Markdown Version | Session Recording

Session Date/Time: 29 Apr 2025 19:00

CELLAR

Summary

The CELLAR Working Group discussed the status of draft-ietf-cellar-tags (tags-16), which has commenced a Working Group Last Call (WG LC), and draft-ietf-cellar-codec (codec-14). Key technical discussions revolved around balancing ideal encoder practices with backward compatibility for existing files, the necessity of clear language regarding optional data, and the timing of WG LCs. Progress on draft-ietf-cellar-chapter-codec is pending the completion of the other two drafts, with a significant point of discussion being the accessibility of normative DVD specifications. Administrative items included a request for GitHub labels for the FFV1-V4 document and the process for deprecating the application/x-font MIME type.

Key Discussion Points

• Administrative Items
  • The Note Well was called to attention.
  • It was reminded that the meeting is being recorded and will be available on YouTube.
  • Previous meeting minutes were approved without suggested changes.
  • The chairs noted ongoing efforts to streamline the agenda and documentation processes (e.g., uploading to the data tracker).
• draft-ietf-cellar-tags (tags-16)
  • The document has had several Pull Requests (PRs) merged, leading to the submission of -16.
  • A two-week Working Group Last Call (WG LC) has been initiated.
  • Discussion occurred regarding extending the WG LC to allow more time for implementers (e.g., from the FMEG community or MKVMerge) to review.
  • The sense of those present indicated no strong reason to extend the WG LC, as the IETF Last Call (ITF LC) provides another opportunity for external review, and current community contacts are already engaged.
  • The inclusion of EBUR 128 loudness tags in -16 was discussed, with the editor expressing confidence in the current approach, noting its alignment with professional broadcast formats (like BWF/BMF).
• draft-ietf-cellar-codec (codec-14)
  • The chair raised three open topics for the editor related to previous review comments.
  • "Should" vs. "Must" Language: A recurring discussion point concerned the use of "should" versus "must" for defining behaviors, particularly for codec private data (e.g., VP9) and subtitle mapping. The WG recognized the need to recommend best practices ("good coders ought to do this") while acknowledging that older, existing files or encoders might not adhere to these. Specifications need to clearly explain how decoders should handle cases where recommended information is missing, emphasizing the IETF principle of "strict in what you send and liberal in what you receive" given the static nature of Matroska files.
  • Security Considerations: Clarification was requested for text stating "the mandatory code private may be missing." It was agreed that "mandatory" was incorrect if the data can be missing, and the text should be refined to explain the circumstances under which codec private data might be missing (e.g., when it's "known" to be absent) and how decoders should reconstruct or handle it.
  • Parallel vs. Serial WG LC: A question was posed about whether to run the WG LC for codec in parallel with tags or serially. The sense of those present was that the two documents are largely unrelated and do not cross-reference each other in significant ways that would necessitate serial processing.
• draft-ietf-cellar-chapter-codec
  • Work on this document is currently on hold, awaiting the completion and review of the tags and codec drafts.
  • A major discussion point was the expected public release of DVD specifications at the National Diet Library (NDL) in Japan. Concerns were raised about the accessibility of these specifications for normative referencing in an IETF document, especially if physical presence is required or if IPR restrictions prevent sharing/digitization.
  • The editor indicated that the VideoLAN community is actively pursuing access and potential digitization/documentation of these specifications if allowed.
  • The chairs noted that the IETF would have difficulty processing a Proposed Standard document with inaccessible normative references. The WG will await updates from the VideoLAN community regarding the accessibility of the DVD specifications.
• FFV1-V4 Document
  • A request was made for GitHub labels (e.g., "V4") to help organize issues and PRs specific to the FFV1-V4 document. The editor confirmed that a "V4" label already exists.
• application/x-font MIME Type Registration
  • There was pushback from the IESG on registering application/x-font as deprecated via a telechat.
  • Discussion confirmed that this is the only MIME type in question and that it should be registered as a deprecated form of font/otf.
  • The chair offered to draft a small, one-page document to facilitate this registration process through the working group.

Decisions and Action Items

• draft-ietf-cellar-tags (tags-16)
  • Decision: The Working Group Last Call will proceed with regular timeframes; no extension is deemed necessary at this time.
  • Action: Steve (editor) to ensure the XML version of tags-16 is tagged appropriately in the repository.
• draft-ietf-cellar-codec (codec-14)
  • Decision: The Working Group Last Call for codec can proceed in parallel with tags.
  • Action: Steve (editor) to address the open topics (VP9/subtitle "should" vs. "must" language, security considerations text clarification) in the codec draft.
  • Action: Jerome (contributor) to review the time code part and other aspects of the codec specification.
• draft-ietf-cellar-chapter-codec
  • Action: Steve (editor) to follow up with the VideoLAN community regarding the accessibility and potential for digitization/documentation of the DVD specifications from NDL and report back to the WG.
• application/x-font MIME Type Registration
  • Action: Robert (chair) to investigate drafting a small, one-page document to facilitate the registration of application/x-font as a deprecated MIME type, leveraging Steve's input for technical details.
• Working Group Chairs
  • Action: Spencer and Robert (chairs) to have a follow-up discussion to ensure the WG is on track.

Next Steps

• The WG awaits updates on the codec draft from the editor and Jerome's review.
• Progress on chapter-codec will resume once tags and codec drafts are further along in the publication process and clarity on DVD specification accessibility is gained.
• The chair will initiate efforts for the application/x-font MIME type registration.
• Chairs will continue to monitor progress and coordinate next steps.