----- Original Message ----- From: "Dave Cridland" <dave@xxxxxxxxxxxx> Sent: Wednesday, March 30, 2016 8:00 PM > On 30 March 2016 at 18:59, Barry Leiba <barryleiba@xxxxxxxxxxxx> wrote: > > For future reference, I tend not to zero-index my sentences. ;-) > > I think that MUST/SHOULD/MAY (the former two tempered by NOT) are > well-understood, although the strength of SHOULD is usually underestimated. > OPTIONAL is probably obvious enough (though its implications may not be), > and SHALL/RECOMMENDED are uncommon enough that they're probably not > understood nearly as well. I would echo Brian's earlier comment that it is SHOULD and MAY that are troublesome, while MUST and SHALL mean what they usually do and so do not require further thought. I read an I-D for a protocol using HTTP as transport that said that if the server found the client not authorised to access a resource, then it SHOULD return a 404 but MAY return a 403. I think that that usage is ok but I also think it not compliant with RFC2119 and it is the kind of scenario that the RFC ought to address. The underlying issue I see for this case is that 'we are using HTTP and have no control over what it says and you will be using an off the shelf HTTP package and so will have no control over what it generates so like it or lump it (strictly a personal interpretation of this). Also, I see a number of ADs pushing for a SHOULD to be accompanied in the I-D by the valid reasons why it will not happen, as opposed to leaving it up to the implementor, and again, I see this as not compliant, but again, I see it as a good idea. Tom Petch > > As a complete side thing, I wonder how this all seems to > > German-speakers, as German uses initial caps for all nouns. I wonder > > if anyone even notices if someone fails to do that. I wonder if it > > becomes puzzling, perhaps in some instances. > > > > Barry > > >