> On reflection, I should worded my last statement as "clients SHOULD use ipp
> schemes, but MAY use http schemes to contact servers. Servers MUST support
> connections using either http or ipp schemes.
Okay. If we're talking about URLs that go in HTTP request and response headers,
I'd agree with that. The big question I have is the URLs that go in IPP
protocol elements. I think they SHOULD (perhaps MUST) be ipp:.
More to the point, regardless of what is done on the wire, I think the user
should always use and see ipp: URLs when referring to a printer.
Keith
> Like I said earlier, I think this will all work, but a detailed I-D will be
> more complete with examples and such.
>> On a different tack, I was hoping we could just get away with using
> different methods for IPP, but I was soundly voted down in a past
> conference call. If the IESG requirement covers more than just being able
> to distinguish IPP traffic from HTTP traffic, then I think a separate
> scheme is the way to go. I'm still re-reading your (Keith) last few
> messages to see if I can extract the exact issue(s) the IESG is concerned
> with. I'm hitting the road tomorrow for our meeting so I hope to have a
> handle on this by Monday.