Thanks all for the info, we are going to run git filter branch next. On Sun, Apr 18, 2021 at 9:28 AM Jed Brown <j...@jedbrown.org> wrote:
> Andy Grove <andygrov...@gmail.com> writes: > > > We started looking at the documentation for git filter-branch and it > > recommends not to use it. It states that "git-filter-branch is riddled > with > > gotchas resulting in various ways to easily corrupt repos or end up with > a > > mess worse than what you started with:". > > I've used it quite a bit (including splicing 25 years of history across > five SCMs) and have found it does what's on the label, that is just often > not what people expect. > > I think it's fine in this case with the caveat that references to the > parent directory (looks like for testing/data/, perhaps other places; > actual license text) will no longer be consistent with the rewritten commit > (potentially with rust/ subdirectory filtered into the base directory). > Some structured editing can be done via script (say, if you want > LICENSE.txt to always be present). If your goal is to have good provenance > with respect to "who authored this and when", but not necessarily "let's > bisect this bug", then filter-branch would make sense. > > > I guess we can decide to run this at any time, so let's discuss this more > > once we have the repos building? > > Yeah, just make sure to do it before people start doing work in their > clones. >
