Ignatios Souvatzis wrote...

> On Tue, Aug 15, 2000 at 12:34:44PM -0700, Pete Naylor wrote:
> 
> > In the case of postfix, you also wind up with a bunch
> > of binaries, which is neither good nor bad in respect to the theoretical
> > consideration above. In practice though, it's ugly and more prone to have
> > one component fail and cause a mess that's difficult for someone unfamilar
> > with the software to diagnose.
> 
> postfix check?

And this should instantly be apparent and intuitive to those people who
are happy and familiar with the existing tool?

> Anyway: after 10 minutes of reading online docs

For many people, that 10 minutes would be wasted because they have no
reason to switch from using the included sendmail.  Furthermore, if they
did have a reason to switch, we'd be doing the world a favour by
encouraging them to consider a number of alternatives instead of pushing
another defacto standard upon them which will no doubt become a liability
in years to come, when some other piece of software is the latest and
greatest thing in the opinion of NetBSD developers.

> and usign vi, I had configured
> postfix to do what I had failed to do in the 5 years before with sendmail:
> make the damned think NOT cause the resolver to trigger the dialup IP
> connection when writing mail. And yes, I've listend to lots of people that
> know and used all their recipes.

This is a success story that I'm glad to hear.  It's an example of why
people should be able to depend on consistency of the tools distributed in
NetBSD while also having the opportunity to easily evaluate alternatives.
There's no reason you couldn't have tried postfix long ago, or other
alternatives which would also easily have met your needs.  By pushing
postfix on all NetBSD users, we encourage them to suffer with a certain
set of functionality that might not suit them, and we cause hassles for
the many folks already happy with sendmail in at least some of their
installations.  This is not an improvement to NetBSD at all, it's just
more of the same problem.  Alternative MTAs are easily accessible, and the
only worthwhile improvements should be in the aspect of making them easy
to plug in and evaluate/use with NetBSD.

-- 
Pete Naylor