dcc_mail_host issues

Vernon Schryver vjs@calcite.rhyolite.com
Tue Jul 1 22:51:16 UTC 2003

> From: Spike Ilacqua <spike@indra.com>

> ...
> I've found couple of issues with the dcc_mail_host code in dcc.m4.
> First, the commas are missing where dcc_mail_host has been added to
> dcc.m4:
> define(`confMILTER_MACROS_CONNECT',confMILTER_MACROS_CONNECT``,{dcc_isspam}, {dcc_notspam} {dcc_mail_host}'')dnl
>           ^----

That's a good point.
Strangely enough, with at least 8.12.10.Beta0, sendmail doesn't seem to care.

I've a change for dealing with persistent bad servers that justifies
a 1.1.40.  I'll include the commas.

> Second Local_check_mail is called with a trailing "." on the address, so
> to get the same effect as mail_hosts it needs to be:
> R$* $| $* .		$: $1 $(macro {dcc_mail_host} $@ $2 $)
> Or maybe the more paranoid (if I remember "$@" correctly):
> R$* $| $* .		$@ $1 $(macro {dcc_mail_host} $@ $2 $)
> R$* $| $*		$: $1 $(macro {dcc_mail_host} $@ $2 $)

Are you sure of that?  Mail_host currently almost always has a
trailling '.' when it gets to dccm.  The only exception I've noticed
is when the domain failed to resolve.  The appropriate checksum
routine in the DCC client library has long ignored the trailing '.'.

> Finally, and tragically, I was wrong, it does not appear work if
> delay_checks is enabled.  I'm not sure how to work around that...

What is the bad effect of delay_checks?   Is ${dcc_mail_host} null?
If so, I'll have to add a kludge to distinguish that case from 
a sendmail.cf that does not know about ${dcc_mail_host}.

I don't recall now why you need delay_checks, or if you said.

Vernon Schryver    vjs@rhyolite.com

More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.