SPDX Tech Team Meeting 2023-01-03
Records minutes from an SPDX Tech Team meeting covering 2023 planning, quality SBOMs, serialization, and profile updates.
What this file does
Records minutes from an SPDX Tech Team meeting covering 2023 planning, quality SBOMs, serialization, and profile updates.
When to use it
- Reviewing SPDX meeting minutes for historical context
- Understanding SPDX 3.0 serialization and security profile discussions
- Tracking SBOM quality metrics and tooling conversations
- Referencing FOSDEM SBOM devroom planning details
SPDX Tech Team Meeting 2023-01-03
Attendees
- Alexios Zaras
- Bob Martin
- David Kemp
- Dick Brooks
- Jennie Kauffmann
- Joshua Watt
- Kate Stewart
- Kevin Cressman
- Matt Roberts (OpenC2)
- Nisha Kumar
- Rose Judge
- Sen Hastings
- William Bartholomew
Agenda
- 2023 Planning
- Updates from subteams
Notes
-
Implementers Forum - next wednesday, start with discussion from DocFest.
- Adolfo meeting in February
- Spec - specific fields.
- Golden Set of examples
-
Looking for Adolfo on update CI/CD to validate examples being checked in.
-
Brining in some of OpenSSF
-
Quality
- What makes a good SPDX document?
- Validates, NTIA minimum
- Agreement of what is a qualilty SBOM - used for different purposes.
- Chainguard hosting a panel discussion, what makes quality SBOM on Jan 18th.
- % of licenses filled, % of noassertion - tooling to measure SBOM quality.
- SBOM's created for specific purpose - for Risk Assessment, may not need Licensesing, and have minimum elements, and content needs to align with NVD for
- Bob - Meeting on 24th/25th - update on NVD & CVE efforts. Panel on VEX and SBOMs. Disconnects to iron out.
- Nisha - biggest frustration is identifying the component itself. Make use of several pieces of meta data. Straightforward for some software components. Tools tackling low hanging fruit, but not sure what to do with other stuff. Package managers don't always give configuration effectively. Tools and infrastructure knowledge are needed for effective triangulation. Knowledge is spread out amongst multiple people. If doing manually - document why this is the right source. OpenJDK vs. Oracle Java, etc. Not everyone knows how to triangulate what packages are where they came from.
- Rose - quality is different to different use cases - guidance based on usage intention of SBOM. Intended Usage of SBOM.
- How to Use different types? Management of issues associated with Quality of SBOM.
- Dick looking for tooling from SPDX to CDX - referred to Steve.
-
FOSDEM
- Will have an SBOM Devroom - getting back answers from those with difficulty traveling.
- All others confirmed. Looking for proposing a new discussion - looking for panel ideas - email to devroom organizers.
- The organizers of the FOSDEM SBOM devroom can be reached by sending email to sbom-devroom-manager@fosdem.org. Please do not hesitate to contact us if you have any inquiry or suggestion for the devroom.
-
Security Working sessions
- Come up with model similar to build profile, and solidifying relationships to be submitted.
- Thomas or Jeff - looking for update from.
-
Dick referenced ICT-SCRM Task Force CISA side
- Software assurance, processing SBOMs.
- What to request and how to interpret what they get back.
- Sub teams working on draft work. Open question how to get external review for some of these deliverables.
- https://www.cisa.gov/ict-scrm-task-force
-
David Update
- serialization, taking offline with recommendations. Some discussions with Max. Asking to add a serialization folder to repo. Serialization concepts. Content in folder. https://github.com/spdx/spdx-3-model/pull/57. Note: this is limited to JSON discussions at this point.
- The README describes Information Model based serialization design, and specific JSON values shown on the logical diagram. https://github.com/davaya/spdx-3-model/tree/serialization/serialization
-
Joshua update
- SBOMs of SBOMs.... needing a serialization format, looking for bare JSON
- Some things need serialization. current blocker.
-
Nisha Update
- TUF - metadata in a certain format, different serializations, no control over serializations, so hashing becomes problematic. How to solve this is ambiguous, and used a work around. Specification for an evelope. Outside metadata. JSON object describing, avoids dealing with serialization. For purpose of integrity checking. Signatures around Data.
- Desire to go to element level, and supporting different serialization formats, is part of our SPDX approach. Canonical form of JSON, is needed.
-
Hardware Profile - targetting to 3.1
- Sen is interested in being included
- William knows folks who will be interested as well.
Next Meeting Topics
- Resolve: Agents & Identities in 3.0 Model approach
- Review of the canonical Serialization rules defined to date, then review PR.
- Update on where we are with Security & Licensing profiles.
- Other PRs against SPDX 3.0 to be included.
What's inside
Attendee list, agenda, 8 topical discussion sections, and next meeting topics.
Change this for your project
- Replace
https://github.com/spdx/spdx-3-model/pull/57with your own PR reference - Replace
https://github.com/davaya/spdx-3-model/tree/serialization/serializationwith your own repo branch - Replace
https://www.cisa.gov/ict-scrm-task-forcewith your own external link if needed
Where it goes
Keep alongside your test suite. Used to define and score model evaluations.
Worth borrowing
- Structuring meeting notes with attendees, agenda, and topical subsections
- Capturing action items and next meeting topics for follow-up
Related Documents
IATA BCBP Standard Compliance
Documents which IATA BCBP fields and barcode formats a Swift library implements, including Version 8 gender code support.
Evaluating AI Agent Systems: Metrics, Benchmarks, and Quality Assurance (2024-2026)
Surveys 2024-2026 metrics, benchmarks, and monitoring tools for evaluating AI agent systems, with recommendations for a self-improving coding agent.
Voice AI Leaderboards, Benchmarks, and Evaluation Gaps (Jan 2025 -- Feb 2026)
Surveys 20+ voice AI benchmarks from Jan 2025, Feb 2026, identifies evaluation gaps, and provides leaderboard data for STT, TTS, and end-to-end voice agents.
LLM Evaluation — Deep Dive
Covers why LLM evaluation is hard, benchmark taxonomy, LLM-as-judge methodology, contamination, robustness, and designing an eval suite for a product.