Ira,
> On Oct 26, 2020, at 8:34 PM, Ira McDonald <blueroofmusic at gmail.com> wrote:
>> Hi,
>> The lack of REQUIRED "printer-current-time" to populate "date-time-at-xxx"
> attributes is an *old* problem - noticed in the first draft of IPP 2.0. In RFC
> 8011, it's still (incorrectly) RECOMMENDED, which breaks the preferred
> "date-time-at-xxx" attributes (the only useful ones in a Job Log) - and RFC
> 8011 is the authoritative source definition in the IANA IPP Registry, so this
> can't simply be fixed in a PWG 5100.x spec (I think?).
So dateTime attributes were all optional in IPP/1.1 and IPP/2.0. And we updated the definition of printer-current-time to include the 'unknown' syntax since it was already explicitly allowed in RFC 2911, just not included in the syntax definition for printer-current-time...
The issue for the registry is that when I updated the registrations for RFC 8011 I didn't update the printer-current-time syntax to match the document. I'll include a fix for that in my next dump for IANA.
The issue for IPP Everywhere is that 1.0 didn't explicitly require it but *did* require date-time-at-xxx - clearly the intent was to require printer-current-time but we missed it. IPP Everywhere 1.1 makes it RECOMMENDED and notes that the omission from 1.0 was an error since we *did* require date-time-at-xxx. IPP Everywhere 2.0 will be able to make it REQUIRED.
Anyways...
________________________
Michael Sweet