Re: [Cbor] draft-ietf-cbor-tags-oid-00.txt

[[e-mail]]

Sean Leonard <[e-mail]> writes:
> Whoops, you are right. I think I was thinking of something else, namely, 
> that bit 8 cannot appear at the end of an encoded OID: it has to end 
> with 0x00-0x7F.

Well, you could delimit a sequence of concatenated OIDs by *prefixing*
each one with 0x80, so you can scan the byte string for occurrences of
0x80 immediately preceded by a byte 0x00-0x7F.  That costs 1 byte per
OID, and is probably the best you can do with a simple system.  OTOH, a
sequence of CBOR-encoded octet strings that are less than ~23 octets
only requires a one byte prefix for each as well, if the OID-ness is
implied by some mechanism.

Dale