Was digging through a project at work today where some guy in 2014 made 100+ commits in a single day and the only one that had a comment said “upgrading to v4.0”.
I’ve had commits called fuck
git commit -m “minor tweaks”
+3,276 -4,724
Bug fixes. Too many to count.
I had one of those and it was two in the night and I was tired and forgot what I did and committed
stuff, I dunno.But normally I’m a good boy and prefix with the ticked id and write down the change and attempted fix.
Conventional commits all the way! Even if I don’t use the keywords (feat, fix, etc.) I always write the comment in imperative tense; the message should tell you what happens if you merge it.
I totally agree.
Right now I’m on a new project with a teammate who likes to rebase PR branches, and merge with merge commits to “record a clean history of development”. It’s not quite compatible with the atomic-change philosophy of conventional commits. I’m thinking about making a case to change style, but I’ve already failed to argue the problem of disruption when rebasing PR branches.
Enforced by pre-commit, conventional commits has cleaned up our commit logs and changelog so much.
I try (my best) to follow https://www.conventionalcommits.org/en/v1.0.0/
deleted by creator
They fluxuate wildly between short and informative messages like “fixed regex validation on property A” and “I fucking hate prettier” when the build pipeline fails because I had a line that was 2 characters too long.
On projects I setup I have prettier run as part of a commit hook. All files will be formatted at all times
You get two options.
Normally it’s a squashed commit of everything in a feature, with a commit message like:
[JIRA-1234] - Descriptive but Concise Name of FeatureBut every now and then it’s multiple commits like:
quick fix Ugh, fix typo fuck fuck why doesn’t it work Oh, I’m stupidFollowed by
fixed formattingfinal formatting fixyou gotta be kidding me, fuck you, detekt!Bro, squash merge
Or if you’re using feature branches, rebase, squash, and force push before opening the MR
Sure, but before squashing you gotta commit
My commits tend to be pretty verbose. Here’s an example log from one of my projects.
I follow the standard imperative style for the commit title, and then I use the body to summarize any important internal changes, reflect on the overall project status (for example, what milestones this commit crosses or what other work it might enable or require), and state what I’m going to work on next. I’m sure some people find it too wordy, but I like having the commit history show lots of details about the overall status.
Edit: I always have a descriptive summary, i.e., never one word commits or similar.
you are a pro and I aspire to be you
I’m not sure I do. I wouldn’t want to read all that just to find the item that broke. Might be faster to read the code.
That’s why
git log --onelineexists ;)I use
alias gl='git log --graph --abbrev-commit --no-decorate --date=format:'\''%Y-%m-%d %H:%M:%S'\'' --format=format:'\''%C(8)%>|(16)%h %C(7)%ad %C(8)%<(16,trunc)%an %C(auto)%d %>|(1)%s'\'' --all'It will change your world.That is sexy. My only problem is that I tend to run my Git operations in a pretty small
tmuxpane on the side of my editing pane, so that layout ends up being too wide to fit well. I’ll definitely keep that alias around for when I have a full screen though!Haha yea I have written a number of
gitanddockeraliases over the years that are permanently in mydotfiles. I’m always inscreenbut perhaps will get into this newfangledtmux.
I like my company’s style:
For issues:
<jira ticket> - [program][deliverable] did this to fix that
Problem: symptoms of the problem that future devs can use to figure out its the same problem
Root cause: why this is broken
Solution: how I fixed it, including the scope
Testing: what testing it has it gone through
I try to follow the BLUF pattern: Bottom line up front. The first line is as short a description of the change (“Re-fixed a bug where a URL without a verb could crash the bot.”) with some detail following (“I thought I caught that a couple of years back…”)
I try to save the detail for the code itself: Comments describe what I was thinking at the time for context, the code is the code. I don’t replicate the code comments in the commit message because having the same thing in two places means having to keep two things up to date, and that rarely goes well.
I always try to capture the reasoning behind why I am making the change. I wrote about this more here https://lencioni.medium.com/the-secrets-to-great-commit-messages-106fc0a92a25
I simply commit to master with the message “git commit”.
Per XKCD, “as a project drags on, my git commit messages get less and less informative.”
I feel like this might be a good case for LLMs… Auto git commit suggestions based on the diff.
@CanadianNomad @lucas Awesome idea, yeah I can totally see that working well :)
There are already some attempts but I don’t think it will work, harmful even. Best case scenario, the AI can understand the code as well as a senior engineer from another company. All they can know without the context is what was changed, which is useless. We need the reason why the commit was made, not what was changed. The info is not there in the first place for the AI to try to extract.
Doesn’t GitHub copilot already do this?
If it does, I haven’t seen it… I’d be happy to test drive it.
Developer Initials - Jira ticket number which includes the project abbreviation and the ticket number - brief description:
DA - HHGTTG-42 - fix question answer format
If you need details you look in the ticket.
Developer intials seems a tad redundant since the commit is tied to author(s). But I guess it is only 2 extra char
Literally about 90% of the commits I’ve pushed on my DayZ server’s configs and mod files are just marked ‘a’. The actual mod updates I almost never have made porch notes for. Trying to be a little more informative for my new D&D based Conan Exiles server.
It still looks better than how I used to name things in flash.
