Thread

OpenBSD Developers Reject Uutils Coreutils

31 points · 41 comments · pjmlp

Loading the complete thread in the background. This saved snapshot is available now. Refresh

airhangerf15 · 1791101000 · focus · HN ↗
This reads like garbage and Lunduke already did better coverage
1. klez · 1791108933 · focus · HN ↗
  Agreed. It's misattributing quotes and talking about OpenBSD releases from more than a year ago as if they were upcoming releases. Something's seriously wrong with this article.
2. holsta · 1791136572 · focus · HN ↗
  <a href="https://drewdevault.com/weird-guys/#:~:text=2026-,Bryan%20Lunduke" rel="nofollow">https://drewdevault.com/weird-guys/#:~:text=2026-,Bryan%20Lu...
  1. bbg2401 · 1791141904 · focus · HN ↗
    Drew is hardly one we should listen to on matters spanning weirdness. He’s a total creep.
bel8 · 1791107519 · focus · HN ↗
Discussion nhere:
<a href="https://www.mail-archive.com/ports@openbsd.org/msg143875.html" rel="nofollow">https://www.mail-archive.com/ports@openbsd.org/msg143875.htm...
Of note is Theo's answer which follows below:
---
```
    Theo de Raadt Sun, 20 Sep 2026 12:03:20 -0700
    
    David Uhden Collado &lt;daviduh...@gmail.com&gt; wrote:
    
    &gt; Stuart Henderson wrote:
    &gt; &gt; On 2026&#x2F;09&#x2F;20 07:01, David Uhden Collado wrote:
    &gt; &gt;&gt; The main goal of the packaging is to make these implementations usable
    &gt; &gt;&gt; as alternatives to the existing GNU utility ports without requiring
    &gt; &gt;&gt; source changes in dependent ports.
    &gt; &gt;&gt;
    &gt; &gt;&gt; For example, uutils-coreutils installs the same g-prefixed command names
    &gt; &gt;&gt; as sysutils&#x2F;coreutils, including gcat, gls, gcp, gdate, gsort, gstat,
    &gt; &gt;&gt; gtail, gtimeout and the other GNU-compatible utilities. They are
    &gt; &gt;&gt; symlinks to the upstream multicall binary, which is installed under
    &gt; &gt;&gt; libexec&#x2F;uutils.
    &gt; &gt; ...
    &gt; &gt;&gt; Each package conflicts with its corresponding GNU implementation and
    &gt; &gt;&gt; declares the GNU port as a secondary @pkgpath.
    &gt; &gt; I don&#x27;t think this is a usable approach for ports.
    &gt; 
    &gt; The truth is, I find these Rust reimplementations quite
    &gt; interesting. Ubuntu 26.10 has already adopted uutils coreutils because
    &gt; the project has reached a level of maturity and stability where it can
    &gt; be used reliably. The other reimplementations are still more of a work
    &gt; in progress.
    
    Smells like agenda.
    
    &gt; I also think they fit quite well with OpenBSD as alternatives to GNU
    &gt; utilities, particularly because they use a permissive MIT license.
    
    Argument is vaguely like: because we already have permissive licenced
    utilities, our user base are really interested in having a second set of
    permissive licenced utilities which are very subtly different.
    
    That makes no sense. Noone wants subtly different behaving binaries as
    part of their workflow.  If someone runs the openbsd ls command as part
    of a pipeline that uses openbsd sed, or openbsd cut, or some other
    openbsd utility and it parses a non-standized output characteristic
    by accident, there are no people in this universe who wants to replace
    that ls with a different ls and get surprised by un-standardized tooling
    behaviour clash.
    
    &gt; I&#x27;m not sure yet whether it&#x27;s possible to install the individual
    &gt; utilities as separate binaries. This is new territory for me, since
    &gt; uutils is structured as a metapackage, and because it&#x27;s written in
    &gt; Rust.
    
    Oh, because it is written in Rust.
    Your agenda is showing.
```
---
1. broodbucket · 1791108283 · focus · HN ↗
  Bit hostile. I get the point that it's unnecessary, but the whole "agenda" thing feels kneejerk. Just say that being written in Rust isn't a good enough reason for adoption.
  1. Matl · 1791108478 · focus · HN ↗
    Theo de Raadt is like that, everything he touches is gold, everything someone else comes up with is garbage, unless he forks it.
    uecker · 1791109566 · focus · HN ↗
    Perhaps, uutils is garbage though, pushed by some only for rather bad reasons (Rust and license). I worry more about projects where it is not directly laughed out of the room.
  2. hnlmorg · 1791109048 · focus · HN ↗
    Yeah, that’s Theo de Raadt for you.
    In fact this feels remarkably tame by his standards.
    The guy is undoubtedly an exceptional talent. But his communication skills isn’t one of them.
    BSDobelix · 1791109890 · focus · HN ↗
    >But his communication skills isn’t one of them.
    It's the projects shield, and pretty good one at that ;)
    hnlmorg · 1791110092 · focus · HN ↗
    You don’t need to be rude to be a project shield.
    Linux went through this evolution with Linus Torvalds, and plenty of other projects have managed to have strong leadership without the hostility.
    BSDobelix · 1791110810 · focus · HN ↗
    >You don’t need to be rude to be a project shield.
    But you can, it's also a really good way to shy away "over-sensible" individuals, if you think that was rude.....well ;)
    >Linux went through this evolution with Linus Torvalds
    Not he's still the old one, just without "retro abortion"
    hnlmorg · 1791112513 · focus · HN ↗
    > But you can, it's also a really good way to shy away "over-sensible" individuals,
    You say that like it’s a good thing but it’s really not. All you’re doing is poisoning your own well. Shitting your own bed. And other analogies that emphasise an own goal.
    The last thing a project with limited broader appeal, and that appeal diminishing with the homogenisation of POSIX platforms, is someone trying their hardest to make their community even less appealing to others.
    > if you think that was rude.....well ;)
    Id already said that comment was tame. What I’m saying is that Theo is more generally rude.
    BSDobelix · 1791117936 · focus · HN ↗
    OpenBSD is still here, with Theo.
    Letting toxic Snowflakes into your Project is "Shitting you Bed".
    >What I’m saying is that Theo is more generally rude.
    Nice, that means if he's doing stupid stuff i can be rude too without having to worry that a CoC or even worse a HR department threatens me to resign or publicly state what a terrible person i am.
    hnlmorg · 1791125724 · focus · HN ↗
    > Letting toxic one-sided-Snowflakes into your Project is "Shitting you Bed".
    Wow. Just wow.
    You do realise that people are complex and are unlikely to agree with Theo on everything. Even yourself. That doesn’t mean they’re toxic nor one-sided.
    You you do realise that my original point was comments like yours put off even people who aren’t one sided nor snowflakes. Because actually most people don’t want to be around conflict.
    Thus the only people you end up attract is fellow toxic personalities. Which is probably why you feel it’s acceptable to use such language in defence of Theo.
    You’re lucky that I happen to know some awesome people who are OpenBSD advocates, because otherwise comments like yours would have pushed me away from the OpenBSD community decades ago.
    BSDobelix · 1791134275 · focus · HN ↗
    >You do realise that people are complex and are unlikely to agree with Theo on everything?
    No problem go somewhere else, or bring a good point, no one said you have to agree with Theo on everything.
    >Because actually most people don’t want to be around conflict.
    That's not a conflict, you are simply to soft and think everything is an attack on you.
    >And believe me when I say I’ve written and tested a lot of software for OpenBSD over the last 2 decades.
    Ehh that's ok i guess?
    >So this is why I say there’s very real risk of rudeness costing projects valuable contributions.
    Sometimes risks have to be taken, it's also funny that it's always people who are "potential" contributors but are in fact never the ones who contribute anything.
    hnlmorg · 1791142720 · focus · HN ↗
    > That's not a conflict, you are simply to soft and think everything is an attack on you.
