Ah, the Internet's promise of global interconnectivity in all its glory.
This kind of SMTP breakage is widespread (and not limited to SMTP) due to
the (ab)use of NAT as a means of firewalling and/or quarantine. The only
thing you can (generally) count on is that port 443 will be open and
unproxied.
The more we talk about SMTP submission, the more I like web-based MUAs.
SMTP should probably just be replaced with an XML-based web service. Absent
that, all we can do is try to impose some accountability for the identities
used in an SMTP submission. Among the mechanisms for doing just that are
SMTP authentication, SPF, DomainKeys, and their ilk.
After that, it's a matter of content filtering (if one considers the
accountable sending party a contextual factor). In most cases, I think
content filtering will have to become more personalized to the individual
recipient. That is to say that one could greatly improve the end user
experience by infusing the communal notions of desirability (e.g. RBLs,
Brightmail) with unique personal knowledge (e.g. to whom have I already sent
email?).
Tripp
----- Original Message -----
From: "M Z R" <mzrahman(_at_)gmail(_dot_)com>
To: "Devdas Bhagat" <devdas(_at_)dvb(_dot_)homelinux(_dot_)org>;
<asrg(_at_)ietf(_dot_)org>
Sent: Sunday, December 05, 2004 1:21 PM
Subject: Re: [Asrg] SMTP AUTH
Actually, there are some cases where submission port (587) also cannot
be used. For example, my personal experience in McDonalds Free WiFi in
Singapore. I tried using my SMTP AUTH server, but they 'hijack' all
SMTP related traffic and send it through their own gateway.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg