hello everyone,
i am sure many others of you have also noticed that there is
an smtpd in the works, but unlike the other projects in
progress (like opencvs) i am vainly waiting for an
announcement about it... i realize it is not ready for use
but neither was opencvs when started so why the secrecy?
it is no doubt a controversial issue, and probably everyone
sees this coming, but i am going to ask anyway: why?
i am not sure if this is only a lightweight smtpd (without
esmtp), it's banner doesn't say so at the moment (can that be
changed please? even spamd does that :]) so as someone
involved in administering mail servers, i am asking, why make
another smtp server? especially when:
0. is is HARD
1. it is being written from scratch by an OS team
2. sendmail being ferociously defended by team members
everytime someone brings up the "lets change smtpd in
base"
3. there is a wonderful alternative with a security
record matched only by openbsd by a religiosly
respected security researcher (so it is really
not the case of "all the alternatives are bloated
pieces of shit" like xntpd/snmpd/quagga etc)
4. there is a paper by the aforementioned researcher
describing how hard is to make a full featured
smtp deamon, and how long it takes for it to
be mature (sorry, cant find the link atm)
5. development forces are stretched thin all the time
what is this smtpd going to add? chroot? privilege
seperation? easy configuration? great support on the mailing
list? great documentation? full support of the standard?
interoperation with other packages like ldap and sql?
postfix is one of the best example for all these.
or is this just about the license?
(i always thought the bsd crowds were far less zealous in this
respect and applied common sense as in use the right tool for
the right job. i am not such a fanatic that adding a non-bsd
licensed package to my server will send me right to bsd hell.)
or will this be another project in the makings for a decade?
or is this simply going for world domination by the devs
with all the french sounding names? (gilles@ + pyr@) :]
contrary to popular belief, this is not an attack on smtpd.
having said all this, actually i am very excited about this
project and can't wait to see it taking off. but at the
same time i would like to hear why it seems like a good
idea for the developers and why they are willing to give
it hundreds of hours of their time... esp with the bar
being as high as postfix...
-f
--
if you have to travel on a titanic, why not go first class?
