Hi,

On Tue, 1 Jun 2021, ZheNing Hu wrote:

> Junio C Hamano <gitster@pobox.com> 于2021年6月1日周二 下午4:54写道：
>
> > ZheNing Hu <adlternative@gmail.com> writes:
> >
> > > [...] But if I want to base the current topic on
> > > zh/ref-filter-atom-type, GGG will send past patches
> > > (zh/ref-filter-atom-type) repeatedly.
> >
> > I thought GGG lets you say "this is based on that other branch, not on
> > the 'master' branch" to solve that exact issue?
>
> I'm not sure...I will try it after I rebasing this topic to
> zh/ref-filter-atom-type.

Yes, it should be possible to rebase your patch on top of one of the
[a-z][a-z]/* patches Junio publishes at https://github.com/gitster/git
(and which get mirrored automatically to
https://github.com/gitgitgadget/git via a scheduled Azure Pipeline), and
then to change the PR base (simply click the `Edit` button next to the PR
title, as if you wanted to edit the title, and you can also change the
base branch).

> I just remember that it looked like something went wrong with my base on
> your patch last time.[1]
>
> [1] https://lore.kernel.org/git/pull.870.v6.git.1613739235241.gitgitgadget@gmail.com/

https://github.com/gitgitgadget/git/pull/870 seems to be based on
jc/diffcore-rotate all right.

If you are talking about including Junio's patch in v5
(https://lore.kernel.org/git/fb4bfd0f8b162e51e71711fe5503ca684f980d58.1613480198.git.gitgitgadget@gmail.com/#r),
I _think_ that there might have been the simple problem of
jc/diffcore-rotate having been force-pushed just before you sent v5, and
therefore you had a stale branch.

To prevent things like that, it is a good idea to set the upstream of your
local branch accordingly (in this instance, `git branch
--set-upstream-to=gitgitgadget/jc/diffcore-rotate`) and ensure to `git
pull --rebase` before force-pushing and submitting.

Ciao,
Dscho