Sendmail vs postfix

Christopher Browne cbbrowne-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org
Fri Nov 11 05:59:53 UTC 2005


On 11/10/05, Ian Zimmerman <nobrowser-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org> wrote:
>
> Lennart> I refuse to go near qmail due to it making to much work for me
> Lennart> to install and maintain it on debian, and I avoid djb's
> Lennart> software on principle because I think he is too weird.  Perhaps
> Lennart> I should consider avoiding some of RMS's software too for the
> Lennart> same reasons (I already avoid emacs, although gcc would be hard
> Lennart> to do without).
>
> That's a funny reason not to use someone's work.  I think you probably
> want to avoid anything I do, as well :-)
>
> No question that qmail (the software) is weird, though.

The "djb property" stirs up a lot of heat.

I see it four, no five ways:

1.  The "djb thing" of promoting the notion of "license free software"
is VERY distinct from the common thing of people thinking it important
to pay attention to the licenses of the software that they use.

While Debian has some "Trotskyite leanings" where the "more carefully
licensed than thou" approach makes it painful to directly participate,
the fact the attention is paid means that Debian doesn't tend to get
blindsided by license issues.

In contrast, while there are BSD folk that spit on the thought of
"toadying to RMS" by licensing things under the GPL, nonetheless the
"BSD way" tends to involve a similar sort of care surrounding licenses
(albeit with a different preferred position).

The djb approach does more than merely spit on them all; it voids
other buffers at that sort of "due diligence."

2.  DJB, being a tenured professor, feels quite prepared to hold
tenaciously to whatever positions he decides to hold, and the flame
wars have been pretty actinic...

3.  The "qmail way" involved DJB effectively doing quasi-mathematical
proofs that parts of what he did are 'provably correct,' which means
that when he feels like holding tenaciously to some position, he has a
pretty strong leg to stand on.

4.  There's a "cult of personality" surrounding him.  Other people,
who are neither tenured professors, nor particularly knowledgeable
about computer security, wish they were him, and imagine (incorrectly)
that they could become like him if they flame brightly enough.

(They are wrong because the only way you can get away with such
flaming is if you are such an intellect that you can earn a Ph.D in
mathematics, and if you are simultaneously political enough to earn
tenure.  Perhaps DJB got tenure out of force of intellect, without
much political investment; if that be the case, that adds to the
"intellectual requirement" that the cultists tend to lack...)

5.  Falling out of the fact that his tools were, indeed, designed by a
mathematician, who puts a premium on people understanding things, the
configuration scheme for qmail puts a huge premium on understanding
the architecture of how qmail works.

Personally, I figure that one of the major points to having other
people write code is so that I *don't* need to understand how it
works.

That has always put me in conflict with how qmail works...  I have
configured it several times on my systems over the years; I have
always found it excruciatingly painful to configure.  Sendmail has
always been easier, at least as far as fiddling with the parts I
*NEEDED* to fiddle with...

All of these things add together to imply that deploying DJB software
can get rather controversial...
--
http://www3.sympatico.ca/cbbrowne/linux.html
"The true  measure of a  man is how he treats  someone who can  do him
absolutely no good." -- Samuel Johnson, lexicographer (1709-1784)
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml





More information about the Legacy mailing list