[Top] [All Lists]

Re: [ietf-smtp] IETF Policy on dogfood consumption or avoidance - SMTP version

2019-12-25 10:43:38
On 12/25/19 11:22 AM, Dave Crocker wrote:

If we want to see mechanisms that focus on the spam originators
and spammers, rather than the receiving end, focusing on people,
and especially mechanisms, that have characterized efforts for
most of the last 25 years of effort may not be the best answer.

If you want to formulate proposals that have any practical basis, it requires knowing quite a bit about the pragmatics of spam today.  There is nothing new about the idea of seeking 'strategic' change; there have been many over the last 25 years.  However such proposals typically are not burdened by requisite knowledge and typically embody exactly the naivete one would expect. Similarly, proposals need to attend to the realities of human factors, which is similarly lacking in typical proposals.

If existing spam filtering practice were working well today, we wouldn't be having this discussion.   More broadly, there's always a tension between doing things the old way, and seeking better ways to approach a problem.   There are always advantages to doing things the old way, especially for companies that have built products around doing that.   There will always be some naiveté with new proposals.   And yet, without new ideas, there's a huge tendency for old approaches to eventually do more harm than good.    This isn't at all specific to email or networking or engineering, it also exists in medicine and in law.

In a sense, of course, you're right.   But I think it's a bit too early for damage control.  If we ever get around to forming a working group we can expect that it will attract interest from everybody who has an existing anti-spam product, all looking for ways to defeat anything that IETF might do that would hurt their market share, no matter how good or harmful their products are.


ietf-smtp mailing list
<Prev in Thread] Current Thread [Next in Thread>