Danny Angus wrote:
> Noel J. Bergman <[EMAIL PROTECTED]> wrote:
> > > +0 This is a temporary measure which resolves the issue, but doesn't
> > > remove the root cause properly.
> > Danny, what are you talking about? Of course it removes the root cause.
> > That is all and exactly what it does. What other root cause are you
> > presuming?
> Yes it removes the root cause, but not in the best way, from where I'm
> sitting it is a pragmatic patch and not one which has actually taken
> account of all of the design considerations.
Huh? The solution is to always avoid using InetAddress for resolution.
That is the fix we use throughout JAMES in both v2.3 (and earlier) and
trunk. In v2.3.0, we missed four lines of code; one line in each handler.
In trunk those lines were replaced with a call to our DNSService to resolve
the names using dnsjava.
So what's the difference between what I did and what is in trunk? In my
patch, I used a static method on the DNS server class. In trunk we changed
the interface, changed each client to lookup the DNS server from Avalon,
changed the assembly descriptor to register each one as needing the DNS
service, and then called an instance method instead of a static method.
That's it. Assembly vs a static method. The only difference between the
two code sets. That's all that we're talking about.
> > I'm not ready to define [DNS service] in the Mailet API.
> No me either. I'd much rather educate than compel.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
