At Wed, 16 Mar 2011 12:17:59 +0000, Peter Rosin wrote:

Hi Takashi,

This is not really important to /me/, on /this/ occation. But I thought I'd ask anyway as it seems strange that the author tag ended up wrong. If it is a symptom of more errors, others may care more. Personally I'm mainly wondering how it could happen, and would like for it to not happen again (especially if I ever send more substantial patches).

The commit e2e9566230 ALSA: AT73C213: Rectify misleading comment. was first sent as a git patch [1] and then again amended with a pair of other (seemingly) trivial comment fixes [2] (the amended fixes were not commited for some reason).

Both patches were sent from my axentia account and both carried signed-off-by tags using the axentia account. Yet the commit is made naming my lysator account as author (the sign-off is still intact).

The patch that seems to have ended up being commited is [1], and what I think has happened is that I used the lysator account in the discussion following the 2nd patch, but I fail to understand how the sender of that one message (or was it a few?) could "leak" into the author tag.

I'm assuming it was just a silly mistake, I'm mearly pointing out the mishap and understand that there is little one can do to fix this buglet this late.

I usually apply patches with git-am, and it takes the original From tag unless explicitly specified in addition in the patch. So was this case, I suppose.

Though, it seems that I overlooked a warning from my check script, which checks the consistency between author and sign-off.

thanks,

Takashi