Re: coments on draft-ietf-sacred-protocol-bss-00

> Regarding error codes, I really don't want to continue that thread,
> but I can't help noting that the current protocol I-D defines new BEEP
> error codes to cater for SACRED specific errors.

hi. i don't think you understand section 8 of the beep spec.

the list of reply codes given isn't an exhaustive list of reply codes that
can occur in an "error" element.

rather, it's a list of codes that the working group agreed on for common
use. for example, you'll see those reply codes (and others) in
implementations of channel 0, some of the sal profiles, the tls profile etc.

if you look at some other profiles that use beep, you'll see that some of
them choose to use the "error" element defined in beep. some don't.

those that do, use the same list or may add some codes.

so, thinking that section 8 is some kind of a totality is wrong.


now, it turns out that there is an issue with the new reply codes defined in
the sacred spec. specifically, rfc 821, which defines "the theory of reply
codes" talks about the significance of the 'x', 'y', and 'z' digits that you
find in a reply code, and the codes that sacred defines, doesn't fit into
that taxonomy. when i get home and can consult 821 in a quiet room, i'll
send a note suggesting what the actual xyz values should be...

/mtr