Skip to content

knowledge(data-modeling): distinguish Modify from primary-key Rename - #274

Open
Jesper Schulz-Wedde (JesperSchulz) wants to merge 1 commit into
mainfrom
jesperschulz-primary-key-aggregation-knowledge
Open

Jesper Schulz-Wedde (JesperSchulz) wants to merge 1 commit into
mainfrom
jesperschulz-primary-key-aggregation-knowledge

Conversation

@JesperSchulz

Copy link
Copy Markdown
Contributor

Summary

Add one focused data-modeling article explaining why changing a retrieved primary-key field and calling Modify is not a record rename, without misreporting legitimate aggregation in a different temporary table.

  • Resolve the actual record type and key before diagnosing the write.
  • Explain the restricted Rename alternative: retain original identity, supply the complete new key, check uniqueness and record freshness, and do not assume OnModify runs.
  • Recognize non-key quantity accumulation followed by one destination insert per group. Merely marking the original table's record temporary does not change its key.
  • Distinguish record-value changes from a published table's key-schema changes.
  • Add original self-contained good/bad companions and select them in the existing evaluation manifest.

Why this belongs in BCQuality

A reviewer can miss a key-changing update or incorrectly carry the same finding forward after the receiver changes to a buffer whose key excludes the accumulated field. That requires interpreting record identity and the full data path, not repeating a compiler/analyzer diagnostic. data-modeling owns record/key consistency; existing published-schema and performance-aggregation articles address different concerns. No skill or protocol changes are needed.

Evidence and applicability

The article cites Microsoft Learn for Rename, Modify, primary keys, temporary tables, TableType, and runtime mapping.

Core guidance applies to all BC versions; TableType = Temporary is explicitly limited to runtime 6.0 / BC 17+. The illustrative samples use a temporary record variable and contain no copied application source.

Motivated by the public microsoft/BCApps#14006 discussion and static inspection of its subsequent buffer correction. This contribution does not claim a new runtime reproduction or that a green CI result proves execution of a particular regression.

Validation

  • Frontmatter/content validator: 0 errors, 0 warnings.
  • Fixture validation, changed-pair coverage, and neutral preparation: passed, 315 cases.
  • New article present in both prepared case candidate lists.
  • Findings-report contract checks and diff whitespace check passed.
  • AL samples compiled or executed: not performed; demonstration-only corpus.
  • Model-based scoring: not performed.

Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
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