search

cityofaustin / atd-vz-data

The technology that powers the City of Austin's Vision Zero program
https://visionzero.austin.gov/viewer/
11 stars 2 forks source link

Data model: crash details: change log #1449

Closed johnclary closed 2 months ago

johnclary commented 3 months ago

Associated issues

This is a bare bones change log for the crash details page.

Testing

URL to test: Local

I guess the main thing to test is that we are are (1) capturing all changes as expected and (2) make sure that this change log is usable and intuitive enough for MVP purposes.

  1. Start the data model stack and run the CRIS import script as described in the collector branch PR.
  2. Navigate to any crash details page and scroll down to find the Record history card. This is the change log. Is the name Record history ok? Do you prefer Changes or Change log?
  3. Click on a row of type create in the Record history table. Verify that a modal appears which displays a table with the record's values.
  4. Try editing a unit and person record, as well as various crash fields. Each time you make an edit, the record history table will update and capture your changes. If the change you made affects a calculated value, e.g., person injury severity, the changes to calculated fields will also be captured in the change log.

Ship list

roseeichelmann commented 2 months ago

One thing thats interesting is how when i updated the injury sev for a person, the change log shows that i updated the est comp cost as well. I guess that kinda makes sense bc they are related and that is a generated field that changes on injury sev update but also its not like i directly updated it so it also was unexpected to see it there? image

edit: just saw you already pointed that out in the pr description and its obviously purposeful. i wonder if theres a way we can make it clear that it wasnt directly user edited? either way this still works perfectly as is for an MVP!

chiaberry commented 2 months ago

I also didnt expect to see the comp cost change at first (then I realized John warned us in the test steps), but I do really like seeing the changes. A way in the future enhancement could be some signaling that connects the calculated value to the field that incited the change.

johnclary commented 2 months ago

Thanks y'all for the feedback! a few changes I made just now:

I will open a follow-up issue to: