On 23/04/16 19:27, Evgeniy Berdnikov wrote:
> Each authenticator must have unique
> "public name", denoting its mechanism (CRAM-MD5, PLAIN, etc).
> That's why I have to pack code lines for different smarthosts
> into one authenticator. It looks messy. Better solution, IMHO,
> would be to write two authenticators in chain, each with
>
> client_condition = ${if eqi{$host}{...}}
>
> Is there some reason to prevent chaining of authenticators
> with the same public name, or is it just a design flaw?
It is deliberately coded, so it's a design decision.
--
Cheers,
Jeremy