    You and I both know this was never about me, and thus I never thought any of our conversation was a personal attack. But nice try on the trolling.
    > it's also funny that it's always people who are "potential" contributors but are in fact never the ones who contribute anything.
    Everyone’s a potential contributor until either you scare them off or they hang about and contribute. There’s nothing funny about that. It’s just what those words literally mean.
    ———
    Anyway, I won’t bother replying to you again. As I’ve said before, you’re just proving my point about Theo. So I feel we’ve reached the conversations natural conclusion.
    Sorry to disappoint you though. Hopefully (for you) your next victim gets their feelings hurt. ;)
    sdcfgy · 1791109993 · focus · HN ↗
    I disagree. It takes a different personality and position to maintain a project like this. If you aren’t the biggest asshole then you’re at the mercy of the biggest asshole, which probably means you’re going to end up making ideological decisions rather than engineering decisions. Especially seeing has how politicised everything is these days.
    Thus has exactly the communication style that is required to support the project.
    And yes I have been on the end of a Theo quip. It’s a badge of honour because ultimately he was right.
    hnlmorg · 1791110367 · focus · HN ↗
    I agree that strong leadership is a requirement. However there’s a huge difference between being strong against pushy arseholes, and preemptively being an arsehole because people make allowances for your behaviour to others.
    This just comes down to effective communication skills. And is evident by the number of other large projects survive without the rudeness.
    Theo isnt an example that deserves emulation and admiration.
    sdcfgy · 1791111001 · focus · HN ↗
    I’m not sure I agree. My own management experience in engineering and software has shown that strong leadership means saying no quite a lot. Ideas come with best intent mostly but dangerous realities. A soft no is a keep trying to most engineering folk. Shutting something down needs a Theo.
    hnlmorg · 1791111598 · focus · HN ↗
    You’re talking as Theo is the only form of strong management. As someone who is also an engineering manager, I can tell you that there are ways of saying “no” without being an arsehole.
    I do agree that sometimes you just need to end the discussion. But that doesn’t mean that arsehole comments should be the default character trait.
    sdcfgy · 1791111954 · focus · HN ↗
    Theo isn’t always an asshole.
    hnlmorg · 1791112079 · focus · HN ↗
    But he is more often than he needs to be.
    Or at least that’s our point of contention.
2. sdcfgy · 1791108832 · focus · HN ↗
  I couldn’t agree more with Theo on this one. I suspect a lot of people interested in doing this stuff haven’t had to maintain anything for a long time (10 years+). This to me feels like the painful switch between SunOS and Solaris we had to do back in the day. Lots of very small differences. And you have to maintain them forever if you have dual stacks of tools.
  Not only that the rust ecosystem imports a HUGE difficult to control surface area.
  Always feels like a vanity project when I see ideological rewrites. Also didn’t they introduce a ton more CVEs doing this on the Ubuntu side.
  1. BSDobelix · 1791110497 · focus · HN ↗
    44 CVE's : <a href="https://corrode.dev/blog/bugs-rust-wont-catch/" rel="nofollow">https://corrode.dev/blog/bugs-rust-wont-catch/
    Also stuff like: "Set Permissions at Creation Time, Not After"
    sdcfgy · 1791110791 · focus · HN ↗
    What a shit show.
  2. JdeBP · 1791111109 · focus · HN ↗
    Interestingly, the mailing list discussion being talked about happened roughly a week after this Hacker News thread highlighted afresh all of the Ubuntu LaunchPad bugs being left hanging; and highlighted a major problem with rm that, probably not coincidentally, suddenly saw fresh attention after being around for years, the same day as the Hacker News entry appeared.
    * <a href="https://news.ycombinator.com/item?id=49697392">https://news.ycombinator.com/item?id=49697392
    sdcfgy · 1791111279 · focus · HN ↗
    I completely missed that thread. Just the first post is horrifying.
