threads / patch / 22039

patchDocumentation: commit: explain the non-meaning of S-o-b

Subject: [PATCH] Documentation: commit: explain the non-meaning of S-o-b

## tl;dr

2 messages between Dec 28, 2009 and Dec 29, 2009. Diffs are folded; open one to read it.

replies: 1people: 2as markdown or json

Jan Krüger· Dec 28, 2009, 18:40 UTC · lore

In the manpage for git commit, the option --signoff is mentioned but there is no explanation of what it actually means. Add a brief hint that S-o-b doesn't have a pre-defined meaning.

Signed-off-by: Jan Krüger <jk@jk.gs>
---
Semi-resend. Nobody commented on this when I first sent it in early
December, so here it is again, with a slightly more verbose
explanation. The repetition is intentional.
 Documentation/git-commit.txt |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)
Show changes to Documentation/git-commit.txt +4 −1
diff --git a/Documentation/git-commit.txt b/Documentation/git-commit.txt
index d227cec..cae510b 100644
--- a/Documentation/git-commit.txt
+++ b/Documentation/git-commit.txt
@@ -114,7 +114,10 @@ OPTIONS
 -s::
 --signoff::
 	Add Signed-off-by line by the committer at the end of the commit
-	log message.
+	log message. This line has no inherent meaning; it is up to the
+	potential recipient of the commit to decide what it stands for.
+	It is typically understood as an assurance by the committer that
+	the commit conforms to the receiving project's commit guidelines.
 
 -n::
 --no-verify::
-- 
1.6.5.3.171.ge36e
Junio C Hamano· Dec 29, 2009, 07:54 UTC · re: Jan Krüger · lore

Re: [PATCH] Documentation: commit: explain the non-meaning of S-o-b

Jan Krüger <jk@jk.gs> writes:
> In the manpage for git commit, the option --signoff is mentioned but
> there is no explanation of what it actually means. Add a brief hint that
> S-o-b doesn't have a pre-defined meaning.

The commands am, cherry-pick, commit, format-patch, and revert all share the "sign-off" and they more or less deliberately refused to define what it means to sign-off a patch; as your patch says, it is up to the project policy.

config.txt gives a description for format.signoff option that is too specific to git and the kernel, which I think is a mistake given the much wider audience we have these days.

How similar should the document for the "sign-off" options to these commands be? I am wondering if we should have the same description copied and pasted to all of them, or avoid repeating ourselves and instead have them refer to a central place, perhaps somewhere in the Tutorial or the Workflow document.

Show 27 quoted lines
> Signed-off-by: Jan Krüger <jk@jk.gs>
> ---
> Semi-resend. Nobody commented on this when I first sent it in early
> December, so here it is again, with a slightly more verbose
> explanation. The repetition is intentional.
>
>  Documentation/git-commit.txt |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
>
> diff --git a/Documentation/git-commit.txt b/Documentation/git-commit.txt
> index d227cec..cae510b 100644
> --- a/Documentation/git-commit.txt
> +++ b/Documentation/git-commit.txt
> @@ -114,7 +114,10 @@ OPTIONS
>  -s::
>  --signoff::
>  	Add Signed-off-by line by the committer at the end of the commit
> -	log message.
> +	log message. This line has no inherent meaning; it is up to the
> +	potential recipient of the commit to decide what it stands for.
> +	It is typically understood as an assurance by the committer that
> +	the commit conforms to the receiving project's commit guidelines.
>  
>  -n::
>  --no-verify::
> -- 
> 1.6.5.3.171.ge36e

← back to recent threads