I wouldn't say these are "basic bugs". The first is specific to using 'rrsync', and the second is when using the rsync daemon, and I can't remember when I last saw a system using that one (yes, I'm aware there are still use cases for using the rsync protocol, but I would consider it pretty obscure nowadays).
You could argue that he should've bumped the version more and should've done a longer beta test, but on the other hand, these were mostly security fixes, and I can understand he wanted to get them out there rather sooner than later (also "doing a beta test" is easier said than done - how do you get people to run a test version of rsync?).
Not saying that's the case here, but in a way that's how it's always worked. Ask for help, get little to no response. But break something and people respond in a hurry, especially if it somehow affects them.
I wonder whether there would have been less complaint had the number been a major increase, say to v4 beta. That amount of change in a minor version number bump seems unusual, but I don’t know rsync enough, I just use the version bundled in my 2/4 year old LTS distros and assume it works.
I do wonder about “The world of software engineering has changed dramatically in the last few months”. Has it really? I’ve been hearing that for a few years now.
A major version bump would have been helpful, and having such a large rewrite be initially a beta could have calmed the hoards.
> The world of software engineering has changed dramatically in the last few months
I disagree, but I guess we shall see how this is going to pan out. We've all introduced schoolboy errors in our time regardless of how long we've been developers. Messing up absolute paths because you only tested on relative happens to the best of us, but generally we say "woops" and try to fix it.
Doubling down and blaming users for your fuck up rarely ends well.
I don't think they blamed the users for that mistake, though. They specifically thanked people for reporting the regressions. It's the rage from bystanders which is totally out of proportion with the scale of the issues that they take issue with.
11 hours since this has been posted and 20 comments. Nobody that's raging really cares about the reasoning. All that matters is: AI use spotted -> time for the warpath.
> for the people saying things like “I’m a PhD from xyz uni and I’m telling your LLMs are just stochastic tools that make everything up and the world will fall apart if you use them”, I’m here to tell you that you are out of date. The world of software engineering has changed dramatically in the last few months.
It indeed has. I now review code written by machines that sucks the life of me, read reports written by a machine, and overall have a miserable life as a software engineer. The pay is good so I'll keep on it for the time being, but you are damn right it changed, for the worse.
The sooner we all collectively realize that the commonly cited "reasons why LLM code is bad" and "reasons poorly architectured, poorly planned, poorly tested code is bad" are a circle on a Venn diagram, the sooner we can hopefully make some effort to escape this sports fan tribalistic nonsense.
> I did not just vibe-code “convert test suite to python”. I’m a software engineer with 40 years experience (yeah, I’m OLD!)
Interestingly enough, I see the trend of people with decades of experience using AI more and more often. I'm in the 20 years club myself, and I do AI-assisted coding every day. It does help, and I'm grateful to have a tool like this to quickly try new ideas and throw them away if they suck. Shaming the author for using AI to help with the CI stuff is baffling to me. Are we witnessing just another ideology-driven tribal reaction on the rise?
I really feel for the developer of Rsync. He did absolutely nothing wrong and got flamed to hell for it. I’m glad he has help now and I hope all of the negativity doesn’t get to him. I would certainly have a difficult time with my mental health if I was on the receiving end of all this hate.
It's sad that the outrage posts got hundreds of comments while this article, from the maintainer explaining the CVEs and test suites, only has this singular comment and is already on the second page.
A lie travels around the globe while the truth is putting on its shoes. Do better, HN.
Maybe because the explanation isn't so much an explanation but an admission that core functionality was broken in a minor version release - and therefore almost orthogonal to the use of AI. If there had been no major regressions, do you think anyone would have complained?
People are (correctly) not going to be held to some kind of lower standard just because they "used AI" and "were fixing security issues".
>If there had been no major regressions, do you think anyone would have complained?
If there had been the same regressions but no AI, do you think there would be a 300 post issue full of people complaining? People are holding the update to a higher standard just because they used AI.
I think he should just retire for real (from rsync) and enjoy the time, who would take all the negativity and keep going for a volunteer project like rsync.
And this has nothing to do with AI. It seems to me that a significant amount of people in IT disconnected from reality.
Go throw a couple potatoes in the ground and see them grow or whatever. Your opinion most likely wont matter nor make any difference.
Please Do Not Vibe Fuck Up This Software - https://news.ycombinator.com/item?id=48342705 - May 2026 (466 comments)
Rsync 3.4.3 has hundreds of Claude commits - https://news.ycombinator.com/item?id=48334021 - May 2026 (81 comments)
A bit of a mixed one here.
The vibe coding got the project attention and it looks like he's going to get the help he needed.
However the "outrage", if you even call it that, wasn't entirely misplaced when pretty basic bugs were introduced by this, such as:
Can't use rsync with absolute paths:
https://github.com/RsyncProject/rsync/issues/922
Links mode is broken:
https://github.com/RsyncProject/rsync/issues/915
The scale of the commits, and rewriting the entire testing framework was pretty big for a "bugfix" revision.
It is entirely misplaced based on the evidence so far.
Neither of the bugs you link are currently traced back to a commit that used AI.
