ietf-asrg
[Top] [All Lists]

Re: [Asrg] 6. Proposals - Pull System (revisited)

2003-11-26 17:18:24
At 5:33 PM -0500 11/26/03, Yakov Shafranovich wrote:
Bill Cole wrote:

At 3:43 PM -0500 11/26/03, Yakov Shafranovich wrote:
Bill Cole wrote:
At 5:04 PM +0000 11/26/03, Bruce Stephens wrote:
Bill Cole <asrg(_at_)billmail(_dot_)scconsult(_dot_)com> writes:
....
If complete consent control can be created in email, with the sender in charge at all times, then that would solve spam, since the sender must actively subscribe to something to receive it. That is one of the motiviations behind pull systems, where a hybrid approach of push and pull is used.


I think that pull systems are less 'email' the more purely they follow 'pull' models. The example of the blogging universe predominantly using RSS as the means of content providers getting material to readers rather than the traditional mailing list with all of its clunkiness is a good sign.

Email may be the 'killer app' of the Internet that can be made to do everything, but it is not designed to do everything well, and it may simply be time to start narrowing what email is allowed to do in order to push the traffic into more suitable media: many-way discussions to NNTP, serial broadcast to RSS.

Can we connect the two: can mailing lists be distributed via RSS, with an ability to receive them in an MUA, and ability to respond to them via email.


'Can' is a fuzzy word. I am sure that such things are possible, but I am not familiar with any particular implementations. After all, the 90's brought us single clients that handle SMTP, POP3, IMAP, and NNTP so seamlessly that many people have no concept of news and mail being different things or of the fact that there are different protocols involved in sending and receiving mail. The place to look for how people are using RSS and creating end-user tools for it is in the blogging world. I am not familiar with all the available tools, but since Mr. Judge says that MUA-integrated aggregators exist and it seems a natural integration, I assume that they do.

This is really getting back to what some people (including me) have been saying for a long time: the solution to spam cannot come from more engineering, because the problem is not a technical one. Restricting broadcast email (by which I mean messages sent via SMTP) is not all that difficult, and there are working examples of systems that do an extremely effective job of it. Unfortunately, broadcast email is not currently all spam, so whacking broadcast email altogether would be a bad thing. If the *social* change can be effected of getting non-spammers who now use SMTP for broadcasting to use other available mechanisms like RSS, the technical aspects of eliminating the spam (i.e. all broadcast email that can't migrate to RSS or some other pure-pull mechanism) are not quite trivial, but are certainly straightforward.

Bringing in another subthread to this: the advantage of ordaining RSS as the preferred protocol/medium for one-way or heavily asymmetric broadcasting (i.e. most marketing and news materials) as well as for guided/led/facilitated discussions (i.e. like most weblogs that take reader comments) and NNTP as the preferred mode for multi-way ownerless discussion group is that unlike the approach of building an all-new mail protocol or hacking more extensions onto SMTP, there isn't a new protocol needing a flag day, and early adopters can get a benefit. As proof, note that we have people making the move away from SMTP mail NOW because of the fallout from the spam wars. Spamcop runs a news server, not a listserver. Blog sites offer users updates via RSS, not email notification.


--
Bill Cole bill(_at_)scconsult(_dot_)com


_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg