On 9/20/2024 11:57 AM, Michael Wise wrote:
Although ... if there is a standard header field that doesn't start
with an X-, would love to know what it would be.
I personally prefer the Comments: header, as documented in RFC-822 and
subsequent, but nobody else seems to use it.
As to the system design …
That’s not something I am in a position to speak to publicly.
1. Note that use of X- as a construct was deprecated a long time ago.
It was -- I thought and think -- an entirely reasonable idea.(*) But it
turned out to create more problems than it solved. If it actually
solved any.
2. My reference to 'standard' was not meant to imply that an appropriate
one existed but that use of non-standard stuff is, at scale, an issue.
(No matter how major the source is...) So, for example, if signaling
through an extra header field is deemed appropriate, fully documenting
it through an independent, public publisher would be worth considering.
Especially if that path might produce something that could be called an
independent standard.
3. Please do NOT use a comments field as an ad hoc standard. It makes
the field useless for actual comments. And I believe this problem has
already been well demonstrated. (I fully appreciate the appeal please
avoid that temptation.)
d/
(*) The construct was not in RFC 733, but was added in RFC 822. There
was no IETF back then, but we did have a public -- ie, within the
Arpanet community, since the Internet wasn't an official, production
environment quite yet -- discussion process. I wish I could remember who
suggested the construct, but as I recall, I really liked it. (Still do,
as an example of creative thining.) I don't remember how much discussion
there was about it, and I don't recall any controversy about it. The
larger lesson to take from it is to think very hard about longer-term
transitions that it might require. In this case, going from
'non-standard' to 'standard' rendered the construct significantly
problematic.
--
Dave Crocker
Brandenburg InternetWorking
bbiw.net
mast:@dcrocker@mastodon.social
_______________________________________________
mailop mailing list
mailop@mailop.org
https://list.mailop.org/listinfo/mailop
