Re: Rename offensive terminology (master)

From: Konstantin Ryabitsev <konstantin@linuxfoundation.org>
Date: Jun 9, 2020, 16:06 UTC
Message-ID: <20200609160650.7e74bcjksrto3rp6@chatter.i7.local>
In-Reply-To: <CAOAHyQx=+fM1FpAv+g3M+j7j4MgLJA03=MGFmXLvZcfJKAEpGg@mail.gmail.com>

On Tue, Jun 09, 2020 at 05:16:57PM +0200, Simon Pieters wrote:

Show 13 quoted lines

> Thank you for your encouraging response, Brian, and the research of
> what the change entails for git.
> 
> I've added Don to the cc, who started to work on implementing this change:
> 
> https://twitter.com/DEGoodmanWilson/status/1269931743320182784
> https://github.com/git-for-windows/git/issues/2674
> 
> Although I think it's reasonable to move away from 'master' regardless
> of its origin, today Tobie Langel pointed me to
> https://mail.gnome.org/archives/desktop-devel-list/2019-May/msg00066.html
> where, one year ago, Bastien Nocera made the case that git's 'master'
> is in fact a reference to master/slave.

Well, he pointed out that Bitkeeper used this terminology. Git doesn't have any internal concept of "slave" -- the only time you see this word used in the codebase is in the test suite, and we should absolutely change that.

I am torn on this issue -- I certainly want the project to be inclusive to all, but English has a lot of concepts that start with "master" and do not trace their origin to subjugation of fellow human beings:

- masterpiece
- masterful
- master's degree
- master copy

Making this change in git seems like attacking the problem at the wrong end.

Branch names are already fully arbitrary in Git -- you can have a repo without a master branch. Perhaps the best way to address it is to introduce a "default branch name" configuration variable, or just work without any default branches and let the next step after "git init" be "git branch".

-K