Re: onwards to git

[Jim Meyering <jim at meyering dot net>]

Andreas Schwab wrote:
> Jim Meyering <jim@meyering.net> writes:
>> One is very much like the existing glibc.git, with over 64k commits on
>> the trunk/master.  The other has the same content, but with aggregated
>> change sets (and usually-better commit log messages[*]), so there are
>> only 13k commits.
>
> It appears the process of aggregation has clobbered the committer in
> some commits.

Thanks for checking!

> For example 03f1681f in glibc1.git has Ulrich as the
> committer with the same timestamp as the committer of the original child
> 031fc0c9 in glibc2.git.

031fc0c9 appears not to be relevant.

The aggregate glibc1 commit, 03f1681f, corresponds to these in glibc2,
so you must mean one of these:

  8ebfbaec5b6ddf227b9a0289dcb96d9bfb46268d linuxthreads/ChangeLog
  65d64b95919286e40b94f4ac60ff38b1edb54d1d linuxthreads/sysdeps/m68k/Makefile

and sure enough, you committed them, yet the rewriter changed the
"Committer:" metadata to Ulrich in that case.  That is because the
rewriting script took no special pains to preserve committer
information.  It usually got it right, but obviously not always.

I've just fixed it so that it does preserve committer name and email.
Will push later today.

> I would also suggest to filter out all commits with the "file ... was
> initially add on branch ..." commit message.

It appears they have been automatically elided, probably
because the patch portion is empty:

    $ go glibc1; git log --raw |grep initially.added
    $

Do you still see vestiges in glibc1?