Sylvestre Ledru who serves as the lead developer of the uutils project for the Rust Coreutils implementation presented at FOSDEM 2026 this weekend on this initiative. Ledru has spoken at FOSDEM in prior years on Rust Coreutils and this year’s talk focused primarily on Ubuntu 25.10’s adoption of it in place of GNU Coreutils.
Ledru’s presentation covered the progress made on Rust Coreutils in recent times and Ubuntu 25.10’s uptake of Rust Coreutils and continuing that for Ubuntu 26.04 LTS. While some bugs have been found as a result of it, they have been fixed rather quickly. Ledru’s presentation also points out some of the popular trolling around Rust Coreutils and ultimately how many of those commenters have been proven wrong
Lol, very first pair of comments. I love phoronix sometimes.
Ah, the duality of man…
Volta raging over any rust post, a classic XD
Replace a perfectly usable GPL software for MIT? Nope. I used to fall for that ten years ago. The social infrastructure of software is more important than the exact tech used. The license is fundamental to that.
What did you replace Xorg with 10 years ago?
Not the person you asked but kwin the GPL licensed Wayland compositor. Why do you ask?
Nothing. Still use it to present day. I generally trust the selection of OS software to the distro developers - Debian and Ubuntu for me. We constantly use software that doesn’t perfecty fit our requirements or wants.
I wasn’t aware that coreutils was going somewhere.
We like the Rust, we hate the cuck license. Simple.
I don’t understand what’s going on with the rust community insisting on cuck licences. Do they love writing on their Mac books so much?
I think a part of it may be that they are from the younger generation like myself, and most of them don’t really know the history of software and FOSS, and MIT is just a safe option for them. I think they haven’t really put in the time to read and undertstand the philosophy and logic behind FOSS and read the licenses and writings.
But why MIT? How come that became the default? Why not GPL? Is Microslop Github suggesting MIT by default?
I’m not sure but theres something in my mind bout MIT being the first suggested license for github.
and also, to be real, you need to do some reasearch and actually understand the GPL license if you want to use it for your project. But with MIT you can just slap it on there and forget. It’s convenient, but like a lot of conveniences, can be very bad.
Why do you need to do research for GPL? It’s the OG opensource license AFAIK that forces users to also opensource their stuff. MIT let’s anybody close source your code and make money with it.
GPL isn’t perfect as it doesn’t solve the funding problem, but MIT is about the worst thing one can do for opensource: do the work for companies, for free, and be OK with never contributing back to the opensource ecosystem.
The OG open-source license was when software was just shared as a convenience, as companies only sold the matching hardware. When AT&T started asking for license fees for UNIX, it all went downhill.
The compatibility stuff really do need understanding.
You can’t use all other projects in your GPL project and your GPL project cant be used within every project.
permissive licenses really do have no compatibility since they forbid nothing and allow everything (which isnt good most of the time) so tgey are simpler to use.
Now that you mention it, it is kinda strange. MIT has been abused before.
as well as Apache
True. I hope rust community gets more mature and uses some good licenses.
You’re a rube if you think corporations can’t throw some money at interns do a rewrite in MIT and bypass GPL that way.
So let them do that. Why should we be doing their dirty work for them?
Unless we’re stupid.
Hmm sorry for calling you a Rube. You make a fair point.
Average Twitter post
What? you expected me to double down? I am trying to make lemmy better than reddit, not worse.
@Maddier1993 XD
So are you saying that the developers should abandon the project if they do not use a license that you like?
It’s not about any of us enjoying the license*, it’s about preserving the integrity of free software. It’s both flattering and disturbing that core utils is popular enough that people have decided to give them away to anyone who would want to take them without ever contributing back. If those people are found out there will be no legal recourse. Those Rust rewrites would inevitably be made proprietary without any credit for the authors.
First GNU coreutils is going to remain GPL-licensed, so nothing that already exists is being given away; the only thing that is happening is that some people have decided to write brand new code. (And it is worth noting that GPL only says that if you share the binaries, you have to share the source code; if your changes are only used internally, you do not have to contribute them back, though you probably want to do so since it makes your life easier down the road when you want to use newer version of upstream.)
