[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Git commit messages
From: |
Florian Dold |
Subject: |
Re: Git commit messages |
Date: |
Wed, 27 May 2020 14:09:53 +0530 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.8.0 |
On 5/27/20 1:58 PM, Daniel Golle wrote:
> Yes, let's please establish something like that. I've tried my best
> myself, but that's like 0.000001% of GNUnet commits -- and of course,
> as was pointed out in a previous debate on that, GNUnet is not the
> Linux kernel and experimentation becomes very slow if you have to
> write a novel on every single step you do...
>
> Another aspect is me struggling with reviewing things for GNUnet
> because often times I can see what a commit does but I have no idea
> what it was **supposed to do**, due to lacking commit descriptions.
> Also knowing whether a commit was meant to be purely cosmetic is nice
> when reviewing. So maybe we can be a bit more differentiated with our
> to be established language of symbols:
> Let's use '- ' prefix for small fixes or follow-up commits, just like
> you suggested above.
> Let's use '# ' prefix for purely cosmetic changes (ie. not functional
> changegs intended).
I would *strongly* suggest not using "# " as the prefix for any type of
commit message, as it's the default comment prefix when editing git
messages ... that'll lead to all kinds of problem. You don't want to
tell people to change that comment character in their git config to,
say, "%" just for GNUnet ;-)
The minus is not very self explanatory. I've been sporadically working
on GNUnet-related stuff since about 2012 and $TODAY is the the day I
learned what the intention of the "- " prefix is (it was sporadically
used before, especially during TUM times).
Why not use something some prefix like "minor: " that is actually
possible to figure out on your own?
- Florian