Meeting Minutes

[lennartbecker-d]

Weekly 2025-06-05

Participants

@mishu-dev
@FelixMoelders
@alexandruiulian10
Oliver Eggerler
Jonas Lammers
@MaximilianToe
@lennartbecker-d
@Erikhu1
@kalu-an

Agenda & Meeting Minutes

• @MaximilianToe, @Erikhu1 to emails , groups , slack (WELCOME 😊)
• @mishu-dev Write functionality – should not be top priority , let’s focus on read at process and qualification level at the moment till clarified at score level what are the current priorities
  o @mishu-dev synch with Continental Persistence team related to their timeline
  o We need feature request + requirements for json write from persistence
• @mishu-dev collect all Continental github ids and forward them to @Becker, Lennart – to setup github organization for json eclipse score
• @lammjo MISRA C++ 2023 – current there is no SCC toolchain in Eclipse SCORE – clang tidy , Coverity , github actions – d-fine please check
• @alexandruiulian10 how to handle the misra findings: by category ex: mandatory > required > advisory ; Conti needs to clarify with baselibs/process guys
• @kalu-an – let’s start from a frozen version – make the changes , and cherry pick or plan major update if needed ,
  o Current used in SCORE baselibs bazel_dep(name = "nlohmann_json", version = "3.11.3")
  o Current last stable 3.12.0 Bazel Central Registry
  o Even with small delta 3.11.3 -> 3.12.0 lot of fixes and features came in
  How do we want to keep track with further fixes, new featues with nlohmann repo?
• @mishu-dev please check if this site helps for description or MISRA rules https://de.mathworks.com/help/bugfinder/ref/misrac2023rule21.6.1.html

Weekly 2025-06-12

Participants

@mishu-dev
@FelixMoelders
@alexandruiulian10
@Erikhu1
@kalu-an

Agenda & Meeting Minutes

• We should follow the following process instead of MISRA
o https://eclipse-score.github.io/process_description/pr-27/trustable/index.html
o Trustable Software Framework (TSF)
o https://github.com/orgs/eclipse-score/discussions/1225
• .rst files can be used for collecting the evidence
• Other possibilities to collect and link the evidence
• project pipeline is in progress , including automation (Erik)
• remove Fabian from all related meeting etc (Mihai)
• unit-tests are needed , compiler checks
• we need to check in the future if google is test is needed for nlohmann json
• exception handling is not desired !!!!
• external incubator repository was set up: https://github.com/score-json/inc_json
• official incubator is still pending: https://github.com/eclipse-score/
• TSF work packages:
o test coverage – tenet to be created – folder creation - d-fine
o assertion behaviors – functional testing (checklist) - Conti to check
• process checklist needs to be searched
o integration example – what is needed for a pull request to be open for merging : at least 1 review , all findings fixed or motivated and all open discussions closed
o whenever downstreaming or fixing bugs on our side we need a process for merging this into score space
• example of docs format process: https://eclipse-score.github.io/score/main/platform_management_plan/software_development.html
Folder structure example
• let's follow the folder structure like this: https://gitlab.com/richardmaw-codethink/safety-monitor/-/tree/main/trustable?ref_type=heads
• let’s set the filenames as corresponding tenet step etc: docs/trustable/tenets/ta-tests.rst
• commit message format: trustable: Add SMAs about alert channels

Weekly 2025-06-25

Participants

@mishu-dev
@FelixMoelders
@alexandruiulian10
@lennartbecker-d
@NilsEb
@kalu-an

Agenda & Meeting Minutes

• https://github.com/orgs/score-json/discussions not working anymore
• Access to external inc-json is available
• We need to have formulated requirements for TSF work: https://github.com/eclipse-score/score/pull/822/files
• TSF open topics:
  o How to structure the document
  o We should create statements that must be short and we should link evidence
   We need complete set of requirements – translated to statements – checked if wright or wrong
   Check for safe layout – we need to proof that software fulfills the requirements based on evidence
  o Lets’ compile a slide deck with our ideas of practical appliance TSF on eclipse score json lib
  o Markus Schu / Paul Sherwood as contact for Eclipse SCORE TSF
• Official inc_json repository is set:
  o https://github.com/eclipse-score/inc_json
  o Check for contributor access on eclipse-json
  o Options:
   github subtree: ex ./nlohmann/json github subfolder (current under investigation)
   or direct fork – more limited because of the score overhead
  o Rediscuss with baselibs how the inc_json repo will be used
  o Fabian and Maximilian should be removed from the meeting
  o nils.eberhardt@d-fine.com should be added to the meeting

[mishu-dev]

Weekly 2025-07-03

Participants
@lennartbecker-d
@NilsEb
@FelixMoelders
@mishu-dev

Agenda & Meeting Minutes

• @mishu-dev contact comm, persistency, logging , feo for additional functional component requirements for json parser lib: topic initiated via email
• @mishu-dev contact infrastructure to clarify json-lib setup: Topic initiated here https://sdvworkinggroup.slack.com/archives/C0894QGRZDM/p1751885984111959
  • submodule/subtree upstream might be problematic
  • org nlohmann - score_json (submodule/subtree for nlohmann ) - baselibs(external bazel reference) - preffered due easy of use
  • org nlohmann - score_json fork (downstream + json lib changes) - score-json for docs and submodule/subtree - baselibs(external bazel reference)
• @alexandruiulian10 new PR findings need to be addressed
• d-fine working on Assertions developers vs integrators for test TSF
• how can we preserve history when migrating from external json to eclipse inc_json
• proposal of commit messages for trustable: trustable: Restructure so we have a top level document
  • ex: https://gitlab.com/richardmaw-codethink/safety-monitor/-/tree/main/trustable/trustable?ref_type=heads

[mishu-dev]

Weekly 2025-07-10

Participants
@lennartbecker-d
@NilsEb
Oliver Eggeler
Richard Trippel
@FelixMoelders
@mishu-dev
@alexandruiulian10

Agenda & Meeting Minutes

• we need to clarify the requirements of the component request (which are appliable to nlohmann json) , the component request is for the eclipse score json component (public api + glue code + nlohmann json). How can we appoint that only selected requirements from the component are needed for nlohmann json (in TSF for example) ?
• we should separate the code from the documentation (Markus Schu) - different repositories
• how we should handle the critical findings from the SCC tools , we need to define our quality goals
• compiler warnings should be fixed
• how to address c++ exceptions , example can we treat exceptions in glue code?
• to be checked if there is an possibility to disable or other workaround for exceptions in nlohmann code - is abort accepted in score?
• https://json.nlohmann.me/home/exceptions/#switch-off-exceptions
• exceptions handling in the baselibs: https://github.com/eclipse-score/baselibs/tree/1d363886de9908fd9228e1bfd86b0ef912e4ee58/score/language/safecpp/aborts_upon_exception

Codethink topics:

• we need to find a way to separate nlohmann requirements from the score json public requirements
• how we should handle the critical findings from the SCC tools , we need to define our quality goals
• which TA and TT should be addressed by integrators or dev-teams
• which github setup is more suitable to TSF

[mishu-dev]

Weekly 2025-07-17

Participants
@NilsEb
@FelixMoelders
Richard Trippel
@mishu-dev
@alexandruiulian10

Agenda & Meeting Minutes

• Decisions should be copied into eclipse score space according to score process
• eclipse score integration of nlohmann/json into eclispe score:
  • https://github.com/eclipse-score/inc_json this needs to be removed @mishu-dev
  • we need a direct fork to be setup from nlohmann/json (Markus Schu prefers this solution) @mishu-dev
  • we add documentation and other infrastructure on it via commits
  • we should use the latest stable release tag
  • we can create tags according to our releases
  • when is qualified we create a tag with trustable ex 1.1-trustable-release (to be discussed with ops
• pull request - we need to converge to completion
• exception handling on nlohmann/json alternates to be checked with score (ex: https://github.com/orgs/eclipse-score/discussions/301)
• @mishu-dev send email with enrolling and getting contributor access to eclipse-score

[mishu-dev]

Weekly 2025-07-24

Participants
@NilsEb
@mishu-dev
@Jonas-Kirchhoff
@kalu-an

Agenda & Meeting Minutes

• eclipse score json fork , waiting for merge (review done): Create Nlohmann JSON FORK eclipse-score/.eclipsefdn#58
• Feature Request: JSON Parser eclipse-score/score#822 - nlohmann does not support schema validation out of the box, there is one package that can be put on top that supports schema validation
• @mishu-dev contact technical circle to check if exception handling via abort is acceptable for nlohmann json library, exception mechanism allows json params
• @NilsEb TSF - setup of initial version of statements , codethink synchronization , trustable graph to be integrated into the documentation , link evidence and setup info for SME
• baselibs issues status board - BAS - Baselibs FT (view)
• @mishu-dev will be on vacation next week