Second, what scenario exactly is it that you are worried about? I want a specific and plausible answer, not just vague allusions.
Finally, if the Rust authors are fine with the possibility that someone will use their code in this way, then who are we to tell them to stop their development when we can continue to use GNU coreutils?
You did not answer my question, and I think it is an important one so I will repeat it: should they abandon the project if they are unwilling to use your preferred license?
I wrote this before I realized that you don’t actually care about the answer, you just want people to shut up about it, so sorry. If you want somebody to do the work you’ll have to do it yourself now. You’ve been given plenty of examples in this thread already
I have not been given a specific example of a scenario involving uutils, I have only been told about scenarios for unrelated and very different projects, and the difference between the situations is significant enough that you can’t simply point to them and declare that your point has automatically been proven. In fact, I would argue that uutils is such a different case that it is implausible that such a scenario could occur and become a big problem.
And yes, people stopping complaining about work being done on a project they are not involved with in every single post discussing it would be a perfectly fine outcome for me. But if they are not going to do that, then I would also be happy with getting my questions answered because I believe that they are relevant.
I think the biggest point you may be missing here is if you start re-writing GNU/Linux (which is what uutils is the first step in doing) with an MIT license, you start making reasons for commercial entities that contribute back out of obligation to stop supporting upstream free software. This is a no brainer to me. As to whether or not anybody should stop writing uutils, the answer is **obviously not. ** The license, however, is free game for scrutiny
Okay, but is this group trying to re-write all of the GPL software in the Linux ecosystem with an MIT license? I ask this because I think that the words “first step” are doing a lot of the lifting in your argument.
And just to be clear, my objection is not to people disagreeing with the license; in fact, as I have said elsewhere–though I hardly expect you to have read all of my comments here!–I think that the underlying criticism is actually reasonable, I just also think that the extent of the concern is exaggerated in practice in this specific case (which is why I keep trying to pin people down on specifics rather than generalities). Again, my objection is that people feel the need to post the same inane comments with varying degree of toxicity (such referring to them as using a “cuck” license) complaining about it in every single post.
Your ignorance is annoying.
It is telling that you cannot tell the difference between having a difference of opinion and ignorance.
They never should’ve gone with a weak license to begin with.
Whoever is suggesting and perpetuating MIT over GPL needs to be tarred and feathered.
So now you are going to start tarring and feathering anyone who decides to start a project using the MIT license, or at least engage in an equivalent level of verbal violence at every possible opportunity?
If they could just use a real licence and even more copyleft (at least something, like EUPL, MPL or GPLv2)
The licence would be significantly better. And would drive a bit more adoption.
So you are saying that the quality of the code and the functionality that it offers would be significantly improved?
(It’s not clear how much more adoption there would be, though a bit more is plausible.)
More people would inclined to contribute or include it in there own projects if it wasn’t a regression in terms of FOSS.
I.e. why contribute to this project that could be forked to create tools that don’t respect the users when the main existing project doesn’t have that flaw?
Given that the Rust community seems to prefer more permissive licenses, I doubt that there many people who would be interested in contributing that be put off by this in practice.
Unless you are telling me that you personally had been really motivated to contribute to this project yourself, but changed your mind when your leaned about the license?
I personally yes. Also I’ve seen hesitation in including them in distros because of it
Fair enough, but in that case, them using a less optimal license is a problem that will solve itself because it won’t be used, so it is not something that needs to be brought up by multiple people in every post on uutils.
People get excited about it, post about it, and disappointed by the reality of the license in constant cycle.
It’s kind of the “missing stair” problem. Those who know, can go around it, but knew people need to be informed that it’s busted to avoid it until an actual solution is presented.
I dont understand.
You are the one saying that the project would be significantly better. I am asking you to translate that significant claim into a set of metrics.
I really appreciate your insightful point as no one else has ever pointed this out before.
It is trolling when it broke production level systems?
