b4 feature request: run command before rewriting history

b4 feature request: run command before rewriting history

[Toke Høiland-Jørgensen]

Hi Konstantin

I recently started switching my workflow over to 'b4 prep' for preparing
patches, and am loving the workflow. Thanks a lot for working on this!

However, I ran into one little snag, that I'm hoping could be fixed with
a new b4 feature: I am also a big fan of 'stacked git' (stgit)[0], and
often use that instead of 'rebase -i' when preparing a series for
submission.

So naturally, I want to combine 'b4 prep' with stgit. The problem here
is that stgit maintains its own state for the branches it manages, and
gets awfully confused if the git history is rewritten underneath it.
It's still possible to use it with 'b4 prep', as long as I remember to
do 'stg commit --all' before running any of the b4 commands. And, well,
guess what I already forgot several times? :)

So, my feature request is this: add support to b4 for a hook command
that is executed before any operation that will rewrite history - and
abort the rewriting if the command fails. That way, I can make sure that
I don't forget to clear the stgit state before running b4, saving me
from the pain of dealing with a confused tool.

Do you think this would be feasible to implement?

Many thanks!

-Toke

[0] https://stacked-git.github.io/

Re: b4 feature request: run command before rewriting history

[Toke Høiland-Jørgensen]

Toke Høiland-Jørgensen <toke@redhat.com> writes:

> Hi Konstantin
>
> I recently started switching my workflow over to 'b4 prep' for preparing
> patches, and am loving the workflow. Thanks a lot for working on this!
>
> However, I ran into one little snag, that I'm hoping could be fixed with
> a new b4 feature: I am also a big fan of 'stacked git' (stgit)[0], and
> often use that instead of 'rebase -i' when preparing a series for
> submission.
>
> So naturally, I want to combine 'b4 prep' with stgit. The problem here
> is that stgit maintains its own state for the branches it manages, and
> gets awfully confused if the git history is rewritten underneath it.
> It's still possible to use it with 'b4 prep', as long as I remember to
> do 'stg commit --all' before running any of the b4 commands. And, well,
> guess what I already forgot several times? :)
>
> So, my feature request is this: add support to b4 for a hook command
> that is executed before any operation that will rewrite history - and
> abort the rewriting if the command fails. That way, I can make sure that
> I don't forget to clear the stgit state before running b4, saving me
> from the pain of dealing with a confused tool.
>
> Do you think this would be feasible to implement?

Hi Konstantin

Friendly ping on the above? :)

-Toke

Re: b4 feature request: run command before rewriting history

[Konstantin Ryabitsev]

On Wed, Oct 30, 2024 at 11:16:15AM +0100, Toke Høiland-Jørgensen wrote:
> > Do you think this would be feasible to implement?
> 
> Hi Konstantin
> 
> Friendly ping on the above? :)

Yeah, I hope to be able to look into this in the near future. Unfortunately,
I've been dealing with less-than-pleasant tasks such as ddos attacks and
trolls over the past couple of weeks, but at least it's finally on the wane as
people move on.

-K

Re: b4 feature request: run command before rewriting history

[Toke Høiland-Jørgensen]

Konstantin Ryabitsev <konstantin@linuxfoundation.org> writes:

> On Wed, Oct 30, 2024 at 11:16:15AM +0100, Toke Høiland-Jørgensen wrote:
>> > Do you think this would be feasible to implement?
>> 
>> Hi Konstantin
>> 
>> Friendly ping on the above? :)
>
> Yeah, I hope to be able to look into this in the near future. Unfortunately,
> I've been dealing with less-than-pleasant tasks such as ddos attacks and
> trolls over the past couple of weeks, but at least it's finally on the wane as
> people move on.

Right, no worries, just wanted to make sure this didn't get lost
entirely. Thanks for looking at it! :)

-Toke

Re: b4 feature request: run command before rewriting history

[Toke Høiland-Jørgensen]

Toke Høiland-Jørgensen <toke@redhat.com> writes:

> Konstantin Ryabitsev <konstantin@linuxfoundation.org> writes:
>
>> On Wed, Oct 30, 2024 at 11:16:15AM +0100, Toke Høiland-Jørgensen wrote:
>>> > Do you think this would be feasible to implement?
>>> 
>>> Hi Konstantin
>>> 
>>> Friendly ping on the above? :)
>>
>> Yeah, I hope to be able to look into this in the near future. Unfortunately,
>> I've been dealing with less-than-pleasant tasks such as ddos attacks and
>> trolls over the past couple of weeks, but at least it's finally on the wane as
>> people move on.
>
> Right, no worries, just wanted to make sure this didn't get lost
> entirely. Thanks for looking at it! :)

Hi Konstantin

Another friendly ping on this? :)

-Toke