404 when trying to stop the timer which has been started before binary upgrade

[bagasme]

Description

I was curious about using timer/stopwatch feature of Gitea. I enabled the timer when the deployed binary version was about 7ae10cb.

Today I build Gitea from version e460b6e and deployed it. When I tried to stop the timer,
Gitea returned 404 on https://%(DOMAIN)s/times/stopwatch/toggle, thus the timer cannot be stopped.

I guessed that commit 90cf07a touched the subsystem in question.

Gitea Version

main (commit e460b6e)

Can you reproduce the bug on the Gitea demo site?

No

Log Gist

https://gist.github.com/bagasme/335d722c4f87f552176b352d2e466f84

Screenshots

Git Version

2.39.1

Operating System

Ubuntu 22.04

How are you running Gitea?

I built Gitea from source using Go 1.20, Node v18.13.0 (installed via nvm) on Debian 11.

The Gitea instance is hosted on a LXD container on my computer.

Database

PostgreSQL

[wxiaoguang]

The generated stopwatch URL is not ideal, if there is no ActiveStopWatch then it outputs incorrect URL

(Hmm, it seems not related to your problem directly, so just FYI)

---

I just played a while, I think it could be just a display bug? Maybe you do not have active stopwatch yet ..... The display bug is confirmed and is still being fixed .....

Could you help to confirm how your problem happens, especially the screenshots? Is there an active stopwatch?

[bagasme]

Could you help to confirm how your problem happens, especially the screenshots? Is there an active stopwatch?

My Gitea instance have implicit ENABLE_TIMETRACKING=true (as the config item isn't specified in app.ini). When it is set to false, the timer is paused; but when I set it again to true, the timer resumes.

Anyway, how can I check if there is active stopwatches?

[wxiaoguang]

Screenshots?

ps: the display bug of #22169 problem would be fixed by #22879 -- if 22169 is the cause.

[wxiaoguang]

Please have a try to see whether 22879 resolve the "display bug" on your side (I guess).

[wxiaoguang]

I would assume that the problem has been resolved?

[wxiaoguang]

I would assume that the problem has been resolved?

Hmm ... maybe there is no more problem. I guess the problem was the UI regression caused by #22169 , it should have been fixed by #22879