Commits page displays unclearly time value

[ZacharyBear]

Gitea Version

1.24.3

What happened?

In the commits page, it list all commits and its pushedDate, I think displays the commit time will be better.

1. From my situation, there're commits published in different time, but they were pushed at the same time. The commits page shows their push-time, not commit time, so they're all "3 Hours" ago.

2. When I click and open each commit, the commit detail page only shows commit time: "Yesterday".

In summary, the commits and commit detail page are showing different time, this could make a mental burden.

How are you running Gitea?

Privacy deployed.

[puni9869]

Checking ... allow me some time on it.

[wxiaoguang]

There are two kinds of times in git commit: "commit time" and "author time"

Try to grep "Author.When" and "Committer.When" through the code base, sometimes they are not used clearly or consistently.

---

gitea/templates/repo/commits_list.tmpl

Lines 64 to 68 in 4a6db5a

	{{if .Committer}}
	<td class="tw-text-right">{{DateUtils.TimeSince .Committer.When}}</td>
	{{else}}
	<td class="tw-text-right">{{DateUtils.TimeSince .Author.When}}</td>
	{{end}}

---

gitea/templates/repo/commit_page.tmpl

Line 143 in 4a6db5a

<span class="tw-text-text-light">{{DateUtils.TimeSince .Commit.Author.When}}</span>

[silverwind]

Some alignment with GitHub on when to use which date may be good. Here's an AI-generated answer on where they show each date, so must be verified before:

[wxiaoguang]

I believe we should always use "committer date" for all these cases.

For most cases, "author date" is the same as "committer date"

If a committer changes the commit, they might have made huge changes by amending, "author date" isn't really needed in such cases, "committer date" isn't worse than "author date".

---

To keep things simple: it's fine to always use "committer date" , most git commands also use "committer date" by default.

[silverwind]

Yes I think commit date is generally what users expect. Author date for contribution-related counting might make sense, but in gitea this is currently push date, which is even more wrong.

[puni9869]

I have started working on this.