[alexandruiulian10]

Weekly 2025-07-31

Participants:
@NilsEb
@Jonas-Kirchhoff
@rtrippel-dfine
@FelixMoelders
@alexandruiulian10
@kalu-an

Agenda & Meeting Minutes:

• nlohmann json fork is available - https://github.com/eclipse-score/inc_nlohmann_json
  • only a development branch is available, we don't have a master branch or the release tags (issue currently raised by Nils)
• test evidence added to the trustable graph
• trudag tool was pitched to SCORE and using it was approved for the trustable pilot - https://github.com/orgs/eclipse-score/discussions/1463#discussioncomment-13943183
• there is a separate test data repository used by the nlohman json; it may either be forked or added as a folder in the future https://github.com/nlohmann/json_test_data
  • we will ask for a fork in order to preserve the history, but will not push it if it is not approved easily
• we can contact Schemmel Alexander for topic related to process or TSF (he is also the safety manager and will do the review)
• josn parser feature request is still ongoing - some new finding from Schemmel Alexander

[mishu-dev]

Weekly 2025-08-07

Participants:
@Jonas-Kirchhoff
@Erikhu1
@mishu-dev
@NilsEb
@alexandruiulian10
@rtrippel-dfine
@FelixMoelders
@kalu-an

Agenda & Meeting Minutes:

• eclipse score nlohmann repository current main branch: https://github.com/eclipse-score/inc_nlohmann_json/tree/json_version_3_12_0
• to be added branch protection
• Trustable graph will be the initial alpha version that will be reviewed and merged into the eclipse score inc_nlohmann_hjson
• current workarea: https://github.com/score-json/json
• Test Data - add a folder or separate branch (observation from Alexander Lanin) : Add JSON test data mirror eclipse-score/.eclipsefdn#64
• if we have issues related to infrastructure best related pull requests you can contact the following slack channel #score-infrastructure-review-requests
• Feature Request: JSON Parser eclipse-score/score#822 , to be discussed with Sankur the finding with custom defined types, please fix
• does a new write component affect us? More expectations and definitions will be needed
• @mishu-dev synch with persistency when they need the qualified write feature
• Alexander will help us with review of the artifacts for qualification
• abort is allowed for unrecoverable error nlohmann json library (confirmation from Alexander) - https://github.com/orgs/eclipse-score/discussions/301
• we need to have a guidance/document/ , we need to push this decision

[mishu-dev]

Weekly 2025-08-21

Participants:
@rtrippel-dfine
@Jonas-Kirchhoff
@Erikhu1
@mishu-dev
@alexandruiulian10

Agenda & Meeting Minutes:

• @Conti check most relevant TSF to continue with (ex provenance)
• we need to prepare the write component request for the json parser: https://github.com/orgs/eclipse-score/discussions/1439
• Bogdan Badicu is the new persistency contact for ECLIPSE SCORE
• clarification exception handling topic (if TRIGGERING abort is legal) - @mishu-dev contact Holger to address this in architecture round
• abort PR https://github.com/eclipse-score/score/pull/1610/files#diff-d9ade5d1259ae8d322db6f148354758f90dda25758f3396c59ca6c49a517e490R73

[mishu-dev]

Weekly 2025-08-28

Participants:
@Erikhu1
@Jonas-Kirchhoff
@rtrippel-dfine
@mishu-dev
@alexandruiulian10

Agenda & Meeting Minutes:

• branch issue with wrong mirror on score nlohmann incubator fixed
• what TENETS are in work . should we create issues for visibility
  • TT-PROVENANCE - conti
  • General topics - d-fine
• md vs rst? If trudag supports only md, should we clean the rst?
  • rst should be cleaned if unneeded , to be checked with committer first (/)
• repo folder structure ex: TSF/trustable/no-json-faults should be moved into a parent container, assertions subfolder?
  • to be checked if this can be reorganized for example ex: TSF/assertions/analysis
• we should use the same component name everywhere JSON-Library vs nlohmann/json , nlohmann-json
  • nlohmann/json should be used in all documentation (/)
• TSF/docs/generated duplicated in TSF/trustable/generated , which one should we keeep , how the score is persisted?
  • TSF/trustable/generated should be removed , and TSF/docs/generated must be excluded - @mishu-dev (/)
• where is the schema for the md (ex reference , evidence structure etc) - references can be used AS EVIDENCE
  • we should set the score , the SME should the check evidence and trustability of instrumentation and can approve it or not (/)

