On Sun, Dec 14, 2008 at 2:18 PM, Stefano Bagnara <[email protected]> wrote:
> Robert Burrell Donkin ha scritto:
>> On Sun, Dec 14, 2008 at 1:55 PM, Norman Maurer <[email protected]> wrote:
>>> Ah ok now I see what makes you wonder... thats right. isLocalEmail
>>> works only after the vut mappings was done. Thats prolly a problem in
>>> some cases..
>>
>> it's the problem in my case :-)
>>
>> (the code's very messy since it's trying to ensure backwards compatibility.)
>
> Sorry but I don't agree on this one.

the user code is very messy - lots of casts everywhere, lots of proxy
layers, no way to implement a clean API, too many user POJOs, too many
deprecated methods. the VUT stuff is annoying since it's tacked on to
this mess.

> The way VUT works is by design. We added some support for easy virtual
> hosting leaving there the backward compatibility, but the underlying
> system is not aware of the VUT so you have to run the VUT in the mailet
> chain before any check for "isLocalEmail" is done.

AFACT VUT is resolved in LocalDelivery but LocalDelivery is - in the
default configuration - only called after an isLocalEmail match

> This applies to every "redirecting" mailet, such as Redirect/Forward and
> not only to the VirtualUserTable stuff.

AIUI these features are not part of LocalDelivery whereas VUT is

> There is not a API to know what the final recipient will be given an
> input email and IMHO there should not be a similar API, because I could
> have some mailet with my own rules accepting some address now and some
> other address later and redirecting them to some address now and some
> other address later.
>
> I and Norman simply extracted the VUT to a service so to be able to use
> the same data both from a mailet and from the in protocol smtp filters,
> but this is only for advanced tuning and there is no out of the box
> solution at all, IMHO.

IMHO if LocalDelivery supports VUT then it should be a core feature
and supported out of the box. if not then create a VUT mailet which
using the service.

> If you have a proposal try to describe it and I'll try to show you the
> issues I see with the proposal ;-)

reduce the complexity: come up with a new clean consolidated user API
without the need for casts etc

- robert

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to