Re: [hatari-devel] Git commits (was: HDC code update patch)

[ Thread Index | Date Index | More lists.tuxfamily.org/hatari-devel Archives ]


Uwe Seimet schrieb:

Don't get me wrong, but let me just say that the recommended length of git
commit messages is 70-75 characters. https://i.stack.imgur.com/uzUqh.png
shows that for the Linux kernel the contributors mostly manage to stay
within this range.

Sorry, but I think you're confusing something. The statistic you posted
above refers to the length of the *summary* line [1], i.e., the first
line of the commit message.

As you can see from this randomly picked Linux commit...
https://github.com/torvalds/linux/commit/413a4a6b0b5553f2423d210f65e98c211b99c3f8
.... the actual message is quite detailed -- and rightly so.

Eero was not asking you to put all details into the summary line, of course.

When writing a commit message, I try to provide detailed reasons for the
change, so that another developer many years in the future can hopefully
understand. Even if this means that I sometimes spend more time writing
the commit message than the actual code change. See, e.g., this commit:
https://github.com/emutos/emutos/commit/819011e

Again, don't get me wrong, but if what you pointed out is the Hatari commit
message policy, it is IMHO not very developer-friendly.

Then again, it's not very maintainer-friendly if they have to do a lot
of routine work such as writing commit messages for contributor's diffs.

I assume that your git workflow contains commits; as you would need
those for a Github PR, too. Then, like Eero says, imho it's not much of
a difference whether you do "git diff" or "git format-patch origin".

Regards
Christian

[1] https://stackoverflow.com/a/11993051
--
Christian Zietz  -  CHZ-Soft  -  czietz@xxxxxxx
WWW: https://www.chzsoft.de/
PGP/GnuPG-Key-ID: 0x52CB97F66DA025CA / 0x6DA025CA



Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/