PM 9/30/03 +0300, Steve Alexander wrote:
>Here's my rationale:
>
>* A112 disappears because it is a registration for exactly the same 
>from->to as A212.
>
>* I put A111 before A110 because B1 is more specific than B0. A B1 can do 
>all of what a B0 can do. Liskov substitutably etc.

Note that historically, Zope interface adapter registries have considered 
an adapter for a more general interface to take precedence over an adapter 
for a more specific interface.  I originally questioned the wisdom of this, 
but then later realized that it was a good idea, for reasons that currently 
escape me.  :(  All I remember is that I found it to be a good idea in 
practice, even though in theory I agreed with your assessment above.  ;)  I 
then wrote PyProtocols to follow the same rule, on the basis of 
"implication distance".

That is, although a B1 is substitutable for a B0, it is *not* a B0.  If you 
ask for a B0, and there is a B0 adapter declared, then that should take 
preference over an adapter to B2.