[urn] Re: Registration request for urn:thread:
Peter Saint-Andre <[email protected]>
| Newsgroups | gmane.ietf.urn |
|---|---|
| Message-ID | <[email protected]> |
On 11/19/24 12:57 PM, Dale R. Worley wrote: > Esko Dijk <[email protected]> writes: >> Also we've updated our registration request template based on the >> review and to provide the reference to this online registry. >> New template: >> https://github.com/ThreadGroup/urn-thread-registry/blob/main/urn-registration-v1.txt > > I'm OK with the concept. As to the formatting of the template, I'd like > to see the lines in urn-registration-v1.txt be wrapped. As my browser > displays that URL, each paragraph is one single line that wanders very > far off to the right, and I suspect for other people it's the same. Just FYI: if necessary, as team lead I clean up registration requests (placing line breaks after 72 characters per line) before I submit them to IANA. > As to the syntax: > > urn-thread = "urn:thread:" sub-namespace *( ":" sn-content ) > sub-namespace = sn-label *( ":" sn-label ) > sn-label = 1*unreserved > sn-content = *( pchar / "/" ) > > A thread URN consists of "urn:thread" followed by one or more sn-label's > and then by zero or more sn-content's. Note that one can't tell the > difference between sn-label's and sn-content's by syntax, as both can be > composed entirely of pchar's. So splitting the sn-label's from the > sn-content's requires additional information. I suppose you might > expect a parser to know all of the possible sub-namespace's, but that is > difficult to maintain. > > Also, colon is a pchar, so an sn-content can contain a colon. That > means that a series of multiple sn-content's cannot be parsed without > additional information. > > I think you may want to revise the syntax specification. Agreed. Peter _______________________________________________ urn mailing list -- [email protected] To unsubscribe send an email to [email protected]