To be fair im NOT blaming the rust util team. I hope the best for them. But it was a bad decision to use something like that to power systems before it was fully tested and ready. It broke many different things in prod at work and we had to switch over to another distro entirely. Which was a lot of work. It made us stop using Ubuntu which is a shame.
Your first mistake was using Ubuntu on a production server. Canonical has made more than enough questionable decisions over the past decade that using Ubuntu for a production system should be a red flag.
Your probably right. It was an old setup but ill own it. I inherited it (mitus touch) so I probably should have put more effort into switching.
Switching can be hard sonetimes. At work we still have a few Ubuntu 20.04 LTS machines that need to be replaced.
Oh yeah! We just switched over to Debian and all good. But theres just all those little tools you never think about haha.
20.04 was nice back in the day. Super stable.
It is trolling when it broke production level systems?
Depends. Were they the ones who put it into production level systems? If the answer to that question is no, then, well, you have your answer already.
https://ubuntuhandbook.org/index.php/2025/10/ubuntu-announced-fix-2510-updates/
This specifically broke things at work and is why we moved off.
Right, so the answer to your question is that it is trolling the uutils devs because Ubuntu was the one that decided to make the switch.
i mean, how many realistically? how many systems are out there using non-LTS releases that would actually run into these edge cases? and auto-updating them in production without triggering the bug first? or maybe i’m a naive corpo
Honestly it was a bunch of docker containers that failed all around the same time. Kinda sucks. Again I blame more Ubuntu support (since we pay for it) than rust or rust utils.I hope to eventually switch all systems to using the library when they hit 100%. Its going to be so fast :)
What broke? If it was a GNU ism that wouldn’t work on *BSD either, than it is your own stupid fault. There are other linux distros that also don’t use the gnu core utils that would break things do.
sudo-rs is a different project than uutils
You know what that is fair. While the change did temp kill our projects it was a combo of things.
Going from GPL to a weaker license was a terrible idea and whoever supported it should be held accountable.
GNU coreutils is not going anywhere or having its license be weakened.
But anyway, I am glad that you are brave enough to be willing to personally punish people for using a license that you do not prefer.
Rust Coreutils Continues Working Toward 100% GNU Compatibility, Proving Trolls Wrong
98 comments
Phoronix, you are the trolls.
Lol, phoronix forums being on point
So it still doesn’t work and there’s no timeline for when it will.
Got it.
That tends to be true of a lot of projects before they are finished.
The cult of Rust developers just won’t give up.
Nor will the cult of Rust haters.
They are coming to steal your IP
I don’t hate Rust. I just don’t like everyone that think it’s the solution to everything. My brother is on the Rust committee, I get it.
So you do not like the uutils developers because they think that Rust is the solution to everything?
I am slightly biased because I spent over a decade writing C99 code that had to have 100% uptime. My point was just that it was written in Rust doesn’t make it better. Carefully coded and audited code in any language is good.
Right, and likewise just because your code was written in C99 does not make it any better than assembly code that accomplished the same task, as long as it was written carefully and audited.
or G’d forbid Perl. Oh i can think of something worse, PHP.
Call to action: calling all toppers
Thank you for introducing needed some levity into the discussion. 😉
The individual is a troll. Its best to move on. Either that or an LLM that needs to have the last word no matter what.
I guarantee this will have a comment 😄
I guarantee this will have a comment 😄
You do know that, when you edit your comment to pretend that you had predicted something in advance, that there is a timestamp showing when you made the change, right?
It’s interesting that there are comments that say things like:
The cult [emphasis mine] of Rust developers just won’t give up.
We like the Rust, we hate the cuck [emphasis mine] license. Simple.
Going from GPL to a weaker license was a terrible idea and whoever supported it should be held accountable. [emphasis mine]
Whoever is suggesting and perpetuating MIT over GPL needs to be tarred and feathered [emphasis mine].
And yet you have decided that it is my comments calling this out as being the trollish ones, as if I were the one being unreasonable.
they are coming for our daughters
😁
