Session Date/Time: 25 Feb 2025 20:00

CELLAR

Summary

The CELLAR Working Group discussed administrative items including agenda adjustments, previous meeting minutes approval, and rescheduling the October interim meeting. A significant portion of the meeting was dedicated to the process for handling the deprecated application/x-font-otf media type, exploring options like Errata, registry updates, or an update RFC ("Abyss document"). The group also discussed improvements to the GitHub repository structure for better discoverability by new contributors and reviewed the status of open Pull Requests (PRs). The control05 document was decided to be dropped from the WG's active work.

Key Discussion Points

Meeting Logistics:
- The chair noted adjustments to the agenda to improve focus.
- Participants were encouraged to share video whenever possible to foster more interactive meetings, with a reminder that sessions are recorded and published to YouTube.
- The IETF Notwell policy was acknowledged.
- Previous meeting minutes were accepted with no comments.
October Interim Meeting Date:
- It was noted that the originally planned October interim date was too close to the IETF meeting week, impacting travel.
- Options for earlier dates were discussed: October 14th or 21st.
- Decision: The October interim meeting was moved to October 14th.
application/x-font-otf Media Type Discussion:
- The group discussed how to formally address application/x-font-otf as a deprecated alias for application/otf.
- Process Options Explored:
  - Expert Review/IESG Approval: Potentially submitting a request to add this to a "deprecated aliases" line in an existing registry for application/otf. This would apply if it's simply an unrecorded fact that was true when the original consensus was made.
  - Errata: Robert clarified that an RFC Editor Erratum is for correcting flaws in publication (e.g., spelling, obvious grammar errors), not for substantive changes or additions that might have been contentious during IETF Last Call. Such changes often lead to a "Held for Document Update" status.
  - Abyss Document (Update RFC): For substantive protocol or behavioral changes (like those mentioned for ebml), a new draft (e.g., draft-ietf-cellar-rfcXXXX-bis) would be the appropriate mechanism, requiring a new consensus call and publication process.
- Current State: RFC 881, which defines font media types and their IANA registry, currently does not have a formal notion or registry for "deprecated media types," even though many were used in the past.
- Question: Should the WG pursue creating a registry for deprecated font media types, or include this information in a future document like the Matroska v5 specification?
- Concern: Scattering information across multiple documents (e.g., RFC 9559 and Matroska v5) could make it harder for implementers to find comprehensive information.
- Next Steps for application/x-font-otf: Robert will discuss this issue with ADs and other relevant experts at the IETF Bangkok meeting to determine the best path forward, and will reach out to Steve for additional information to inform these conversations.
GitHub Repository Organization:
- Robert raised concerns about the accessibility of the GitHub repository for new contributors, specifically finding which files correspond to which documents and how to make changes given the mix of Markdown and XML sources.
- Steve explained that a Makefile defines the generation process, and all XML artifacts are generated on merge actions, though not easily visible. Markdown is the primary editing format for most text.
- Suggestion: Organize the repository into document-specific directories (e.g., five folders for the five current documents), containing all necessary parts to generate each document. This would significantly improve discoverability for contributors.
- Action (Steve/WG): Consider repository reorganization (document-specific directories) and improvements to documentation (e.g., in README) to guide new contributors.
- Action (Steve/WG): Explore ways to make generated XMLs/artifacts more visible from GitHub actions.
Open Pull Requests (PRs):
- There are approximately a dozen open PRs related to spec-tags.
- PR #960, related to parsing numbers for tags, might be tricky and requires input from Moritz (who develops widely used software).
- Action (Spencer): Review the PRs tagged with spec-tags. Steve will provide a link.
control05 Document:
- Decision: The Working Group decided to drop the control05 document as an active work item. Robert will initiate the process to release it from the group.
Other Pending Items:
- Status of F1 and GPU decoder is still pending.
- No updates yet on the DVD format books from the National Diet Library of Japan.
- Action (Steve/WG): Poke the list if information about the DVD format books from NDL Japan becomes available.

Decisions and Action Items

Decisions

The previous meeting minutes were approved without comments.
The October interim meeting date was moved to October 14th.
The control05 document has been dropped as a CELLAR Working Group item.

Action Items

Robert: Discuss the process for handling the application/x-font-otf media type (registry update, Errata, or Abyss document) with relevant ADs and IETF experts at the IETF Bangkok meeting. Follow up with Steve for any necessary information.
Steve:
- Provide Spencer with a link to the spec-tags related Pull Requests for review.
- Consider reorganizing the GitHub repository into document-specific directories to improve discoverability for contributors.
- Explore ways to make generated XMLs/artifacts more visible from GitHub actions.
- Consider adding documentation to the repository's README to guide new contributors on how to find and edit files.
Spencer: Review the spec-tags related Pull Requests (e.g., #960).
Working Group (General): Monitor and inform the mailing list if any news regarding the DVD format books from the National Diet Library of Japan becomes available.

Next Steps

Further discussions and determination of the appropriate IETF process for application/x-font-otf will occur, potentially involving ADs and other experts at the upcoming IETF Bangkok meeting.
Review of open Pull Requests, particularly those concerning spec-tags, will proceed.
Potential improvements to the GitHub repository structure and documentation will be explored to enhance contributor experience.