John C Klensin wrote:
I don't disagree with anything Ned said in his note, but
want to reinforce the principle that trying to make too-fine
distinctions in this area won't work in practice and may
lead to even more confusion.
Clearly Hector and SM will speak up what they had in mind,
a 5xx *instead of* 354 (but then the 4xx is odd), or a 5xx
*after* 354 (as in the strawman example), or maybe both...
My assumption "after 354" differed from Ned's assumption,
unsurprisingly our conclusions also vary.
I like Ned's RFC 2920 example, and knew that I should have
looked it up, but it is none of the RFCs I know "by number",
and I was lazy. Where is Bruce when I need him to keep me
Your example is also nice, and if somebody posts a draft
with "retry clarifications" I hope that will also discuss
the PIPELINING fine print.