Skip to content

docs(issues): fold in the Egeria team's response and our corrections (2026-10-05) - #433

Merged
dwolfson merged 2 commits into
odpi:mainfrom
dwolfson:docs/egeria-team-response-2026-10-05
Oct 6, 2026
Merged

dwolfson merged 2 commits into
odpi:mainfrom
dwolfson:docs/egeria-team-response-2026-10-05

Conversation

@dwolfson

@dwolfson dwolfson commented Oct 5, 2026

Copy link
Copy Markdown
Member

Summary

Records the Egeria team's 2026-10-05 response to the open server-bug list in PYEGERIA_ISSUES.md, plus three corrections to our own entries. Docs-only.

Caveat on the response itself: the Egeria team's fixes are in two PRs on oak2026 that are not merged to main, so no build contains them, and they tested nothing against a live platform.

Corrections to our own entries (each verified in source or history):

  • ISSUE-125: the cause is an argument-order bug at OMRSRepositoryContentValidator.java:877 (getAttributeTypeDef(sourceName, guidParameterName, guid, ...) into an overload expecting (guid, name)), not a GUID map. Our earlier hypothesis was wrong and is replaced.
  • ISSUE-89: rsa.key-id does not stop the signing key being regenerated (RSAGenerator always makes a new key pair; the id is only a label). The original claim was wrong.
  • ISSUE-79: probably a pyegeria bug. TemplateRequestBody.deep_copy defaulted to False from 2026-01-09 until fix(sdk): leave deepCopy, isOwnAnchor, startsWith and ignoreCase to Egeria's defaults #399 (2026-09-29, first released in v6.1.23); the model is serialised with exclude_none, which keeps False; and create_folder_element_from_template builds its body without deepCopy. That yields an element with no ResourceConnection, i.e. a null assetConnector. Still to be confirmed live.

Responses recorded: 90, 112, 117, 124, 125 valid with fixes pending merge (124 also needs updateTime); 102 and 108 not reproduced by the Egeria team (with our answers to what they asked for).

Test plan

  • n/a (docs only). Corrections checked against Egeria source at 449ad06894 and pyegeria's git history.

🤖 Generated with Claude Code

dwolfson and others added 2 commits October 5, 2026 13:50
…(2026-10-05)

The Egeria team answered the open server-bug list (source oak2026 fb3d6fce53; fixes are in two unmerged PRs, so
no build has them, and nothing was tested live). Recorded per entry: 90, 112, 117, 124, 125 valid with fixes pending
merge; 102 and 108 not reproduced by them; 95 and 85 confirmed.

Corrections to our own entries: ISSUE-125's cause was an argument-order bug at OMRSRepositoryContentValidator:877,
not a GUID map (our hypothesis was wrong, confirmed in source); ISSUE-89's claim that rsa.key-id stops key
regeneration was wrong (RSAGenerator always generates a new pair); ISSUE-79 is probably a pyegeria bug, since
deep_copy defaulted to False until odpi#399 (2026-09-29) and the folder helper never set it.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Signed-off-by: Dan Wolfson <dan.wolfson@pdr-associates.com>
… pending

The entry still said 'fixed on branch ..., pending PR/merge'; that branch merged in odpi#421 and shipped in 6.1.28 (and
6.1.29). Re-verified read-only on the 2026-10-05 rebuilt platform: get_catalog_targets returns 'No elements found'
in all three formats with no server error.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Signed-off-by: Dan Wolfson <dan.wolfson@pdr-associates.com>
@dwolfson
dwolfson merged commit ac82678 into odpi:main Oct 6, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant