December 2022 Production (PUBLISHED) vs June 2023 PreProduction (Version 4 - COMPLETE RE-START OF SWISS RELEASE, AS IF THIS IS A "FIRST TIME" RELEASE) traceability
SUMMARY COMPONENT REPORT: https://docs.google.com/spreadsheets/d/1Q18jq4eHU4-H8hou3U3Pb-4l1jGgqRNSIk-1KvTfrIo/edit#gid=1
MSSP TICKET: https://jira.ihtsdotools.org/browse/MSSP-2072
Differences found in package Comparison | Number of RF2 records impacted | Related JIRA ticket(s) | Rationale |
|---|---|---|---|
| RefsetDescriptor files | 10 | 8 records had their UUID's refreshed + new effectiveTime of 20230607 (but NO CHANGES TO ACTUAL RECORDS) as expected, due to NEW "FIRST TIME" RELEASE:
2 records however, DID NOT GET REPLACED IN NEW JUNE 2023 RELEASE as expected???
| |
| ModuleDependency files | 5 | 5 records removed completey, replaced with nothing!! SO ALL WRONG!! NEED TO REVERT BACK TO EXTERNALLYMAINTAINED INPUT FOR THIS FILE TO SEE IF THIS FIXES IT!! | |
| Readme file | all x prefixes added in, plus PreProduction naming convention applied, as expected. | ||
| Association files | 12 | 2 new records added/updated for 20230607, 10 records (from all different effectiveTimes) removed ...as expected ???????????????? due to exact match in Summary Component Stats report for this cycle (see link above) | |
| AttributeValue files | 11613 | 6860 new records added/updated for 20230607, 11613 records (from all different effectiveTimes) removed ...as expected ???????????????? 10753 records have been added here for the Common French descriptions (11000241103) - question is why?!! The snapshot of this file had only 20 records in it for the December 2021 release! However when I run the June 2022 release, it now has over 10000 records! (there’s only been 200 odd Description changes in this cycle!) The attributeValue config in the manifest is still the same as it was in December 2021: <file Name=“xder2_cRefset_AttributeValueSnapshot_CH1000195_20220607.txt”> To be fair, looking at the Common French Description file (https://prod-release.ihtsdotools.org/api/centers/ch/products/snomed_ct_switzerland_releases[…]s/xsct2_Description_Snapshot-fr_CH1000195_20220607.txt), it looks like there were about 10254 records inactivated in the December release - however these were Born Inactive (in the first time release) and so I wouldn’t expect the system to create new attributeValue records for these? But perhaps it has to if it can’t find existing attributeValue records for the old inactive records?? CHECKING WITH QUYEN...Quyen thinks it's a snowstorm bug, so I've raised ticket: ISRS-1254 - Getting issue details... STATUS Michael confirmed them to be all valid in ISRS-1254 | |
| Language Refset (de-ch) files | 29491 | 27928 new records added/updated for 20230607 - ***** FOR REFSETID: 2041000195100 ***** 1563 records (from all different effectiveTimes) removed ..NOT as expected due to Summary Component Stats report for this cycle (see link above) showing:
IS THIS TO BE EXPECTED BECAUSE WE'VE "STARTED FROM SCRATCH" IN THIS RELEASE?????? DOES THIS MEAN THE SCS REPORT WILL BE INVALID, AS WILL SHOW NOW INACTIVATIONS BECAUSE IT'S A "BRAND NEW RELEASE" WITH NO "PREVIOUSPUBLISHEDPACKAGE" TO COMPARE THE CONTENT TO???? | |
| Language Refset (it-ch) files | 2603 | 1292 new records added/updated for 20230607 - ***** FOR REFSETID: 2031000195108 ***** 1311 records (from all different effectiveTimes) removed ...as expected ???????????????? due to exact match in Summary Component Stats report for this cycle (see link above) | |
| Language Refset (en) files | 182 | 96 new records added/updated for 20230607 - ***** FOR REFSETID: 900000000000508004 + 900000000000509007 ***** 86 records (from all different effectiveTimes) removed ...as expected ???????????????? | |
| Language Refset (fr) files | 0 | 0 records added/updated, as expected as this is Common French Language file?????? (but then shouldn't we be moving some into fr-ch, or were these already in December - no they weren't there, they requested no preferences for Common French, just Swiss French) SUMMARY REPORT EXPECTING 11 - this looks like a problem with the Summary Component Report - possibly just because we're comparing to a DIFFERENT previous published package though, because we've generated the new one?!! | |
| Language Refset (fr-ch) files | 260,288 | 161,454 new records added/updated for 20230607 - ***** FOR REFSETID: 2021000195106 ***** 98,834 records (from all different effectiveTimes) removed ...as expected ???????????????? due to exact match in Summary Component Stats report for this cycle (see link above) | |
| Concept files | 46 | 25 new records added/updated for 20230607, 21 records (from all different effectiveTimes) removed ...NOT as expected due to Summary Component Stats report for this cycle (see link above) showing:
IS THIS TO BE EXPECTED BECAUSE WE'VE "STARTED FROM SCRATCH" IN THIS RELEASE?????? DOES THIS MEAN THE SCS REPORT WILL BE INVALID, AS WILL SHOW NOW INACTIVATIONS BECAUSE IT'S A "BRAND NEW RELEASE" WITH NO "PREVIOUSPUBLISHEDPACKAGE" TO COMPARE THE CONTENT TO???? Some of the records removed appear to be absolutely valid, as they are previously inactive records (and therefore there's no need to keep them as "born inactive" records) (eg)
Many more appear to be valid as they're simply "removing" these records to be replaced with the same record just with new effectiveTimes (eg)
Yet more appear to be added in validly as brand new concepts for this authoring cycle (eg)
However, that leaves 1 active record that has just been removed for no apparent reason:
DID RORY PERHAPS INTENTIONALLY MISS OUT ALL CONTENT FOR THIS MODULE (WHEN IMPORTING THE CONTENT INTO THE FRESH NEW CODESYSTEM), BECAUSE THIS IS THE COMMON FRENCH MODULE (11000241103 | SNOMED CT Common French translation module (core metadata concept) |) AND THEREFORE WE'RE NOW GOING FOR AN APPROACH THAT JUST IMPORTS THE COMMON FRENCH CONTENT STRAIGHT INTO THE SWISS MODULE INSTEAD AND THEREFORE THERE IS NO LONGER A DEPENDENCY THERE ON THE COMMON FRENCH MODULE, AND THUS NO REASON TO INCLUDE IT IN THE SWISS PACKAGE???? | |
| Description (de-ch) files | 29596 | 28030 new records added/updated for 20230607 1566 records (from all different effectiveTimes) removed ...as expected ???????????????? due to close match with TOTAL DESCRIPTION Counts in Summary Component Stats report for this cycle (see link above) | |
| Description (en) files | 68 | 40 new records added/updated for 20230607 - 28 records (from all different effectiveTimes) removed ...as expected ???????????????? due to close match with TOTAL DESCRIPTION Counts in Summary Component Stats report for this cycle (see link above) | |
| Description (fr) files | 185,262 | 7479 new records added/updated for 20230607 - 177,783 records (from all different effectiveTimes) removed ...as expected ???????????????? as planned in INFRA-8967 in order to correctly split out the Swiss French records into their own "fr-ch" Description file, leaving ONLY the Common French records in this "fr" Description file. ...as expected due to close match with TOTAL DESCRIPTION Counts in Summary Component Stats report for this cycle (see link above) | |
| Description (fr-ch) files | 158774 | 158,623 new records added/updated for 20230607 - 151 records (from all different effectiveTimes) removed ...as expected ???????????????? ...as expected due to close match with TOTAL DESCRIPTION Counts in Summary Component Stats report for this cycle (see link above) | |
| Description (it-ch) files | 2637 | 1326 new records added/updated for 20230607 - 1311 records (from all different effectiveTimes) removed ...as expected ???????????????? due to close match with TOTAL DESCRIPTION Counts in Summary Component Stats report for this cycle (see link above) | |
| Relationship files | 64 | 37 new records added/updated for 20230607 - 27 records (from all different effectiveTimes) removed ...NOT as expected due to Summary Component Stats report for this cycle (see link above) showing:
IS THIS TO BE EXPECTED BECAUSE WE'VE "STARTED FROM SCRATCH" IN THIS RELEASE?????? Some of the records removed appear to be absolutely valid, as they are previously inactive records (and therefore there's no need to keep them as "born inactive" records) (eg)
Many more appear to be valid as they're simply "removing" these records to be replaced with the same record just with new effectiveTimes (eg)
Yet more appear to be added in validly as brand new concepts for this authoring cycle (eg)
The good news here is that the only record related to the now unnecessary Common French module (11000241103) was already inactive, and therefore gets removed anyway:
| |
| Relationship Concrete Values files | 0 | 0 records added/updated ...as expected due to exact match in Summary Component Stats report for this cycle (see link above) | |
| OWL Expression files | 46 | 25 new records added/updated for 20230607 - 21 records (from all different effectiveTimes) removed ...as expected ???????????????? NOT AS EXPECTED - THE SUMMARY COMPONENT REPORT (linked) SHOWS 2 new and 1 inactivated expected????) Maybe because it's comparing to a DIFFERENT previous published package though, because we've generated the new one?!! Yes, manual checks show 1 inactivations + 1 addition since previous publication... | |
| TextDefinition (de-ch) files | 0 | 0 new records added/updated for 20230607 - 0 records (from all different effectiveTimes) removed ...as expected ???????????????? | |
| TextDefinition (en) files | 0 | 0 new records added/updated for 20230607 - 0 records (from all different effectiveTimes) removed ...as expected ???????????????? | |
| TextDefinition (fr) files | 1 | 0 new records added/updated for 20230607 - 1 records (from all different effectiveTimes) removed ...as expected ???????????????? | |
| TextDefinition (fr-ch) files | 0 | 0 new records added/updated for 20230607 - 0 records (from all different effectiveTimes) removed ...as expected ???????????????? | |
| TextDefinition (it-ch) files | 0 | 0 new records added/updated for 20230607 - 0 records (from all different effectiveTimes) removed ...as expected ???????????????? |