[Jonas-Kirchhoff]

The relevant metadata within an item of the trustable graph are the following:
---
level: 1.1 # automatically generated
normative: true # item is plotted with trudag plot, and used to calculate the score
# normative: false # item is not plotted with trudag plot, and can not be linked to any normative item
references: # list of references
- type: my_custom_reference # custom reference specified in .dotstop_extensions/references.py
input_1: foo # data for constructor of my_custom_reference
evidence: # object for singular evidence, list is illegal; evidence is called with flag --validate
type: my_custom_validator # custom validator specified in .dotstop_extensions/validators.py
configuration: config_for_my_custom_validator
---
Every other metadatum is currently, to the best of my knowledge, ignored, or stored implicitly in .dotstop.dot.

[Jonas-Kirchhoff]

It is possible to move the items of the trustable graph to any location within the project's root folder; in particular, a reorganisation of the TSF/trustable folder is possible. There, it is important to not touch the .dotstop.dot-file manually and to not rename any of the files.

[mishu-dev]

Weekly 2025-09-04

Participants:
@Erikhu1
@rtrippel-dfine
@Jonas-Kirchhoff
@alexandruiulian10
@mishu-dev
@FelixMoelders

Agenda & Meeting Minutes:

• integration into a full system topic:

  • https://sdvworkinggroup.slack.com/archives/C0894QGRZDM/p1756459527374819
  • https://pad.codethink.co.uk/i9_IsuxqRPq7IqVZh90Uow?view
    Tom Mewett [12:25 PM](https://sdvworkinggroup.slack.com/archives/C0894QGRZDM/p1756459527374819) S-CORE integration problem statement & solutions Hey all - I wrote a summary of the problem we are having with S-CORE integration (with examples), and Codethink's suggestion of solutions. https://pad.codethink.co.uk/i9_IsuxqRPq7IqVZh90Uow?view S-CORE currently does not support integration into a full system. Its build processes are not designed with this in mind. System integrators will want to build S-CORE components with as many dependencies as possible (e.g. libraries and compiler) coming from the system. They will use their choice of meta-build/integration tool to build the system piece-by-piece inside a sandbox. This is standard practice when building a Linux system or distro. However, many features of Bazel are designed so it can do its own integration, such as downloading dependencies, defining specific toolchains, etc. S-CORE is using these features, which make the build processes unsuitable for integration. See Appendix: Examples of S-CORE components retrieving their own dependencies below. [...] Let me know if you have any questions (edited)

• score repo cleanup:

  • https://github.com/eclipse-score/inc_json - should be deleted - https://github.com/orgs/eclipse-score/discussions/1239#discussioncomment-14294653
  • https://github.com/eclipse-score/inc_nlohmann_json - this is the official one

• TSF Cleanup: https://github.com/score-json should be replaced with the official one once the TSF changes are first merged into Eclipse score fork - create issue?

• 05.09.2025 first look with Alexander Schemmel on current TSF instrumentation

• additional repos for the test data + documentation - the final decision is - 1 repo for source and documentation + 1 separate for test data

[mishu-dev]

Weekly 2025-09-11

Participants:
@rtrippel-dfine
@Jonas-Kirchhoff
@Erikhu1
@alexandruiulian10
@mishu-dev

Agenda & Meeting Minutes:

• Currently we are covering only read ops in nlohmann/json
• write ops will be documented incrementally in the future: write pull request: https://github.com/eclipse-score/score/pull/1711
• @mishu-dev what is the read/write split on nlohmann library at nlohmann / eclipse score level?
• @mishu-dev TSF read only for first step qualification - check for decision in communication/baselibs meeting minutes !!!!
• at least 2 days for estimations for TSF serialization operations
• in case that write /serialize is fulfilled 100 % in json library glue code , should also ensure that related calls from nlohmann json should be 0
• check with Alexander and Markus , the impact of read only qualification in TSF

[Erikhu1]

Weekly 2025-09-18

Participants:
@Jonas-Kirchhoff
@Erikhu1
@rtrippel-dfine
@alexandruiulian10

Agenda & Meeting Minutes:

• Write functionality needs to be certified but timeline for this has not been decided. Needs to be aligned on what actually is requested (New RFC8259, I-JSON format RFC7493, old JSON RFC7159, Binary JSON, etc).

