[comp.mail.sendmail] IDA

terry@hq.af.mil (Terry Bernstein) (04/10/90)

What is IDA????

thanks,

Terry

Lovstrand@Xerox.COM (Lennart Lovstrand) (04/11/90)

> What is IDA????

The IDA Sendmail Enhancement Kit is a set of sendmail source code
patches and configuration files originally developed by your's truly
while at the Department of Computer and Information Science*,
Linkoping University, Sweden in 1987.  Since then many people have
contributed with additional features and improvements.

Included in the package is: direct pathalias support via dbm files, yp
support, support for bidirectional alias files, multi-token class matches,
local recipient dwimification, various bug fixes + a lot more.

The IDA Kit (verison 1.2.8 for sendmail 5.59 (still, sorry)) is
available for anonymous ftp from arisia.xerox.com in pub1/ida.tar.Z.
Versions for sendmail 5.61 also exists on various locations on the net.

For further information, see the Kit's README file.

--Lennart Lovstrand
Rank Xerox EuroPARC (but interested in other job opportunities :-)

*) "Institutionen for Datavetenskap" in Swedish, hence "IDA".

jbryans@beach.csulb.edu (Jack Bryans) (03/18/91)

I got a lot'a local static about giving sendmail-IDA a try, all the way up to
"It'll never work", esp. wrt decnet (more on that later).  Enough to give me
the impression that IDA may have had a bad rep some time back.  Whatever it may
have been, it's just fine now.

There was a lot we wanted out of our mail setup.  Recognition of "hosts"
sharing a mail spool & common yp passwd should allow mail to be treated as
local, including delivery.  To the outside world, we'd like to have all of our
"From"s appear the same regardless of the worstation being used at the time.
We'd like to be able to route mail addressed to juser@our.domain to the correct
sub-domain & host, even VMS & "blue" ones.  A few other things like decnet
support'd be nice too.

I was hoping that IDA'd help us get a little closer to the above.  It didn't.
It gave us the whole ball of wax, & then some.  In addition, it looks like
we'll seldom if ever need to dig into the .cf rule-sets.

It looks like someone dug into the src to make it port to a lot of platforms.
The hacking into the .c's I had to do w/vanilla Berkeley 5.64 was unnecessary;
it compiled in Ultrix 4.1 w/o a whimper.

A few notes for the next IDA 1st timer:  The examples in the DBM-Guide are in
dbm "parse" or "make" form.  The keys _should_ all be entirely in lower case.
To get decnet support, uncomment the define MAIL11V3 in conf.h, include defines
for DECNETNAME & DECNETNODES in your .m4, AND make sure all the decnetnodes
have been defined to ultrix ncp.  Piece'a cake.  I've seen recent posts asking
how to get decnet support -- hope this helps.

I haven't been able to get the DECNETXTABLE option to do anything for me.  It's
sparsely commented, but suggests that it'll use decnet routing for hosts that
are both decnet & TCP.  Perhaps someone who's had luck w/it'll post a follow
up.

Jack

rickert@mp.cs.niu.edu (Neil Rickert) (03/18/91)

In article <JBRYANS.91Mar17182038@beach.csulb.edu> jbryans@beach.csulb.edu (Jack Bryans) writes:
>I got a lot'a local static about giving sendmail-IDA a try, all the way up to
>"It'll never work", esp. wrt decnet (more on that later).  Enough to give me
>the impression that IDA may have had a bad rep some time back.  Whatever it may
>have been, it's just fine now.

 Thanks for the support.

>It looks like someone dug into the src to make it port to a lot of platforms.
>The hacking into the .c's I had to do w/vanilla Berkeley 5.64 was unnecessary;
>it compiled in Ultrix 4.1 w/o a whimper.

 Paul Pomes at uiuc did a lot of work on that.  He will appreciate your
comments too

>I haven't been able to get the DECNETXTABLE option to do anything for me.  It's
>sparsely commented, but suggests that it'll use decnet routing for hosts that
>are both decnet & TCP.  Perhaps someone who's had luck w/it'll post a follow
>up.

  I am currently experimenting with this.  If you have a recent version of
'rcs', and fetch the rcs archive of my rulesets from mp.cs.niu.edu:pub/cf
you will find my experiment hidden in a branch somewhere.

-- 
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
  Neil W. Rickert, Computer Science               <rickert@cs.niu.edu>
  Northern Illinois Univ.
  DeKalb, IL 60115                                   +1-815-753-6940