I wouldn't say these are "basic bugs". The first is specific to using 'rrsync', and the second is when using the rsync daemon, and I can't remember when I last saw a system using that one (yes, I'm aware there are still use cases for using the rsync protocol, but I would consider it pretty obscure nowadays).
You could argue that he should've bumped the version more and should've done a longer beta test, but on the other hand, these were mostly security fixes, and I can understand he wanted to get them out there rather sooner than later (also "doing a beta test" is easier said than done - how do you get people to run a test version of rsync?).
>The vibe coding got the project attention and it looks like he's going to get the help he needed.
I guess vibecoding certain software is now the new way of "instead of asking for help, write the wrong solution instead".
Not saying that's the case here, but in a way that's how it's always worked. Ask for help, get little to no response. But break something and people respond in a hurry, especially if it somehow affects them.
I wonder whether there would have been less complaint had the number been a major increase, say to v4 beta. That amount of change in a minor version number bump seems unusual, but I don’t know rsync enough, I just use the version bundled in my 2/4 year old LTS distros and assume it works.
This does once again feel like a case of https://xkcd.com/2347/
I do wonder about “The world of software engineering has changed dramatically in the last few months”. Has it really? I’ve been hearing that for a few years now.
A major version bump would have been helpful, and having such a large rewrite be initially a beta could have calmed the hoards.
> The world of software engineering has changed dramatically in the last few months
I disagree, but I guess we shall see how this is going to pan out. We've all introduced schoolboy errors in our time regardless of how long we've been developers. Messing up absolute paths because you only tested on relative happens to the best of us, but generally we say "woops" and try to fix it.
Doubling down and blaming users for your fuck up rarely ends well.
I don't think they blamed the users for that mistake, though. They specifically thanked people for reporting the regressions. It's the rage from bystanders which is totally out of proportion with the scale of the issues that they take issue with.
11 hours since this has been posted and 20 comments. Nobody that's raging really cares about the reasoning. All that matters is: AI use spotted -> time for the warpath.
> for the people saying things like “I’m a PhD from xyz uni and I’m telling your LLMs are just stochastic tools that make everything up and the world will fall apart if you use them”, I’m here to tell you that you are out of date. The world of software engineering has changed dramatically in the last few months.
It indeed has. I now review code written by machines that sucks the life of me, read reports written by a machine, and overall have a miserable life as a software engineer. The pay is good so I'll keep on it for the time being, but you are damn right it changed, for the worse.
The sooner we all collectively realize that the commonly cited "reasons why LLM code is bad" and "reasons poorly architectured, poorly planned, poorly tested code is bad" are a circle on a Venn diagram, the sooner we can hopefully make some effort to escape this sports fan tribalistic nonsense.
All of this noise and velocity will provide significant cover for the Jia Tans of the world.
> I did not just vibe-code “convert test suite to python”. I’m a software engineer with 40 years experience (yeah, I’m OLD!)
Interestingly enough, I see the trend of people with decades of experience using AI more and more often. I'm in the 20 years club myself, and I do AI-assisted coding every day. It does help, and I'm grateful to have a tool like this to quickly try new ideas and throw them away if they suck. Shaming the author for using AI to help with the CI stuff is baffling to me. Are we witnessing just another ideology-driven tribal reaction on the rise?
I really feel for the developer of Rsync. He did absolutely nothing wrong and got flamed to hell for it. I’m glad he has help now and I hope all of the negativity doesn’t get to him. I would certainly have a difficult time with my mental health if I was on the receiving end of all this hate.
It's sad that the outrage posts got hundreds of comments while this article, from the maintainer explaining the CVEs and test suites, only has this singular comment and is already on the second page.
A lie travels around the globe while the truth is putting on its shoes. Do better, HN.
Maybe because the explanation isn't so much an explanation but an admission that core functionality was broken in a minor version release - and therefore almost orthogonal to the use of AI. If there had been no major regressions, do you think anyone would have complained?
People are (correctly) not going to be held to some kind of lower standard just because they "used AI" and "were fixing security issues".
>If there had been no major regressions, do you think anyone would have complained?
If there had been the same regressions but no AI, do you think there would be a 300 post issue full of people complaining? People are holding the update to a higher standard just because they used AI.
I'm showing that it was only published an hour ago.
5am here in the UK, midnight Eastern US on a Tuesday night? I can see why it wouldn't have gained much traction yet.
> A lie travels around the globe while the truth is putting on its shoes. Do better, HN.
Stop baiting.
Probably best to spell that “baiting”
Thanks! :D
We'll see if you're right, and if HNers can redeem themselves. I'm not holding out much hope, though.
Related:
Please Do Not Vibe Fuck Up This Software
https://news.ycombinator.com/item?id=48342705
I think he should just retire for real (from rsync) and enjoy the time, who would take all the negativity and keep going for a volunteer project like rsync.
And this has nothing to do with AI. It seems to me that a significant amount of people in IT disconnected from reality.
Go throw a couple potatoes in the ground and see them grow or whatever. Your opinion most likely wont matter nor make any difference.