Subject: Re: Mail list envelope sender address
To: D. J. Bernstein <>
From: Christian Kuhtz <>
List: current-users
Date: 11/25/1996 17:27:37
D. J. Bernstein wrote:
> Metzger is stunningly ignorant of how mail works.

Tsk, tsk, tsk.  Glasshouse.

> He continues to claim that parallelism does not reduce latency.

You do not seem to grasp the basic concepts of parallelism.  In an
environment with a constrained pipe, parallelism puts a much higher
demand on the pipe and increases latency much more than a sequential

And because of that, your argumentation is absurd.

> He continues to claim that there is no security problem with throwing
> out messages that have duplicate Message-IDs.

That's what we have Message-ID:'s for.  Read the RFC.

> Sometimes there are several recipients. Sometimes a RCPT response takes
> several minutes. The result is that, on occasion, a message is delayed
> by more than an hour on a single hop. That's unacceptable.

The fact that an RCPT takes several minutes means you have a constrained
pipe; in this particular case probably more likely at the end points of
the pipe itself.  You will not reduce latency by blasting more data
through a constrained pipe.  In fact, you will increase latency because
the capacity of the pipe is already exhausted, otherwise you wouldn't
have latency in the first place.

Aside from that, you are free to tune your mail configuration to honor
the latency limits you would like to impose.

> > no, Dan, I don't believe you no matter how many graphs you come out with
> Some people don't profile code because they aren't worried about speed.

All your profiling doesn't do you any good if you completely neglect
logic and your consideration of resource constraints.  It doesn't matter
how long the code spends at a particular place as much as the IO it is
generating.  And if your IO is constrained, your code can be as
efficient as it wants to.  You don't seem to grasp that very basic idea.

Internetworks in general are not unlimited, latency-free or low-latency
systems.  They usually are filled up to the rim, since this is a
supply/demand driven system which tends to find its equilibrium at the
rim.  This is unfortunately not an ideal world.

> Metzger doesn't profile because he doesn't believe the profiler.

I love to profile.  However, this has zip to do with profiling.

> Metzger and Fair are allowed to criticize qmail, and I can't respond?

If you were to stop bashing and actually presenting an argumentation, I
think people would be a little more relaxed about it.  I for one would

Best regards,

Christian Kuhtz <>
Network/UNIX Specialist
Paranet, Inc.