test
", "time": "2024-07-23T00:30:19Z"}, {"author": "Akira Tsukamoto", "text": "hi
", "time": "2024-07-23T00:32:12Z"}, {"author": "David Brown", "text": "BTW, the \"notes\" for this session take me to notes for COSE.
", "time": "2024-07-23T00:36:29Z"}, {"author": "Dave Thaler", "text": "https://notes.ietf.org/notes-ietf-120-suit?edit
", "time": "2024-07-23T00:36:42Z"}, {"author": "David Brown", "text": "That looks good. The \"PDF\" link is from COSE, though.
", "time": "2024-07-23T00:37:13Z"}, {"author": "Deb Cooley", "text": "that's weird.
", "time": "2024-07-23T00:39:06Z"}, {"author": "Carsten Bormann", "text": "A statement has effects, an expression also returns a value
", "time": "2024-07-23T00:46:29Z"}, {"author": "David Brown", "text": "I wouldn't consider \"statement\" and \"expression to be old vs new.
", "time": "2024-07-23T00:46:36Z"}, {"author": "David Brown", "text": "re Carsten's comment.
", "time": "2024-07-23T00:46:53Z"}, {"author": "David Brown", "text": "They have different and specific meanings.
", "time": "2024-07-23T00:47:01Z"}, {"author": "David Brown", "text": "To clarify, are was the intent of \"utf-8\" to say it isn't actually a UCI, but something encoded as UTF-8.
", "time": "2024-07-23T00:50:56Z"}, {"author": "Dave Thaler", "text": "If you say it's a \"URI\" then it's pct encoded by definition. Otherwise it's not a URI (if it has any i18n)
", "time": "2024-07-23T00:52:31Z"}, {"author": "Dave Thaler", "text": "\"suit-parameter-uri
\nA UTF-8 encoded URI Reference [RFC3986] from which to fetch a resource.\"
It should say, when percent-decoded, it's UTF-8.
", "time": "2024-07-23T00:55:10Z"}, {"author": "Carsten Bormann", "text": "The URI is ASCII
", "time": "2024-07-23T00:55:20Z"}, {"author": "Carsten Bormann", "text": "I would recommend to retain that restriction
", "time": "2024-07-23T00:55:47Z"}, {"author": "Carsten Bormann", "text": "I would recommend not to do that
", "time": "2024-07-23T00:55:59Z"}, {"author": "Dave Thaler", "text": "Why would a SUIT parser want to ever percent-decode it? It's for machine access only. The only reason I know of to pct decode is for human display which the document says we don't do here.
", "time": "2024-07-23T00:56:28Z"}, {"author": "Carsten Bormann", "text": "because that is the way we handle URIs in CoAP options.
", "time": "2024-07-23T00:56:59Z"}, {"author": "Dave Thaler", "text": "It sounds carsten like you disagree with the statement \"URIs are for machine access only\"
", "time": "2024-07-23T00:57:05Z"}, {"author": "Dave Thaler", "text": "er SUIT URIs
", "time": "2024-07-23T00:57:15Z"}, {"author": "Carsten Bormann", "text": "That is not what I'm saying.
", "time": "2024-07-23T00:57:21Z"}, {"author": "Dave Thaler", "text": "I haven't heard a technical reason to say UTF-8. \"Because COAP did\" isn't a technical reason I can understand.
", "time": "2024-07-23T00:57:58Z"}, {"author": "Dave Thaler", "text": "so maybe you can elaborate on what the issue is?
", "time": "2024-07-23T00:58:10Z"}, {"author": "Carsten Bormann", "text": "You cannot use URIs that don't decode to UTF-8 with CoAP. I'd say that is a pretty good reason.
", "time": "2024-07-23T00:58:28Z"}, {"author": "Dave Thaler", "text": "I think you can, you just can't use the coap uri tag. But I don't think we have anything using coap now.
", "time": "2024-07-23T00:59:30Z"}, {"author": "Dave Thaler", "text": "with suit
", "time": "2024-07-23T00:59:46Z"}, {"author": "Carsten Bormann", "text": "CoAP has options, Uri-Path, Uri-Query, etc.
", "time": "2024-07-23T01:00:02Z"}, {"author": "Dave Thaler", "text": "lets say someone wants to use SUIT with another protocol that only permits (say) UTF-16LE pct-encoded URIs.
", "time": "2024-07-23T01:00:33Z"}, {"author": "Carsten Bormann", "text": "Don't use that protocol.
", "time": "2024-07-23T01:00:45Z"}, {"author": "Dave Thaler", "text": "shouldn't a suit manifest be generic and usable with any protocol?
", "time": "2024-07-23T01:00:47Z"}, {"author": "Carsten Bormann", "text": "Yes. including CoAP.
", "time": "2024-07-23T01:00:58Z"}, {"author": "Dave Thaler", "text": "why would coap be special per se?
", "time": "2024-07-23T01:01:05Z"}, {"author": "Carsten Bormann", "text": "It has some relevance in the constrained space
", "time": "2024-07-23T01:01:19Z"}, {"author": "Dave Thaler", "text": "pick any 2 protocols with mutually exclusive encoding requirements. SUIT manifest should still work.
", "time": "2024-07-23T01:01:22Z"}, {"author": "Dave Thaler", "text": "so it can be used with CoAP as long as you use UTF-8 encoded URIs. But it doesn't mean that SUIT manifests that are only distributed via other protocols have to follow that constraint.
", "time": "2024-07-23T01:01:50Z"}, {"author": "Carsten Bormann", "text": "You can design the protocol to use EBCDIC. The IETF has decided to use UTF-8 some 25 years ago, and where we have a choice, we should heed that.
", "time": "2024-07-23T01:03:59Z"}, {"author": "Carsten Bormann", "text": "Ignore idnits.
", "time": "2024-07-23T01:05:19Z"}, {"author": "Dave Thaler", "text": "theres also no requirement that pct encoding has to be a string at all. And I don't believe SUIT manifests should require that either. They're just URIs.
", "time": "2024-07-23T01:06:48Z"}, {"author": "Carsten Bormann", "text": "That is a design decision you can make. It is somewhat hostile to people trying to use the protcol, which is why I'd recommend against it.
", "time": "2024-07-23T01:12:45Z"}, {"author": "Carsten Bormann", "text": "There probably is a normative reference?
", "time": "2024-07-23T01:17:34Z"}]