On Thu, May 21, 2026 at 7:41 AM Akihiko Odaki
<[email protected]> wrote:
>
> On 2026/05/21 4:37, Philippe Mathieu-Daudé wrote:
> > On 19/10/22 14:44, Markus Armbruster wrote:
> >> Philippe Mathieu-Daudé <[email protected]> writes:
> >>
> >>> The amsat.org domain is having issues with DMARC / SPF / DKIM:
> >>> https://lore.kernel.org/qemu-devel/
> >>> CAMVc7JUy5NeEN0q=4zfZvn_rppgqn9wicV1z=tsluhks3ry...@mail.gmail.com/
> >>>
> >>> Consolidate all of my MAINTAINERS entries on my work address.
> >>>
> >>> Signed-off-by: Philippe Mathieu-Daudé <[email protected]>
> >>> ---
> >>>   .mailmap    |  4 +++-
> >>>   MAINTAINERS | 62 ++++++++++++++++++++++++++---------------------------
> >>>   2 files changed, 34 insertions(+), 32 deletions(-)
> >>>
> >>> diff --git a/.mailmap b/.mailmap
> >>> index 1f7319b70b..35dddbe27b 100644
> >>> --- a/.mailmap
> >>> +++ b/.mailmap
> >>> @@ -70,7 +70,9 @@ Paul Burton <[email protected]>
> >>> <[email protected]>
> >>>   Paul Burton <[email protected]> <[email protected]>
> >>>   Paul Burton <[email protected]> <[email protected]>
> >>>   Paul Burton <[email protected]> <[email protected]>
> >>> -Philippe Mathieu-Daudé <[email protected]> <[email protected]>
> >>> +Philippe Mathieu-Daudé <[email protected]> <[email protected]>
> >>> +Philippe Mathieu-Daudé <[email protected]> <[email protected]>
> >>
> >> I think this retroactively credits all your contributions to Linaro.
> >> Intentional?
> >
> > Certainly not intentional. I was seeing this as a git limitation, but
> > Alex pointed to me we are doing an invalid use of git-mailmap, since
> > it is described as:
> >
> >    ... the file .mailmap [...] is used to map author and committer
> >    names and email addresses to canonical real names and email addresses.
> >
> > so more for typos, not developers changing email addresses. So our
> > "replace old addresses by a more recent one" section might not be
> > the best use. I think it was useful for the get_maintainer.pl fallback
> > option: when a file is not cover in MAINTAINERS then the script look
> > at git history. If all files are well covered, then we shouldn't need
> > this fallback anymore, nor this .mailmap translation.
>
> Git's own .mailmap contains multiple "replace old addresses by a more
> recent one" entries. For example, the latest commit for that file is
> exactly for that purpose:
> https://github.com/git/git/commit/4aa72ea1f64e8ddcd1865c76b24591c0916c0b5d
>
> The situation looks same for .mailmap of Linux kernel too:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/.mailmap?id=1bb54043ff309795c90ccadd8a6e6b13ac40ec4e
>
> My understanding is that Git's handling of .mailmap is designed to find
> out email addresses to be used to reach out. For example, someone may
> git-blame and find I wrote some buggy commit, and ask me to review a fix
> for it; in that case it is nicer if git-blame displays my latest contact
> point (@rsg.ci.i.u-tokyo.ac.jp) instead of one for my previous employer
> (@daynix.com), even though I authored the commit for Daynix.
>

This is my understanding as well, it's not for employer attribution,
it's for name change and/or email change.

People who wish to use git log for employer attribution reasons must
use --no-mailmap. The original commit author in the tree does not go
away!

> >
> >>
> >>> +Philippe Mathieu-Daudé <[email protected]> <[email protected]>
> >>>   Stefan Brankovic <[email protected]>
> >>> <[email protected]>
> >>>   Yongbok Kim <[email protected]> <[email protected]>
> >>
> >> [...]
> >>
> >
>
>

-- 
Manos Pitsidianakis
Emulation and Virtualization Engineer at Linaro Ltd

Reply via email to