Should the commit author of a squashed commit be me, or the only author of the squashed commits?

[Byron]

When squashing copilot commits (and there are always at least two, one is empty), it puts my identification as author by default.
Then I have to copy the copilot author string and change it, which is a hassle by now.

I could use the --author argument when invoking the squash, but would prefer if I could make it use the dominant author, or only author automatically, either by adding a new flag, or by default.

Is this something you could imagine, or am I missing something?
Thank you

[jpgrayson]

This is a good question, @Byron. Thanks for bringing it up.

Currently the author of the squashed patch will revert to the repo's configured author unless all of the patches being squashed, including the base patch, have some the same author.

So when squashing pairs of copilot patches, I would expect the squashed patch to retain the copilot authorship.

But a quick experiment shows me that this isn't happening; the author is resetting to the repo's configured author as you are seeing.

So I think there's a bug.

make it use the dominant author

That's an interesting idea too. I'm interpreting it to mean the author most represented in the patches being squashed. I haven't formed a strong opinion about whether that's a good idea though.

[Byron]

It's rare to say this, but it's great to hear that this is a bug :D.

This means it will get squashed eventually and it will respond exactly as I would have expected.

When the author isn't uniquely decidable as there are multiple ones, I think the only unambiguous solution is to put in the configured author, but auto-add Co-authored-by trailer lines to mention every original author. We probably wouldn't want to casually discard original authorship like I originally mentioned, without thinking much obviously 😅.