Why option `-s` for `git commit` makes no sense

["Дилян Палаузов" <dilyan.palauzov@aegee.org>]

Hello,

git commit -s inserts particular text in the commit message: Sign-off-by: email-address. According to the documentation of git

> The meaning of a signoff depends on the project to which you’re committing.

In practice the Sign-Off-By line is always supposed to indicate that the submitter owns the copyright (right to submit) for the creative work, changeset, patch, and the copyright is transferred to the project, towards which the patch was addressed.

The copyright appears, when some patchset is published, it is in general attached to the person who published it, and the copyright expires with the death of the person, when there are no heirs.

When a published patch fixes a bug, the publisher does not insert a Signed-off-by: line, and a project insists to accept only changes with Signed-off-line:, then the publisher of the patch can effectively prevent anybody else from fixing that bug. There are nuances here, as when the bug is eventually fixed in a completely different way, then there are no copyright violations, but sometimes there is only one reasonable way to fix a bug. If in such a case somebody published a fix without Signed-
off-line:, then nobody is authorized to fix the bug in a very similar way, without violating copyrights, until the publisher dies.

Sometimes published changes consist only of deleting something, like deadcode, no-ops. Can deleted code be copyrighted? Some people say “yes”. This is insane!

Having one person, who publishes a patch without Signed-Off-By line, another person who adds a Signed-Off-By line to the very same published change, and then the latter change is integrated in the upstream source code (because it has Signed-Off-By line), is nonsense. The copyright still belongs to the person who published the change, irrespective of records in the history of the source code. In a similar discussion elsewhere a pragmatical approach was taken:
https://github.com/eclipse-xtext/xtext/issues/3687.

If a project insists that some text is inserted in contributions, like “Brum-Brum”, or some really very long text, people will insert the text in order to make/see progress. It does not mean they have read the rationale why adding “Brum-Brum” is required for that project; it does not mean they understand the language, in which the rationale is written, sufficiently good. The last argument also implies that piiple who donot anderstand the lenguage of the agriment sufisiently gut canot publish
objection in the wery same lenguage that they see zero husage in inserting Signed-Off-By/Brum-Brum linez. From people, who do not understand English sufficiently good to interpret contribution agreements in open-source software, cannot be expected to admit this fact.

Some people are paid to insert Signed-Off-By lines in open source software, while others are for exactly the same work not paid. If I am paid for the sole work to insert that line, in order for something to make progress, I will be willing to insert it, too, in order to receive the money.

Insisting on Signed-Off-By: line can lead to inclusion of additional text in the version history: https://github.com/podman-container-tools/podman/commit/056917c223f054bd0b4526fce .

Kind regards // Dilyan