**Session Date/Time:** 13 Jun 2023 14:00 # [SCHC](../wg/schc.html) ## Summary The SCHC working group meeting covered several critical topics, starting with administrative updates and the status of LPWAN documents. A significant portion of the meeting was dedicated to reviewing and resolving open errata for RFC 8724 and RFC 8824, with specific technical points discussed and agreed upon. The session also featured a detailed presentation on optimizing CID (Context ID) delegation for SCHC to achieve more compact representations and facilitate future extensions, prompting a discussion on the working group's role in managing these allocations and aligning with the `core-set` document. Finally, a brief overview of the SCHC OAM (Operations, Administration, and Maintenance) draft was presented, highlighting new features like proxying ICMP messages and generating error reports, though detailed discussion was deferred due to time constraints. ## Key Discussion Points * **Administrative Updates**: * The agenda was adjusted to discuss Errata before CID delegation, with no objections raised. * All active LPWAN documents have been moved to SCHC, and those in ISG review are being handled. RFC 8724 and RFC 8824 are in edit state, and their completion will lead to RFC publication for both. * **Working Group Item Status**: * **SCHC over PPP**: The chair noted that a call for adoption had not yet been initiated and requested Pascal to send an email to the working group to start the adoption call. * **LPWAN Architecture as SCHC Architecture**: This document's continuation under SCHC was affirmed. * **OAM Document**: To be discussed later in the meeting, and formally adopted in the next month. * **Bob's Draft**: Discussion deferred to the next IETF. * **Errata Discussion (RFC 8724 and RFC 8824)**: * Nine errata were discussed, with five originally editorial errata reclassified as technical by the RFC Editor. * **Errata 48824 (Target Value (TV) definition)**: Discussion focused on the appropriate term for TV when `CDA` is `value-sent` and `MO` is `ignore`. Options considered included "empty", "undefined", "not sent", "not set", and "zero length value". The group leaned towards "not set" or "unsetting" for consistency. Marco suggested that "empty" could be ambiguous (e.g., zero-length string). Anna clarified that "not sent" was likely a misspelling of "not set". * **Errata 7391 (CoAP Option user vs. message)**: It was clarified that the text regarding "CoAP option seems like can use a message while the user here and is the client" was confusing. A rephrasing to focus on the situation and the "real user" (client) was proposed and generally accepted. * **Errata 94 (MO field digit '1')**: This was identified as a historical typo, an "equal 1" for a matching operator that does not exist. The proposal to remove the "equal 1" was accepted. A suggestion to replace "path" with "LM" was considered a separate issue. * **Errata 25 (Compressing CoAP type rules)**: The existing text implying `type` is always sent was noted as inconsistent with cases where it's not sent (e.g., `confirmable`). An expansion of the text to align with the table rules was accepted. * **Errata on Client Compression, CoAP Options Plural, CoAP Option Behavior**: * Rephrasing "compressing the client" to "compressing the request" was accepted as editorial. * Correcting "Oscar option" to "Oscar options" (plural) was accepted as editorial. * Clarifying the behavior of CoAP options regarding `iPath` (not mandatory in requests, not supposed to be in responses) was accepted as technical. * **Other Editorial Errata**: Other minor typos were presented and considered uncontroversial. * **Overall Errata Resolution**: The group generally agreed that the errata were correct or required minor rephrasing (e.g., 7391). Marco thanked the group for the thorough review. * **CID (Context ID) Delegation**: * A presentation detailed an optimized CID allocation process for SCHC, aiming for very compact representation of YANG data models, especially for extensions. * The current `core-conf` Delta encoding could lead to larger values when mixing leaves from different YANG models, resulting in more bytes for Delta representation. * **Proposal**: Instead of allocating CID ranges per RFC, the working group should ask IANA for a pool of addresses (e.g., 500-1000 CIDs). The WG would then manage manual allocation within this pool to optimize Delta encoding (e.g., placing frequently extended elements closer to the center of a block to allow smaller Deltas). * Measurements showed a significant reduction in representation size (up to 10%) compared to public ranges or standard IETF space. * Request to ISG for using very small seeds for common values (e.g., Matching Operator, CDA) was also suggested. * Karsten (co-author of `core-set`) expressed support for the approach, emphasizing the need to ensure the `core-set` document's text recognizes and supports such elaborate allocation schemes rather than solely recommending a default procedure. * Discussion arose regarding *who* manages these allocations: the working group, a Designated Expert, or IANA. It was noted that `core-set` currently places the onus on the Designated Expert working with the WG. * **OAM Draft (Operation, Administration, and Maintenance)**: * Due to time constraints, only a brief overview of the OAM draft (focused on ICMP messages) was given. * **Key Features**: * **Proxy functionality**: The SCHC Core can directly answer ICMP Echo requests (pings) for a device for a defined period, without forwarding them over the LPWAN, to check device liveness. * **ICMP Generation**: The SCHC Core can generate compressed ICMP error messages (e.g., "Rule not found", "Port not found", "Packet Too Big") and send them to the device if an error occurs during compression or fragmentation. * **New Matching Operators (MA)**: `match-rule` and `reverse-match-rule` were introduced to facilitate ICMP processing. * A concern was raised about the potential for rule ordering to affect reverse compression, but the speaker clarified that the device uses the Rule ID and is not concerned with exact bit-level matching. ## Decisions and Action Items ### Decisions * The agenda for the meeting was adjusted to discuss Errata before CID delegation. * For the errata concerning RFC 8724 and RFC 8824: * Most proposed errata corrections were accepted by the group. * For errata 48824, regarding Target Value (TV), the group agreed to use "not set" or "unsetting" consistently in the text. * Errata 7391 required a minor rephrasing before full acceptance. * The SCHC working group confirmed its acceptance to continue work on the LPWAN Architecture document as the SCHC Architecture. ### Action Items * **Pascal**: Send an email to the SCHC working group to initiate the call for adoption for the SCHC over PPP document. * **Chairs (Pascal/Alex)**: Discuss the errata acceptance with Eric and post the formal acceptance of the errata (with the minor rephrasing for 7391) to the mailing list. * **Working Group**: Continue discussion on the proposed CID delegation strategy, including the working group requesting a pool of CIDs from IANA, the management of manual allocation, and the potential for ISG-reserved common values. * **Karsten**: Work on ensuring the `core-set` document's text properly recognizes and supports the proposed elaborate CID allocation procedures for SCHC. * **OAM Draft Authors**: Continue work on the OAM draft, including offline discussions on the proposed features, and prepare for its adoption call next month. ## Next Steps * Formalize the call for adoption for SCHC over PPP. * Address the remaining discussions and refinements for the CID delegation strategy, including the relationship with the `core-set` document. * Progress the OAM draft towards a working group adoption call, incorporating feedback from the brief discussion and further offline exchanges. * Consider developing a "base" document for future comprehensive errata and document updates.