Hi Carsten, Thanks.
Recall the (conflicting) design objectives: 1. Optimize for expected deployment. 2. Simple processing. Reduce special cases in the encodings. Addressing 1, the current proposal has two special encodings of ‘name’, EUI-64 and IP addresses written as hex text strings, where the lower case are from the aviation application Bob mentioned. Encoding general upper case hex text strings as tag 23 would be a third special encoding, which works against 2, but it is no major complication. I’ll made a comment on the PR with this content. But I don’t see the problem with keep using the untagged CBOR byte string to represent lower case hex text strings. No new CBOR tag is needed, less overhead. Göran From: Carsten Bormann <[email protected]> Date: Thursday, 16 January 2025 at 16:31 To: Göran Selander <[email protected]> Cc: Robert Moskowitz <[email protected]>, [email protected] <[email protected]>, Joel Höglund <[email protected]> Subject: Re: [COSE] I-D Action: draft-ietf-cose-cbor-encoded-cert-12.txt On 2025-01-16, at 13:55, Göran Selander <[email protected]> wrote: > > • text strings of byte strings (even number of characters ‘0’-‘9’ and > ‘a’-‘f’) are encoded as (untagged) CBOR byte strings. We have tag 23 for representing hexadecimal strings as byte strings (which I think this is about) [1]. [1]: https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.rfc-editor.org%2Frfc%2Frfc8949%23name-expected-later-encoding-for&data=05%7C02%7Cgoran.selander%40ericsson.com%7C7643b1e3350545bc85b208dd3642c9c6%7C92e84cebfbfd47abbe52080c6b87953f%7C0%7C0%7C638726382695597010%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=Wa0lII4nuwuVssH6q8QZMMVLt6K2%2Bjof%2BGXvgHDs6s8%3D&reserved=0<https://www.rfc-editor.org/rfc/rfc8949#name-expected-later-encoding-for> However, this is for uppercase strings. About half of the hex strings I’m seeing are lowercase. I’m assuming here that makes a difference in the X.509 signature? I propose that we add a (1+1) tag for those two (e.g. 108 for ‘l’ as in lowercase). (No help with mixed case, but I haven’t seen those.) Grüße, Carsten
_______________________________________________ COSE mailing list -- [email protected] To unsubscribe send an email to [email protected]