• In glue code there are currently two parsers, called JSON and I-JSON:

  • Clarify whether it is intended to use I-JSON
  • Glue code needs to be adapted if usage of I-JSON is intended
  • Clarify that nlohmann/json does not satisfy I-JSON (one example is key-duplicates)

• Feedback from first review round of TSF documentation has been addressed. Joint review session planned with Markus Schu and Alexander Schemmel on 22/9.

• Rough estimate for documenting write functionality (serialization operations): 5-15 days.

• Docs-as-code insights to be presented and aligned with Alexander Lanin next week.

[Erikhu1]

Correction: The "I_JSON" in the glue code is for "interface JSON" and not "internet json". As such, point 2 can be ignored. Nevertheless, it should be clarified that we only follow RFC 8259, and that all other RFCs are explicitly out of scope (in particular 'RFC 8949 - binary JSON' and 'RFC 7493 - Internet JSON')

[mishu-dev]

Weekly 2025-09-25

Participants:
@Jonas-Kirchhoff
@Erikhu1
@mishu-dev
@alexandruiulian10

Agenda & Meeting Minutes:

• it was confirmed that we should focus only on read operation for the first TSF qualify iteration. We should handle the (uncovered) write part afterwards separately - clarified in baselibs meeting
• we should stay requirements (ex: RFC8259) - clarified in baselibs meeting
• related to requirement separation
  • we need to keep nlohmann requirements the same
  • Alexander Schammel said he will analyze how we can proceed with these requirements
• feedback from 2nd review round in progress.
• Docs-as-code - we should stay with trudag and integrate into bazel - @mishu-dev try to find a way to automate trudag tooling instrumentation

[Jonas-Kirchhoff]

Weekly 2025-10-02

Participants:
@Jonas-Kirchhoff
@alexandruiulian10

Agenda & Meeting Minutes:

• regarding splitting of feature request: Alexander Schemmel and Markus Schu are to be consulted on suggestion
• status of TSF: awaiting review from Alexander Schemmel and Markus Schu
• request to move to within S-CORE (cf. https://github.com/orgs/eclipse-score/discussions/1239#discussioncomment-14561292) under investigations, possible hindrances: custom pipelines, speed of development

[mishu-dev]

Weekly 2025-10-09

Participants:
@Jonas-Kirchhoff
@alexandruiulian10
@Erikhu1
@rtrippel-dfine
@mishu-dev

Agenda & Meeting Minutes:

• regarding splitting of component request - Alexander Schemmel will take care of it , review requested - Split JSON-Library in two components eclipse-score/score#1846
• validation requirement - to be checked and stakeholders notified @mishu-dev
• status of TSF: Alexander Schemmel has issues with Trudag runtime environment - to be notified when valid solution is available
• request to move to within S-CORE (cf. https://github.com/orgs/eclipse-score/discussions/1239#discussioncomment-14561292) under investigations, possible hindrances: custom pipelines, speed of development - committer rights should be discussed for d-fine colleagues so the integration process will not be blocked by lack of committer approval

[alexandruiulian10]

Weekly 2025-10-16

Participants:
@Jonas-Kirchhoff
@alexandruiulian10
@Erikhu1

Agenda & Meeting Minutes:

• validation requirement still awaiting response from persistence and logging
• truedag related issues of Alexander Schemmel were fixed; awaiting confirmation
• d-fine is preparing the migration of json repo in eclipse score - https://github.com/eclipse-score/inc_nlohmann_json

[mishu-dev]

Weekly 2025-10-23

Participants:

@Erikhu1
@LucaFgr
@Jonas-Kirchhoff
@rtrippel-dfine
@mishu-dev
@alexandruiulian10

Agenda & Meeting Minutes:

• Validation target to return explicitly procedural return values instead of abort (to be checked): https://github.com/orgs/eclipse-score/discussions/1239#discussioncomment-14750150
  • nlohmann/json check can be used in the score glue code
• abort / ( general exception handling topic ) raised by persistence should be clarified by them in general architecture round
• TSF migration : successfully tested , Andrey to be contacted with support for migration
• Let's keep this meeting and please provide if there are topic for the current week till Thursday morning. If not we will cancel the occurrence