    About a year ago I switched from Debian to FreeBSD because even that was getting wonky. Glad I’m far far away from this crap.
3. nasso_dev · 1791109044 · focus · HN ↗
  the conversation here makes way more sense than the quotes from TFA
  whoever wrote/generated the article does not seem to understand email syntax
4. ninjin · 1791115622 · focus · HN ↗
  [delayed]
bewareofscams · 1791108675 · focus · HN ↗
[dead]
1. ninjin · 1791109753 · focus · HN ↗
  [delayed]
  1. JdeBP · 1791110470 · focus · HN ↗
    Of course, on OpenBSD it is GNU Core Utilities themselves that already are the alternative suite of SUS utilities. So it's an argument for carrying alternatives to the alternatives.
2. sdcfgy · 1791110174 · focus · HN ↗
  On paper yes but it introduces different problems. The trade off isn’t one that everyone universally believes is worth it.
3. yjftsjthsd-h · 1791147212 · focus · HN ↗
  Couple things:
  Comparing uutil's CVEs per year against GNU coreutils paints a very poor picture. I haven't checked OpenBSD coreutils, but I suspect it's even more ahead. Rust doesn't appear to be a silver bullet.
  Even if rust uutils was more secure, I don't think it's available on all OpenBSD platforms, which really caps how useful it can be.
hn97833 · 1791109104 · focus · HN ↗
I wonder might this be partly due to licensing concerns given uutils have acknowledged they've copied GNU GPL code. From their latest release notes (<a href="https://github.com/uutils/coreutils/releases/tag/0.12.0" rel="nofollow">https://github.com/uutils/coreutils/releases/tag/0.12.0):
"Also, some contributions had carried comments and code over from GNU coreutils, against our own rule that uutils is an original implementation. We reviewed the tree and rewrote what we found, in comments and in the few functions that followed the C too closely"
1. jmclnx · 1791126613 · focus · HN ↗
  The other issue no one brings up us rust is not available on all Platforms OpenBSD supports. Plus on i386 (32 bit), rust based applications cannot compile due to memory issues, for example Firefox.
ninjin · 1791109310 · focus · HN ↗
What is this rubbish? We have "media" coverage trying to stir up drama about what happens on mailing lists now?
Go and read the thread yourself [1] and you will see that the reason it was rejected was that no one saw a reason to take on the maintenance burden when you already need to have GNU Core Utils in ports to be able to build a heap of other ports. Focusing on theo@ calling out the submitter for having their judgement clouded by "Because Rust" misses the point.
[1]: <a href="https://www.mail-archive.com/ports@openbsd.org/msg143875.html" rel="nofollow">https://www.mail-archive.com/ports@openbsd.org/msg143875.htm...
Besides, Rust already has a solid presence in ports and I know several experienced OpenBSD developers dabbling in it. But the language will never make it into base in its current form as OpenBSD expects self hosting. I think everything in base is Posix(-like), apart from Perl.
1. pseudohadamard · 1791183479 · focus · HN ↗
  There's also a pile of odd things about the article:
  > OpenBSD 7.6 shipped in October 2024; 7.7 is expected in April 2025.
  > As of 2024, it passes the majority of GNU coreutils test suites
  This is for an article dated September 2026. It's not obviously AI-written but the dates are completely out of sync, it's like a lot of it was written in 2024 and then other bits were added a week ago. A quick Google of the author's name also doesn't turn up anything that could be her, so possibly it is just AI slop.
jmclnx · 1791127919 · focus · HN ↗
>7.7 is expected in April 2025
A bit confusing since the Article it self is dated Sep 30, 2026. Almost makes me thing it is partially LLM.
duendefm · 1791135878 · focus · HN ↗
F* Rust. If AI finds every bug then Rust choice is pointless for those who still care about programming. Real programmers that like to code have no reason to lose time with Rust. Let's leave Rust for vibe coders.
1. bananaquant · 1791148915 · focus · HN ↗
  Real programmers write assembly code, using a magnetized needle and a steady hand.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.