Re: [PATCH v4 0/5] Advice on checkout dirty files

From: Junio C Hamano <gitster@pobox.com>
Date: Mar 11, 2026, 16:38 UTC
Message-ID: <xmqqikb2qq20.fsf@gitster.g>
In-Reply-To: <pull.2233.v4.git.git.1773219028.gitgitgadget@gmail.com>

"Arsh Srivastava via GitGitGadget" <gitgitgadget@gmail.com> writes:

Show 23 quoted lines

> This is my submission for microproject [GSOC]
>
> This patch adds a new advice type ADVICE_STASH_BEFORE_CHECKOUT to help users
> when they attempt to switch branches with local modifications that would be
> overwritten by the operation.
>
> The new advice follows the same patterns established by existing advice
> functions such as advise_on_updating_sparse_paths(). When triggered, it
> lists the affected files and suggests using git stash push/pop to save and
> restore local changes.
>
> The advice can be silenced with:
>
> git config set advice.stashBeforeCheckout false
>
> Changes:
>
>> advice.h: add ADVICE_STASH_BEFORE_CHECKOUT enum value advice.c: add
>> "stashBeforeCheckout" to advice_setting[] and implement
>> advise_on_checkout_dirty_files() function
>> Documentation/config/advice.adoc: document the new advice key
>
> Signed-off-by: Arsh Srivastava arshsrivastava00@gmail.com

Even though no developer is perfect, when you are presenting your updated work, armed with wisdom borrowed from your reviewers' comments on your earlier attempts, you are expected to take the opportunity to pretend to have written a series of patches that are perfect logical progression towards the final shape of the code without detours, change of plans, and fixing earlier mistakes made in the series.

Please do not throw a drunken-man's-walk series at us. For example, I see that [PATCH 3/5] literally removes what was added by earlier patches. This is not a place for you to show how you made wrong turns before arriving at the final shape of the code.

The final series accepted by the project will have to stay in our history for later developers to see in "git log" output to learn from, and a series being clean logical progression is a must for that to happen.

Also, what is "updation"? Is it a standard English word, or some dialect of an LLM origin?

Show 27 quoted lines

> Arsh Srivastava (5):
>   advice: add stashBeforeCheckout advice for dirty branch switches
>   advice: add stashBeforeCheckout advice for dirty branch switches
>     [GSOC]
>   unpack-trees: suggesting 'git checkout -m <branch>' with its
>     repercussions
>   Updating tests and unpack-tress.c [GSOC]
>   File updation [GSOC]
>
>  t/t6439-merge-co-error-msgs.sh | 6 ++++++
>  t/t7406-submodule-update.sh    | 3 +++
>  unpack-trees.c                 | 9 +++++++--
>  3 files changed, 16 insertions(+), 2 deletions(-)
>
>
> base-commit: d181b9354cf85b44455ce3ca9e6af0b9559e0ae2
> Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-git-2233%2FArsh123344423%2Fadvice_on_checkout_dirty_files-v4
> Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-git-2233/Arsh123344423/advice_on_checkout_dirty_files-v4
> Pull-Request: https://github.com/git/git/pull/2233
>
> Range-diff vs v3:
>
>  1:  eb5639dbc3 = 1:  eb5639dbc3 advice: add stashBeforeCheckout advice for dirty branch switches
>  2:  e88c851701 = 2:  e88c851701 advice: add stashBeforeCheckout advice for dirty branch switches [GSOC]
>  -:  ---------- > 3:  4237b9667d unpack-trees: suggesting 'git checkout -m <branch>' with its repercussions
>  -:  ---------- > 4:  b25ea22410 Updating tests and unpack-tress.c [GSOC]
>  -:  ---------- > 5:  2ef7d5a3d6 File updation [GSOC]