[feedback-report] On redacting data from ARF messages

[J.D. Falk]

On Dec 28, 2009, at 6:41 AM, Steve Atkins wrote:

> My view is that ARF is a container format. As long as the message/rfc822 part makes some pretense to be an email, leaving the details of that as part of the agreement between the creator and the consumer of the feedback loop fits into that, and conveniently leaves the question of whether the creator has falsified any of the data out of scope of the ARF format itself.

I like this angle; it points the blame squarely to the redactor (rather than the RFC authors) and may move the conversation from "redaction should be outlawed!" to "if you don't like that an ARF stream is redacted, you can stop receiving it."

> Also, discussing it will add significant complexity to the document without adding value nor affecting the behaviour of any implementors one jot. They're going to do what they do regardless.

But if we don't discuss it in the draft initially, we'll have to discuss it within the WG every freakin' time someone new gets annoyed by redacted ARF messages.

--
J.D. Falk <jdfalk at returnpath.net>
Return Path Inc