“It never reimplements git — it shells out to the system git CLI and rebuilds commits with git commit-tree, reusing each commit's original tree so file contents are provably never changed.”
Glad the LLM noted this - I was worried this would reimplement git
One of the most annoying quirks of LLMs for me is the insistence on, after being corrected, loudly noting what wasn't done. After getting it to change something stupid in code, it will leave a comment in the code boasting about how it doesn't do the stupid thing, which to future readers reads like an 'asbestos free!' label on a cereal box.
Lets assume one of two worlds - one where AI is constantly improving and one where the tooling is a dead end. In the first world the AI generated comments are no better than the commentary that would be generated in the next six months - in the second world the AI generated comments will obscure the human written comments.
In what world are AI generated comments actually value adds?
If the AI (or the person) ran into a non-obvious constraint, or some externality, or there’s a plan that just isn’t implemented yet- then the comment wouldn’t be able to be divined by the AI. But yeah just pointing the AI at something and saying “write me docs” will get you something at best equally good as what’s going to be available.
One of the reasons this is there is the "psychology" of the LLM performs better with these kinds of affirmations. If you've made a mistake once you're likely to make it again, so leave a note.
This indeed pollutes the code with a bunch of nonsense. One of the things I have it maintain instead is a "findings" document with lessons learned and such.
I have a contrarian take on this: if most people working on the codebase are using the same model, then these comments serve a very useful purpose: they tell the model's future self that despite its tendency towards this solution, it is incorrect and here's why. It encodes useful context right when the model needs it next to avoid making the same error.
I imagine this style of commenting is partially responsible for model improvement in SWE work, a trained trait through feedback, since it makes the comments actually useful. Not to you or me, but to an LLM.
Some of these errors are obvious to a human, that's why we find the comments unnecessary. But if everyone's using LLMs and said model tends to make this error, then they are very useful. It's like a comment you would leave for another human to help them avoid a footgun. It's just a LLM footgun, not a human one.
The problem is that the list of things one shouldn't do is infinite. The things one should do are narrow and get narrower along two independent axes of clarity/wisdom: system requirements and engineer seniority. LLMs overindex on any words given, so you only want to give them negative guidance around truly repeated, almost common-sense pitfalls. But in a large distributed system often the changes come from all different angles, and each agent will find it's own unique failure modes. Those comments will dilute critical context for diverse agents far more than they will help on average, at least for the systems I'm working on, given current claude code chattiness.
One way to clear some of the low-hanging slop is to just have a separate agent code reviewing and pruning any comments that don't stand on their own purely in the context of the diff, but it still doesn't catch it all.
I agree, but that information should be in the commit, which the model also has access to. Of course it's more token intensive to look through commits, but "good" models might know when some more research about a particular bit of code they find weird would be a good idea. Thoughts?
If LLMs are any good, like many of you claim, they won’t need these comments. They will ingest the code and know what was done. They won’t need comments about what wasn’t done.
For what it's worth, I've used Claude to write a lot of smaller utility tools that I wouldn't normally have the time to sit down and implement, but which also solve straightforward problems.
One small example is an extension for the `gh` cli where I give it a PR number and multiple branches and it cherry-picks that PR to those branches, or lets me know if some of the branches don't work. For managing multiple release branches it's very handy, but not so handy that I would have spent a lot of time and energy doing a good job of it.
This seems like a similar sort of thing - a user friendly way to do something that most people don't need, and most people who need it probably don't need it very badly.
I haven't used claude to replace my actual skills - researching, solving problems, etc. - but I do use it to replace all the stuff which is essentially 'spend a few hours reading API docs and writing boilerplate code'.
That is in fact the difficult part. You have to explain it all coherently while adjusting to the reader's frame. No one does that for no good reason. It is hellish work.
Effective and clear communication is really important and often really hard for engineers. It is said that one goal of stackoverflow was to help programmers learn how to write through practice, as it’s both very hard and very critical to their effectiveness:
I mean, given that the English you wrote above didn't make it clear that was the only part you were complaining about, it's worth it to consider that maybe precision is hard in natural language regardless of how it's written
I'm reverse engineering a cherished video game from my childhood. I got to the audio engine part and things got pretty difficult. Managed to map out the locations of the audio data but figuring out which song was which was going to take a long time.
Claude somehow wrote a bespoke one-off Python script to play back the songs straight out of the ROM. I give it the ROM and a hexadecimal address and it just plays it. Converts the data from the ROM into PCM audio samples and pipes them over SSH into termux on my phone where they are played back via SoX play.
Yes, Claude wrote a literal sound synthesizer so I could do this. Claude painstakingly tuned the goddamn synthesizer too, he tweaked it over and over again until the output sounded right based on annoyingly human feedback like "these notes are supposed to be in the background but they are too loud".
This tool turned a painstaking reverse engineering task into a nostalgic listening experience that brought literal tears to my eyes. I mapped out every single song in the game this way.
Without LLMs, none of this would ever have happened.
i definitely used to write bash functions for this sort of thing tho. no GUI, but it was useful to do stuff like obscure my working hours (make it look like i did personal stuff after-hours. make it seem like stuff took all day instead of an hour..like adding sawdust to meat).
like you said, niche and probably unnecessary. and if i were to have published them (even in a gist!) i would have written the docs myself instead of forcing people to read LLM-speak. That's just lazy.
Ha ha fair. But yeah shelling out to git and reusing the original tree object was deliberate means file contents can't change even if I screw up elsewhere.
This is, funnily enough, something I've been periodically searching for for nearly a decade, and even renewed my search yesterday. For those wondering, my use case is a model Parliament (context: https://news.ycombinator.com/item?id=43474850) and deciding on using git to store the laws, with the commits representing the Acts of Parliament themselves. The problem is that we need to guarantee that every commit has the correct metadata: we do not care at all about the conservation of commit hashes. And if we've accidentally missed a law and need to interactively-rebase it into the git history, it should NOT reset all the subsequent commit dates to the current timestamp. Basically, we're wanting to use git as a historical archive. I'm delighted to have found this.
It looks like the screenshot was an actual photo of someone's monitor. I'm left wondering why print screen wasn't utilized. And for some reason it really makes me not want to touch this project.
2. Sometimes I split a Git repo into two, and I want to preserve the commit history (authors and dates) on the files that end up in the new repo - effectively replay the history of just one folder from the original to help create the new one.
3. I occasionally mess up cleaning up and merging a PR from an open source contributor such that the work is incorrectly credited to me. I'll fix the commit so that it credits the right person.
Yes, for instance it could be useful if someone’s been accidentally committing with a bogus email address. Or if you’re cleaning up some sort of mess and would like to recreate an accurate history.
Obviously as with any “force push” situation, the usual caveats apply if this is on a branch others have checked out, but if the mess is big enough, it is sometimes worth it to send everyone a script to reset or just to have people delete and re-check-out.
Yes. Absolutely! Many times I had a wrong configured terminal and messed up the author. Often I work in different projects, privately and in my work, and I use different authors for both
Very rarely. But it can come up if you archive change history in git for things that aren't originally tracked in git.
Take laws for example. Legalize[1] is a project that documents all laws into a git structure where each distinct change is a commit with the proper timestamp of tha change. Sure, this specific project builds that entire git history once and it's unlikely a retroactive change required such edits. But do something like this for historical sources where information is still coming in, and you might need to change a date when new findings occur.
- An agent messed with my git config for some reason so I had to go back and fix the commit authorship
- I created a PR mostly based on work that someone else did, so I reauthored most of the commits to be in their name
The few times I've wanted to completely rearrange a repo, It's been enough just make a backup branch and then dump the log from there and pull what I want.
I have script somewhere I think it's called git-cherry-replace, which soft resets a commit id and copies the commit message and optionally populates GIT_COMMITTER_DATE and whatnot.
What I described is probably simple enough that an LLM can implement it.
But unless you're trying to retroactively reconstruct something to preserve the exact meaning, you should consider whether a rebase even makes sense.
Please put up this or another screenshot as the first thing you see in the README - please!
I'm reluctant to clone a repository and spend time on it just to find out the project looks crap and I don't want to use it. A screenshot saves other people's time :)
its vibecoded. just point another vibe based harness at it and youll have it in an hour or two. this thing is as trivial as you can get get, scope wise
Two recent threads where I sing some praises of git rebase -i, specifically linking it to a spreadsheet. Awesome to see this pop up, feels serendipitous!
I built a small Tauri app that just shows your commits in a table and lets you edit message/author/date directly, plus find & replace with regex if you need to fix a bunch at once. It backs up the branch before it rewrites anything doesn't touch file contents, just metadata.
Glad the LLM noted this - I was worried this would reimplement git
reply