View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0005803 | CaseTalk Modeler | Validation | public | 2026-07-25 22:22 | 2026-07-27 14:44 |
| Reporter | Marco Wobben | Assigned To | Marco Wobben | ||
| Priority | normal | Severity | feature | Reproducibility | have not tried |
| Status | resolved | Resolution | fixed | ||
| Target Version | 15.x | Fixed in Version | 15.0 | ||
| Summary | 0005803: Deciding a uniqueness constraint does not ask whether the same fact can recur over time | ||||
| Description | When a uniqueness constraint is settled for a fact type, the question whether the same participants can be involved again at a later time is left implicit. Two people can marry, divorce and marry again; a treatment can be stopped and started again; an employee can be hired twice by the same employer. In other cases a single occurrence is exactly what is meant. This cannot be derived from the model: both readings are legitimate and look identical. It is a decision the modeller has to make, and it belongs with the temporal nature of the type, where periods of validity are already recorded. Expected: while deciding the uniqueness constraint in the wizard, the modeller is asked about this, the answer is kept as the temporal setting of the type, and documentation and generated output reflect it. Where nothing has been declared yet, the model quality report can point that out, so the question is answerable and can be settled rather than repeated. | ||||
| Tags | No tags attached. | ||||
| CaseTalk Edition | Corporate | ||||
| related to | 0005802 | resolved | Objectified fact types with one participant are classified as kind instead of an aspect | |
| related to | 0005809 | resolved | Temporal setting cannot record a decision moment without also declaring an effective period | |
| related to | 0005810 | resolved | Marco Wobben | UC Wizard asks the same questions again for each language of a fact type |
|
This requires some sort of temporal wizard, in which questions are asked about facts over time. A marriage has a historical timeline; an article has one price but can be adjusted over time and pro-actively into the future; a fact can be known before it is registered; .. |
|
|
Wizard flow. The UC wizard (FORMS/_FUCWIZ.PAS) is a Yes/No state machine over candidate UCs (sValidateSmallUC -> DisplayUC -> confirm, with mStatement and mQuestion), so the follow-up is a new state after a UC is accepted rather than a new dialog. Q1, existing: can these facts occur at the same time? -> settles the uniqueness constraint. |
|
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2026-07-25 22:22 | Marco Wobben | New Issue | |
| 2026-07-25 22:22 | Marco Wobben | Relationship added | related to 0005802 |
| 2026-07-26 10:07 | Marco Wobben | Note Added: 0008618 | |
| 2026-07-26 10:11 | Marco Wobben | Summary | Warn when an objectified relationship cannot occur twice between the same participants => Objectifying a fact type does not ask whether the relationship can recur over time |
| 2026-07-26 10:11 | Marco Wobben | Description Updated | |
| 2026-07-26 11:20 | Marco Wobben | Note Added: 0008626 | |
| 2026-07-27 09:20 | Marco Wobben | Relationship added | related to 0005809 |
| 2026-07-27 10:26 | Marco Wobben | Summary | Objectifying a fact type does not ask whether the relationship can recur over time => Deciding a uniqueness constraint does not ask whether the same fact can recur over time |
| 2026-07-27 10:26 | Marco Wobben | Description Updated | |
| 2026-07-27 11:09 | Marco Wobben | Note Edited: 0008626 | |
| 2026-07-27 11:09 | Marco Wobben | Note View State: 0008626: public | |
| 2026-07-27 13:22 | Marco Wobben | Relationship added | related to 0005810 |
| 2026-07-27 14:44 | Marco Wobben | Assigned To | => Marco Wobben |
| 2026-07-27 14:44 | Marco Wobben | Status | new => resolved |
| 2026-07-27 14:44 | Marco Wobben | Resolution | open => fixed |
| 2026-07-27 14:44 | Marco Wobben | Fixed in Version | => 15.0 |