Client/server 2119 blurriness

[Michael Jones]

A review of the use of the 2119 keywords is in order, per Tim's note below...

From: [e-mail] [mailto:[e-mail]] On Behalf Of Tim Bray Sent: Monday, January 28, 2013 9:38 AM To: [e-mail] Subject: [Openid-specs-ab] Client/server 2119 blurriness

If this were in the IETF, you'd get pushback because in the basic/implicit client specs, there are loads and loads of places where it says MUST or REQUIRED but you're not doing normative-2119-stuff, you're instructing client authors what they can reasonably expect from the server; the MUST/REQUIRED applies to OP behavior not client behaivor.

The problem, if it's a problem, could be fixed by making a whole bunch of MUSTs into musts. I’m not sure it’s really a problem, but as a reviewer of way too many IETF drafts, it did feel a little jarring. -T

[Michael Jones]

Add a note to the 2119 section:

When the 2119 language applies to the behavior of OPs, it is there for explanatory value to help client implementer's understand the expected behavior of OPs.

[Michael Jones]

Fixed #716 - Client/server 2119 blurriness

→ <<cset a4f975324dad>>