You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I recently updated my homebox instance with the latest changes and since the update, maintenance list is completely broken: filters do not work anymore as expected. When selecting "Completed", it show me some entries which are only scheduled (without any completion date).
I first thought it was related to #367 but after reverting this PR from my fork, the issue is still there.
I had a look directly in the database and I have a mix of different date formats (with and without timezone):
There was also a change regarding date handling (&_time_format=sqlite added everywhere).
I'm still digging to find the change responsible for the bug.
How can the maintainer reproduce the issue?
For now, the way to reproduce the issue is to upgrade a running instance to a newer version.
Deployment
Docker (Linux)
OS Architechture
x86_64 (AMD, Intel)
Deployment Details
No response
The text was updated successfully, but these errors were encountered:
I found the commit responsible for the database issue: 1a4e98a. (#430)
I'm still investigating to check if this is responsible for all the other bugs or not.
EDIT: after reverting this commit in my fork, all issues are gone. Maybe PR #430 is totally fine on a new fresh database, but for sure a migration script would be required for already running instances.
First Check
Homebox Version
main (a1dabaa)
What is the issue you are experiencing?
I recently updated my homebox instance with the latest changes and since the update, maintenance list is completely broken: filters do not work anymore as expected. When selecting "Completed", it show me some entries which are only scheduled (without any completion date).
I first thought it was related to #367 but after reverting this PR from my fork, the issue is still there.
I had a look directly in the database and I have a mix of different date formats (with and without timezone):
There was also a change regarding date handling (
&_time_format=sqlite
added everywhere).I'm still digging to find the change responsible for the bug.
How can the maintainer reproduce the issue?
For now, the way to reproduce the issue is to upgrade a running instance to a newer version.
Deployment
Docker (Linux)
OS Architechture
x86_64 (AMD, Intel)
Deployment Details
No response
The text was updated successfully, but these errors were encountered: