The whole UNIX certified thing is a bit odd to start with, but those addendums are really just saying that because the UNIX compliance tests are thoroughly stuck in the past, a bunch of security measures (that users expect to exist on modern systems) have to be disabled to run the test suite.
SIP is not some irrelevant detail, nor is it an obviously good "modern security practice". A system that the hardware owner cannot modify is not really on the same wavelength as traditional UNIX.
The hardware owner can modify it all they want, disable SIP. For a large swath of average and non-average users that will not be modifying the closed-source OS they're running, it is obviously good.
Traditional UNIX is does not mean having an open source core you're intended to tinker with. Traditional UNIX was entirely proprietary, connected with well-defined text-based interfaces. This is entirely orthogonal to user control and manipulation of the OS.
Traditional UNIX was entirely proprietary, with source you're intended to tinker with.
AT&T was forbidden by its consent decree from productizing Unix. Licensees received the source code, on an as-is basis, with no official support from AT&T.
Sure, if by “you” you mean universities, Sun, DEC, IBM, etc.
As a non corporation end user with a sublicense from one of the above licensees (which is the apt analogy for an individual user in my opinion) you most certainly did not receive the source or have the rights to it.
I regard "traditional Unix" as the 1970s era. There were no sublicensees. You bought the source code from AT&T. That was the only option.
You then likely hacked the source code to make it work on your machine.
I can understand that others might include the commercialization of the 1980s and the Unix wars as part of "traditional Unix", but even then, my IRIX system ca. 1990 supported end-user kernel extensions, to handle new types of devices. Our imaging hardware was not "connected with well-defined text-based interfaces".
Sure, but that just means UNIX vendors had access to the source code.
It doesn't mean end users had access to the source code.
In fact, vendors would have been prohibited from distributing AT&T source code to customers who weren't themselves AT&T licensees, just as CSRG was prohibited from distributing BSD to customers who weren't licensees before 4.4BSD-Lite.
The same goes for most other proprietary source-available products. For example, game developers can't open source Unreal Engine projects even though Unreal Engine source licenses are available to end users for free.
You would absolutely normally get source to the vast majority of the unix components with a unix install.
At a bare minimum, you would almost have to have source to the vast majority of the kernel because there was no kernel modules. Sysgenning for a particular system involved compiling the kernel for that set of hardware.
I'm literally working on an early sun system emulator right now, heavily assisted by the standard sec tape that came with the standard software distribution of sunos, which has plenty of AT&T code.
As jeremyjh wrote at https://news.ycombinator.com/item?id=50032733 , "A system that the hardware owner cannot modify is not really on the same wavelength as traditional UNIX.". Since traditional Unix was source-available, with the expectation that people would hack on the source, I fully agree with jeremyjh's characterization.
I think you are talking about a later era in Unix history? The early Unix licensees were educational institutions and commercial sites, not vendors or resellers.
Every one of those academic and commercial end users got the source code.
> "In 1973, AT&T released Version 5 Unix and licensed it to educational institutions, and licensed 1975's Version 6 to companies for the first time, ... Anyone could purchase a license, but the terms were very restrictive; licensees only received the source code, on an as-is basis. ... Unix was entirely unsupported by AT&T"
From the same link, "Perkin-Elmer began distributing Wollongong Unix for Interdata as Perkin-Elmer Edition 7, the first version supported by a computer company". That was in 1980.
So while yes, "Traditional UNIX was entirely proprietary" and could not be redistributed to non-licensees, it was also provided with "full user control and manipulation of the OS", and not merely "connected with well-defined text-based interfaces", quoting jdiff at https://news.ycombinator.com/item?id=50033080.
Also worth keeping in mind that macOS is intended for use with commercial third party software that’s repeatedly been proven untrustworthy, doing things like prodding system internals and installing dodgy kernel extensions where neither is warranted.
SIP is as much defense from the likes of Adobe and Google as it is from more traditional malware.
For now SIP can be disabled (on mac, not of course on iOS that is completely closed) but I wonder for how much. It's clear that the desire of Apple is to make macOS a closed system like iOS, installing third party software outside the AppStore is more and more difficult, with Gatekeeper that asks you to enter the system settings and warn about "potential insecure stuff" each time you want to run an application downloaded from the internet not signed by Apple (yes I know that if you disable SIP you can get rid of Gatekeeper as well).
I think that piece by piece Apple wants to go into direction of making macOS just a desktop version of iOS, that is closed and impossible to modify. They will use of course security as an excuse, the practical thing is that with iOS they succeeded because they went to the market with an OS that was already closed, but making users accept that in new version downloading a software from the internet and installing it it's no longer possible takes time. But they are getting there soon to me.
"""POSIXLY_CORRECT (originally proposed as POSIX_ME_HARDER) is a historical environment variable used in GNU utilities to force strict compliance with the POSIX standard."""
the netbsd gallery (https://netbsd.org/gallery/) had a picture of somebody using that on windows xp (iirc) in order to compile pkgsrc packages in a "posix compliant" environment. the picture seem to be gone now :(
This comment made me feel awkward for two reasons.
First, Linux was never GNU - the GNU ecosystem has always been a significant part, but if you were obstinate enough, you could get away without using any GNU tools (which is even easier today).
Second, you give just two examples. A few years ago we had a trend to rewrite everything in Rust with these newer tools usually not being part of the GNU project. Sure, you can replace everything now. However, most GNU tools are here to stay for a long time especially in enterprise distros where backward compatibility is important.
So no, Linux has never been GNU, and whatever fluctuations we had over decades with innumerable fights like egcs vs gcc etc., the GNU ecosystem is and will long be part of Linux distros.
Over time I see fewer and fewer gnu tools in my standard mainstream Linux. The tooling I used 25 years ago was far more gnu based than today, and crucially the FSF view of what free software is, and by extension the gnu public license, seems to be constantly retreating.
25 years ago it was reasonable to say the typical Linux machine was “gnu/linux”. Today it’s not.
The relevant text from the article, (the rest of this comment is a direct quote from the article):
So, if you want your installation of macOS 15.0 to pass the UNIX® 03 certification test suites, you need to disable System Integrity Protection, enable the root account, enable core file generation, disable timeout coalescing, mount any APFS partitions with the strictatime option, format your APFS partitions case-sensitive (by default, APFS is case-insensitive, so you’ll need to reinstall), disable Spotlight, copy the binaries uucp, uuname, uustat, and uux from /usr/bin to /usr/local/bin and the binaries uucico and uuxqt from /usr/sbin to /usr/local/bin, set the setuid bit on all of these binaries, add /usr/local/bin to your PATH before /usr/bin and /usr/sbin, enable the uucp service, and handle the mystery issues listed in the four Temporary Waivers.
Then, and only then, is your macOS 15.0 actually UNIX® 03-certified.
This is batshit insane. I can guarantee you with 100% certainly not a single macOS installation in the entire history of macOS – let alone when just counting macOS 15.0 – has implemented even half of these changes. I’m sure there is a small number of people who have System Integrity Protection disabled permanently, and an even smaller number of people who have enabled the root account, and an even smaller number of people who have done both of those things – but that’s it. All the other changes are far too obscure and specific to be of any use to anyone.
I had no idea there was a built-in way to make the filesystem case sensitive. My employer probably does not want me to spend time reinstalling my OS for this now but maybe next time I'm given a MacBook for a job I should do this...
Curiously, the iOS filesystem is case-sensitive. A bit of a trap for developers on macOS. I’d expect plenty of third-party stuff to break switching the system drive to case-sensitive.
You think that you do, like I thought that I did, but I doubt that you want it in reality. Last I checked, which was probably under a decade ago(?), there were a bunch of things on macOS that misbehave on a case-sensitive FS, both first- and third-party. I reinstalled same-day.
That’s a bit of a weird article and I’m not 100% sure what its point is.
> This is batshit insane. I can guarantee you with 100% certainly not a single macOS installation in the entire history of macOS – let alone when just counting macOS 15.0 – has implemented even half of these changes.
Well, maybe that means it doesn’t really matter and especially not to the extent that the author seems (?) to be making it out to be. If no one disables System Integrity Protection in order to be UNIX® 03-compliant, maybe that means that to the extent that people care about macOS being “a UNIX”, they don’t care specifically about the parts of the specification that are incompatible with System Integrity Protection.
The explanation I remember seeing many years ago, no idea how true it is, was that this was entirely about purchasing.
There were (are?) companies and especially parts of the government where it was pretty difficult/impossible to purchase anything that wasn’t on an official pre-approved list. Trying to get anything new added was a massive problem.
One of the things that was on the list was something like “a computer running Unix“. Tada. Macs now run Unix so you can easily buy them.
Like other people here, I really don’t think true compliance mattered to basically anyone for their actual use. The idea that it was just box checking so they were easier to buy in some institutional environments makes much more sense to me.
I can see that. Before linux, there were lots of versions of unix from all kinds of vendors.
I'll bet all kinds of government contracts needed some way to compare vendors bidding on a contract
for example:
- These ran UTX-32, Gould's version of Unix based on a BSD 4.2 kernel developed by Gould's UNIX development centre in Urbana-Champaign (IL) to support multiprocessor systems
- AIX is based on UNIX System V with 4.3BSD-compatible extensions.
- DG/UX 1.00, released by Data General in March 1985, was based on UNIX System V Release 2 with additions from 4.1BSD.
- HP-UX (from "Hewlett Packard Unix") is a discontinued proprietary implementation of the Unix operating system developed by Hewlett Packard Enterprise
- Xenix is based on the operating system UNIX System V, which Microsoft licensed from AT&T Corporation in the late 1970s.
Indeed. And, weirdly, there are items that show up by that 03 standard alongside macOS 26 (e.g. "HP-UX 11i V3 Release B.11.31") that also show on the front page. Is it a latest version kind of thing?
Yea — said as much in another comment. That looks like it could be the rule — for example your HP-UX version there appears to be the latest, final HP-UX.
I swear I've seen these "[Latest macOS release] does not appear in UNIX certification site" threads every year since the release of Big Sur in 2020.
It takes several months for the certification to appear on the site. What would be newsworthy is if we see a release next year without this release getting certification.
Apple originally certified Mac OS X as a UNIX because they used the word UNIX all over their website and the Open Group came knocking; they decided it was cheaper/better for goodwill to do the conformance and pay the license fee than to fight the case in court.
Looking now, I see almost no uses of the word UNIX on the website except the docs archive, so I guess they’ve decided to abandon both the certification and the use of the UNIX trademark in macOS.
Maybe. But I don't remember seeing them use the word for the last decade, and the last-but-one release was certified. My first guess would be that it's too new, and hasn't passed certification yet.
It is strange now, but Apple made a big deal of marketing it as UNIX back then. Like at the front-page of apple.com. UNIX had a good reputation as solid and serious.
The reason Apple went to the trouble was not really a marketing exercise, but rather because a lot of U.S. government contacts required it, or a waver. This is the same reason versions of Windows were also UNIX certified. In neither case did this certification actually make them more "standard" in the way that the rules originally envisioned.
So Apple's use of UNIX in marketing was to make sure they never ran into that as a disqualified on a government contract. My impression is that those contracts largely no longer require that (one small measure of common sense has caught up).
This comment should be at the top. The title implies a non-issue: macOS silently disappears because the poster didn’t try to change the page's badly designed filter. The latest version of macOS probably will appear in months.
If this is an intentional move, Apple is just acknowledging reality. UNIX(tm) is no longer so relevant compared to the early days of OS X, when Linux was still in its infancy and the Unix certification gave Apple street cred among developers. Most developers these days target Linux for production. That's why Microsoft developed WSL, not WSU.
SFU 1.0 and 2.0 relied on the MKS Toolkit by MKS/PTC, and SFU 3.0 later adopted the Interix (OpenNT) subsystem, which Microsoft acquired from Softway Systems in 1999.
The original POSIX Subsystem (psxss.exe) was Cutler's team who implemented a minimal POSIX.1 subsystem server alongside Win32 and OS/2. This was done primarily to satisfy US Federal Government procurement requirements (FIPS 151-2).
True, though to Microsoft's credit, the existence of the POSIX subsystem, even as a proof-of-concept, ensured the underlying kernel design supported the development of products like Interix, just as Windows on Alpha et al. ensured NT remained reasonably portable to future platforms like ARM, unlike say OS/2 where the only non-Intel release (for PowerPC) was basically a prerelease lacking core OS features like networking.
I think they are saying that "Unlike OS/2 itself (as a competing MS and then IBM product) had less thought and design went into it as a portable system, whereas NT was designed portable ground up"
That's correct, I put them in the same bucket though, as SUA was also based on Interix. So I consider SUA more as a rebranding, not something completely different like WSL.
IIRC SUA required recompilation. Putting the motto "Windows NT is a UNIX" to life was challenging, and introduction of incompatible SUA added multiple versions maintenance to the challenge.
Linux has been effectively Unix™ for me ever since the days when Dennis Ritchie indicated that he used Linux for everything on his web page. Late 2000s at the latest, maybe early 2000s. Unix enough for Dennis should be Unix enough for anyone.
I've been using Linux for 30+ years now and have always considered it free (as in speech) Unix. For my purposes, the difference was always more legal and philosophical than practical. Of course, I was a student back then and didn't need to run the commercial software of the time.
Well he and others moved away from UNIX, created Plan 9 and Inferno.
In these efforts they also tried to create successor languages to C, Alef and Limbo (for userspace), and never cared that much about anything past C89.
He has settled with Linux because that is the winner of UNIX wars, which predated Linux wars on Distrowatch.
As for the original UNIX,
"Using Unix is the computing equivalent of listening only to music by David Cassidy"
Yes, Unix = Linux for all intents and purposes in 2026. When I started in the early 90s, there was a galaxy of Unices - SunOS/Solaris, HP-UX, AIX, and many others people people don't even remember (ConvexOS anyone?). Heck, back then if you wanted to run a particular software for your industry, you might have to buy and install a specific proprietary Unix.
Today? It's all Linux or Windows (of course, now also Web/SaaS). All these different Unices have fallen by the wayside in favor of Linux. Yes, some are still around but the mindshare has been in Linux for the last 20 years.
I suspect you could rip out a ton of autoconf rules and no one would notice.
It wasn’t the Unix subsystem, but was a subsystem… it implemented the various Linux syscalls and binary format directly and forwarded them to the NT kernel, similarly to how the various win32 syscalls call into the NT kernel. So WSL was a “peer” of the win32 subsystem in the same way SFU or SUA was. But it wasn’t implemented “via” SUA.
Then in WSL2 they just switched to using a VM, which is way more boring but probably less work to keep up with (I’m sure the Linux kernel has a lot of stuff that needs translating/implementing and it’s simpler to just use a hypervisor)
This is fine and I hope it's permanent. Unix certification never really captured what once made "OS X is a Unix" attractive or exciting to many.
What I care about, at least, are experiences of ease, convenience, compatibility, control, order, and freedom around a programmable system that invites the blurring of the lines between user and developer, between quotidian use and automation, and between decoration and reshaping. Such experiences still involve many of the same elements as they did two decades ago: a composable language of commands, commonly observed conventions, well-organized hierarchies of files, formal software freedom, powerful package management systems, great in-band/offline documentation, etc. But Unix certification isn't the center of gravity around which such experiences revolve, if it ever was. And for that matter, neither is macOS.
While it may simply acknowledge reality, I’ve already moved my main development off macOS to a NixOS workstation months ago. Too much friction in an OS that is clearly trying to be more “consumer” than “creative” at this point.
Which is a damn shame, because AI dev is going to result in a massive influx of new software right when Apple makes it ever more difficult to do so on its platform.
I still make Mac builds, but it is more of an afterthought on a secondary Mac build machine now.
You know something has peaked when they become less open and more “clawbacky”; that is basically collecting interest on an investment (marketshare) instead of remaining more open (at the cost of some things). “Increased security” is always the rationale.
APFS is mandatory now. I recently used OCLP to coerce macOS 15 "Sequoia" onto my now-elderly 2015 27" Retina iMac. It's a capable machine: 4 cores, 32GB RAM, 1TB SSD as the system drive and 8TB HDD for data.
I kept my home directory on the HFS+ HDD. So, fast OS, slower but very capacious personal drive. It was a very simple hack and everything automatically obeyed. Worked fine up to Monterey (macOS 13), the last supported version.
Not now.
Lots of apps wouldn't run and the OS couldn't save lots of settings. I had to go through the first-login wizard every single time.
In the end I moved my home directory to the system volume on the SSD, and then moved all the big folders back out to HFS+ HDD (Documents, Downloads, Pictures, Music, Videos, etc.) This took 2 days of work but it did the job. Everything works now and the machine is noticeably faster, but now, I'm using 2x more SSD space and less HDD space.
The point being that the POSIX texts require setting classic Unix values on HFS+ drives, and I am not sure all of those still work on APFS, and that will make passing harder.
I think they might do it anyway, and soon, macOS 27, on Arm64 only, might pass... but in an even more wildly nonstandard config.
I see macOS 26 listed under the Unix 03 standard. Was macOS 27 listed and then removed? The submission title + site contents combined do not give enough info to assess what exactly was silently removed.
The this is much ado about nothing. Every single entry on the page you linked goes to its product page with the exact Unix standard it is certified as. Once you select a filter, you cannot get back to unfiltered list without the back button on the browser.
This is just a site rendering issue. Rather than silently removed, it's more like some bug was introduced. macOS 26 is still Unix 03 certified.
Well Unix as a standard is completely dead. macOS as a Unix is completely dead. I expect another branch in the already complicated as hell tree of Unixes.
yeah I guess there isn't really much to move to... a new custom Kernel and Userspace? extremely unlikely - I guess it's here to stay, despite not being formally Unix.
I'm suspecting some forces at Apple would be perfectly content with ripping most of the BSD core and the assorted CLI out and just have Linux containers spawned in VMs, WSL2-style.
I haven't tried it. colima also simplifies linux VMs which is more of an analog to WSL2 than just running containers. It has really nice integration with Incus as well, if you like persistent system containers.
It's distinctly problematic there unfortunately. Particularly the moment you add any UI toolkits to it. I've had so many problems with stuff that uses Qt because Qt takes a while to catch up with things Apple broke inside their Cocoa internals. Very irritating platform to work on.
Edit: love how I get downvoted for facts?! They literally broke Cocoa so it was throwing hopeless useless exceptions at the bottom of the stack when they changed the native file open boxes to run in another process. Two bits of software I use couldn't raise an open file dialog window any more.
I've been a software engineer for 25 years, doing Linux stuff at startups and big companies (the biggest even). In all my time as a software engineer, I've seen maybe 3 people who didn't use a MacBook Pro for work. Apple makes the bulk of its money from tech companies and schools and they cater very well to both. If it stops being Unix-like, they lose the thing they make the most money from. They are smarter than that.
You still end up in uncanny valley often enough, with commands that are missing, or working unexpectedly differently. It often feels like macOS could be made so much better for developers, but Apple just doesn't care because people buy their devices anyway.
It’s a computer. I can make a computer go beep if I want, that is not the issue.
The fact that someone who didn’t grow up at the time when open source was an idealistic concept instead of free support to corporations needs to learn about the gnutils and how to install them with homebrew is what makes me sad.
Why do we seriously not have a command-line package manager on macOS yet? Windows has one by now. It’s just sad to me that you need a manual checklist of post-setup steps to turn a Mac into a useful machine for software development.
The point I was making was that knowing what GNU means in the first place probably requires some cultural background that many people lack, especially younger ones.
Right, and the point I'm making is that you can learn this from Google once you notice there is a difference from how common commands work on Linux and MacOS. You don't need to know about GNU, you just have to wonder why its different.
It’s not that. It’s that their lawyers won’t let the engineering side distribute gnu tools. Period.
At Apple, if you even use GPL tools in your workflow you have to seek approval, and most often there’s pushback of “why won’t this (alternative that isn’t GPL) work ?”
Their lawyers aren’t stupid, they understand the GPL and its limits, but they’re simply doing a risk assessment- which one is less risky, the GPL variant or the BSD. For a corporation, that’s generally BSD.
I don't think apple makes the bulk from professionals at all, the iphone is their flagship product and their laptop line has been expanding towards casual markets with budget models.
The money they make from MacOS is a footnote compared to the money they make from phone hardware, the app store transaction tax, and ads. To a first approximation, the only reason that Apple cares about MacOS is so that people can make iOS apps.
This is a poor approximation, as there would be no need for such an extensive Mac product lineup if this were true. The Mac mini and one laptop product line with two or three sizes would be more than sufficient, and continuing development of first-party Mac applications like Logic and FCP would serve little purpose.
They don't make any money on macOS anymore. You get it preinstalled on their computers, you can't buy it separately, you can't install it on other computers, and major upgrades are free.
It is a part of the computer that people pay for. This is like saying they don't get any money from Apple Silicon because its already installed on the motherboard.
They absolutely do not want their developers to mass migrate to another platform. Walk into any SF startup and you'll see 100% Mac laptops, except for the Finance guy who needs Excel on Windows, and that one dev who insists on Linux, and all those devs build webpages that work on Safari, and iOS apps before they (maybe) eventually get around to making an Android app.
Apple is highly incentivized to keep devs on Macs, even if it turned out to be a loss leader for them.
And so are developers on Linux. 99.5% of your backend stuff will run just fine on macOS without any modification, but it's been a good practice for a long time now to run stuff in containers (Docker/OrbStack/podman/etc.) in order to have a reproducible environment free of contamination. My development process in macOS is basically no different than it was on Linux.
It already feels a bit like it's two separate operating systems on one machine. It's still usable, but I am starting to notice odd behaviours. Files created in the terminal not showing up in the UI, there seems to be no notification from the filesystem to Finder, informing it that the terminal has created a file.
Developers and operations people are already starting to move to Linux. More and more I see colleague either switching to Linux, or have a mac laptop for administrative tasks and then a Linux workstation, or VMs, where they do their actual work.
> Developers and operations people are already starting to move to Linux.
I’ve been in this industry for a long time now. Linux always had more credibility among devs and ops folk. This “starting to move” view is weird.
My view is that OS X was acceptable by these teams because it was close enough to Linux and corporate could deal with it.
When I worked in corporate IT Linux was simply not an option for a primary workstation because there was no standard configuration or management tooling that made everyone happy. Macs were an easy compromise.
I personally don’t see macos anyhow competing with Linux or to be anyhow comparable to Linux in each others sweat points. linux dominated server side, so be its fair victory but it’s continually shit kind of clone of many things done in MacOs desktop experience. It just kept cloning Apple since forever to repeat a glimpse of that experience even during Next times with WindowMaker and then with many incarnations of modern desktop systems to compare with OSX and MacOs continually and in that direction. Linux is open, it’s different, that comes with value but still a little handicapped brother to MacOS desktop experience with continually f*up fundamentals even if you include wayland as the base. So is MacOs in servers but there Apple just resigned last year and stopped pretending they have servers. Both have only one thing in common, shell with base posix tools. Nothing else. And none are as open as original idea for Unix as this idea is long dead and linux tries to make Stallman’s wet dream about GNU system come true as his own kernel implementation did not move forward much it at all since mid 80s.. that’s the only UNIX reality that matters.. rest is bureaucracy and 5000 pages of POSIX standard nobody cares about as this is more about C libraries than kernel’s internals.
I don’t think it makes business sense for them to mass produce these for a consumer market but a data center of M5 chips sounds amazing for the workloads they’re running.
I don’t think the open source model will ever make an Apple quality UI because that’s the result of highly opinionated executives like Steve Jobs who don’t need to compromise or allow customization.
I’m glad both exist. I use them both and my life is good
Microsoft just shipped a full-fledged first class Linux container runtime as part of Windows[1]. I don’t think it even made front page of HN, which shows how far windows has fallen as a developer environment, but it is absolutely something Apple should be offering as a first class capability on OS X. The compatibility of the host OS file system with being mounted into docker images makes it a better environment for running such things.
Has it really, though? Because outside the HN bubble, Windows still seems to be the most common developer environment for most things that don't specifically target Linux or Apple platforms, and a popular option even then (e.g., in game dev).
> which shows how far windows has fallen as a developer environment,
Be mindful of living in an echo chamber. HN is one. There is a startup/SV-oriented culture here, which selects for a specific niche of people that is not representative of the general population. Going by this forum, you'd also think Java and especially C# were dead languages by how many votes and comments their threads get. Yet in the real world, Windows dev environments dwarf Linux, and Java/C# are everywhere in enterprise.
> Files created in the terminal not showing up in the UI
You might wanna look into that because that’s not how it works at all. I frequently switch between terminal and GUI apps and have not once noticed Finder failing to catch up with whatever it is I’m doing in the terminal.
Strange, because I have the same issue on two different machines. It doesn't happen all the time, but frequently enough that I assumed that Finder does some re-scan every now and then (or when you open a new Finder window).
In some alternative dimension Ternus decides to realize the full potential of the Mach kernel by embracing Hurd in place of XNU. Users are discouraged because the new system is not enough like GNU/Linux.
macOS as Unix is not dead – the compatibility with developer and unix infrastructure it has is an important reason why the company I work for uses MacBooks (at least I believe so).
We've found it to be a much worse experience than windows for dev stuff. It's a different OS and architecture to the target env (Linux). Aligning with that makes life considerably easier.
Your production is not going to be the same as everyone else's. It's also not going to be the ideal for everyone else's, or even possible for everyone else's.
The differences between ARM and x86 are probably less relevant for the majority of server-side development than API and admin interface differences.
In practice, anyone seriously developing Linux software on either macOS or Windows is using VMs or containers running the actual Linux kernel, at least for testing, and the difference is that Microsoft has put far more effort into integrating specific support for these things into the OS.
If you're using a language compiled to bytecode or interpreted, then the OS and arch don't matter a whole lot (so long as you don't have to deal with Window's weird filename conventions or other oddnesses). If you're using compiled-to-bytecode languages, there's a huge advantage in forcing your code to be cross-arch safe from day one, lest ye accidentally end up with something that can only build on the exact CPU in your devs' laptops or can only run on the flavor of WSL you're using that year.
Writing on Mx Mac and deploying on Intel Debian is a little harder in the beginning, but spares you oh, so many footguns down the road. Eventually you might want to deploy to a different OS or arch for cost or convenience reasons. That's a really bad time to find out that your project can't be made to work there without heroics.
Why not develop on the same env? I mean if I were you my macbook would just be an ssh client to the x86-64 dev box that emulates the target env. So many advantages to offshoring the compute to some server (my preference is your own hardware ofc but theres flexibility in cloud). For one, the laptop is not getting hot just being a dummy ssh client. I can spill a coffee on it during a build and nothing happens, build goes on and I am just out a thin client I can replace immediately with anything with ssh installed, my cellphone included if I am so desperate.
I think he means macOS as official, certified Unix is dead.
macOS as a more Linuxy-friendly place to work on Linuxy-type software, sure. Definitely true compared to Windows. But no one cares if macOS is blessed as Official UNIX.
How many people are writing software that they need to be able to run on Linux, Solaris, AIX, and HP-UX in 2026?
I would guess that your company isn’t actually interested in the UNIX compatibility specifically, more that they’re interested in laptops that can run a similar toolset.
They probably also buy MacBooks since they’re reliable, performant, high-endurance (battery life) general purpose business laptops that employees like using and see good productivity with.
Does your company care about compatibility with UNIX® infrastructure or that it's compatible with Linux/BSD infrastructure?
People care that macOS has a lot of compatibility with Linux/BSD, but most don't care that it's UNIX® - just as people don't care that Linux/BSD aren't UNIX® compatible. Yes, Linux/BSD are very similar to UNIX® and mostly compatible, but they don't conform to the specification.
You (probably) don't care about the UNIX® specification. You care that your MacBook has good compatibility with Linux/BSD infrastructure. Many people assume that Linux/BSD are UNIX® and just don't pay for certification. They aren't UNIX® compatible. They are extremely similar, but they don't follow the spec 100%.
Sorry for putting the ® everywhere, but people often use UNIX when they mean nix-like operating systems and when we're talking about whether people care about macOS as a UNIX® or not, it matters whether were saying "people want macOS to be UNIX®" vs "people want macOS to have good compatibility with nix-like operating systems such as Linux and BSD like it does today." People usually care a lot about the latter and not about the former.
Apparently it was for legal reasons. Apple was sued by the Open Group for describing Mac OS X as Unix® so they decided it was cheaper to bring it up to spec rather than to fight the lawsuit.
Enterpise espeically banking need standard and certification for reliability thats why , also certain Linux distro had certified before from China based on RHEL.
This is UNIX 03 certified OS with OSS alternatives from I gathered on internet :
- Apple OS X 10.7 - Apple macOS 26 , provide OSS Darwin 9 - Darwin 26
- Oracle Solaris 10 / 11 , provide OSS OpenSolais , then Illumos
- EulerOS 2.0 , based on RedHat Enterprise Linux / RHEL 7 and 8
- Inspur K-UX 2.0 , based on RedHat Enterprise Linux / RHEL 5
- Inspur K-UX 3.0 , based on RedHat Enterprise Linux / RHEL 7
Linux can’t be listed, because Linux is not an operating system. It’s just a kernel. Unix specifies a whole operating system, which includes things like libc, cat, sh etc.
At best a distro could be certified. But that certification would not automatically extend to other distros.
More than 1 Linux distributions have been tested, passed, and listed in the past... But you need to pay again every year, and recertify every new version, and they didn't bother keeping it up.
I described them and what this meant in some depth in this 2023 article:
We just ssh/http to the real Linux machine that runs our things anyway. These days with commoditized VPN and symmetric residential Gigabit internet, that Linux box does not have to be in the cloud either (specially with LLMs guiding the sysadmin tasks).
Any thin client with a browser, an ssh client and an internet connection will do the job.
From that Wayback page, it looks like only one version of macOS is certified at any one time: the latest version. Since macOS Golden Gate was only recently released, perhaps it's a process blip and we'll see Golden Gate listed soon.
Interesting - when I visit the original page from this post and choose "Filter by Standard" I can see Tahoe under UNIX 03, but only AIX is listed under UNIX V7.
This suggests that different parts of the website aren't necessarily in sync with other parts, information wise...
There's a long story behind it, but Apple started saying that Mac OS X was UNIX, but UNIX was a trademark. So Apple decided that they either needed to alter OS X to conform to UNIX standards or buy the trademark. They chose the former.
> Is this Apple just not wanting to pay for the certification anymore or does it actually change anything for us devs?
It's probably not about paying and more about tiny differences. Will it change anything for you? Almost certainly not. In fact, it might make things better for you. If anything, Apple's adherence to the UNIX spec can sometimes mean subtle differences in tools between macOS and BSD/Linux. Do you want things like grep or sed that conform to the UNIX spec or that work like grep and sed work in the real world?
And these differences are often small edge cases that don't come up much, but a spec is a spec.
The money is inconsequential for Apple, but at this point it probably makes more sense to just stop calling it UNIX and not worry about shipping a version of a tool exactly meets the UNIX spec - it'd be more useful for their users if it matched the behavior on Linux/BSD.
That grep/sed difference on mac caused me issues so many times, always thought it was just Apple doing their own thing.
I didn't realize it was a spec thing.
It's neither, really. OS X takes its command line utilities from BSD world, while Linux's mostly come from GNU. They're independent implementations, and GNU utilities are as a rule much more featureful than their BSD equivalents. Or much more bloated, depending on your point of view.
Both aim to conform to the POSIX, but POSIX is a very spartan spec, people routinely use non-POSIX features without realizing it.
While strict UNIX compliance only exists with a few esoteric operating systems these days, the POSIX standard still has value. To wit, next time I get in an argument with the GCC or clang devs that the newest version of their compiler no longer compiles my open source software, or likewise if the userspace of a given Linux distribution no longer compiles my open source software, we can resolve whether it’s a bug with my software or their software by looking at the POSIX (or C99) spec and hashing it out.
For example, I once got in to an argument about how to handle code like this:
#include <stdint.h>
uint32_t rotate(uint32_t i, int bits) {
return (i>>bits) | (i<<(32-bits));
}
Whether this should return i if we rotate by 0 bits (like it does in GCC) was a discussion I had with the clang developers, and, lo and behold, we get this:
If the value of the right [shift] operand is negative or is greater than or equal to the width of the promoted left operand, the behavior is undefined.
If the quotient a/b is representable, the expression (a/b)*b + a%b shall equal a
And:
a result that cannot be represented by the resulting unsigned integer type is reduced modulo the number that is one greater than the largest value that can be represented by the resulting type
Likewise, POSIX also means that every system call my open source code does, with the exception of chroot() and setgroups(), is defined in POSIX [1].
The entire environment to make my open source software only depends on POSIX utilities to build, and, indeed, it can build in both a Busybox based and a GNU based environment. For example, none of the filenames in my project have a `-` (dash) in them because, until POSIX 2024, one could not have a `-` or `/` in a filename used by Make, and yes my code builds with a strictly POSIX Make.
I no longer use Python for any of my open source because Python doesn’t have a stable standard to develop against; namely the Debian team got on my case I was using a Python2 script in 2020 when Python2 was made unsupported and I’m not going to play catch up should Python3 (or Perl) ever encounter the same fate Python2 has.
Instead, I write my scripts in Lua 5.1 these days. While Lua 5.1 isn’t a formal standard, the original implementation of Lua 5.1 (or a later version of Lua) has been implemented many times in many languages. There are two independent C implementations of Lua 5.1 in C, and there are also implementations in Go, C#, Javascript, Rust, and so on.
[1] /dev/urandom is also not defined in POSIX, but this is a runtime, not compile time issue, and writing a separate utility to put random bits in a normal file called /dev/urandom is one way to solve the issue.
This doesn't mean anything. The official Unix registry is completely irrelevant to the modern world. Nobody uses OSX pretending it's compatible with Solaris. Like Microsoft, Apple only cared about this certification because it used to be a checkbox on certain government procurement tenders.
When you're small, you need the certificates and badges others give you to be seen and be validated. I really don't think anyone needs to become familiar with Apple or macOS and that any of these certificates now matter a tiny amount to a multi trillion dollar corporation.
Given the amount and volume of software that mac has, the fact that almost all new interesting stuff are released for mac first or even mac alone, I also don't think it needs to attach itself to bigger names for platforms.
I think the financial burden of Ubuntu Mobile, Unity 8, and Mir; and the immediate rise of docker and WSL made canonical fully shift to server, so much so that their website no longer is about a better os for people, but what can it do for your cloud. Other distros also followed its path and Linus became all in on servers and IoT. I see windows as the biggest pile of trash in the world. Infinitely inconsistent, buggy, patched worked, visionless and rotting. Honestly that is one piece of software that is worse that AI slop.
And so what is there for Apple to compete with honestly? Haiku? Solaris?
And if you're honest, macOS has sadly dropped in quality too...
But the reason why I wrote my comment was just that you gave off the feeling that you consider macOS to be the only relevant desktop operating system, which seemed like a very strange view to have given the reality of it's userbase.
It'd matter if this certification came with an extensive test suite guaranteeing some degree of compliance. Instead, you have macOS defaulting to an antediluvian bash as /bin/sh full of ridiculous bugs like this:
In many ways, macOS is kind of driving towards irrelevance.
Apple wants it to be more like iOS (so they can control what you do on it) while users want a normal computer they can do computer things on without it spewing ads or applying political pressure when someone tries to hold Apple accountable (e.g. the EU trying to make Apple stop abusing their market position).
> In many ways, macOS is kind of driving towards irrelevance.
You mean the second most used desktop OS with about 15% of the global consumer OS market? About 30% of consumer desktops in the US alone? That “irrelevant” macOS?
Indeed it is one in the same. The middle layer of MacOS has been rotting for years, the introduction of gatekeeper and SIP has made downloading and running programs or observing your system a pain. Software developers are being pushed toward App models and many other examples exist. For people who simply need a computer that works MacOS is moving futher away from that goal with every release.
I won't be surprised if you will have to boot into recovery, disable SIP, come back to user-space and run those unsafe 'ls' and 'cat' commands you were trying to live dangerously with on '~/Library', and then go back to recovery again to renable SIP to stay safe. You may also have to do that to install every software you want to install and use on your (is it?) mac other than the ones that are from companies that had tied up with and signed Apple's USA certificaiton (i.e User Safety Alliance™; involving non-publicly disclosed deals), or from the smaller devs who were not small enough and could pay a few thousand dollars annually and sent Apple 3 kidneys annually to publish softwares with very reduced scope and access on the device/OS. The data transfer between your mac and those app endpoints will be so protected that even your own self-destructive hands and insufficient intellect won't be able to gather what is happening beneath the Apple's hardware hardcoded encryption. All for safety beyond the UNIX way. So not all surprising.
Have you ever considered, oh maybe using a Mac rather than making up complete bollocks(*)? Or perhaps looking up the well-documented and fully supported procedure for installing unsigned software? Or maybe noticing that the procedure is usually only needed for applications like FreeTube where the developer has a reason to remain anonymous? Yes, you can disable SIP. But you will struggle to find a need to do so.
If that’s what you get reading that comment then maybe ask your lawyer to pump some grey grease into the upper few inches next time you two are at a petrol err gas station.
> usually only needed for applications like FreeTube
No, looks like that won’t be sufficient. Pumping grey won’t be sufficient. Ask the lawyer to also do a “recent top posts on hn” search for you, assuming their marbles don’t yet melt at first sight of fruit company criticism/rant.
> you will struggle to find a need to do so.
Right. A member of the “you are holding it wrong” + “this is how it’s supposed to be used” genus spotted. Not your fault then. Here you go: yeah, you are right. Consider yours truly enlightened and in your immense intellectual debt.
The certification was always a bit odd, applying only to a configuration nobody would run in practice: https://www.osnews.com/story/141633/apples-macos-unix-certif...
The whole UNIX certified thing is a bit odd to start with, but those addendums are really just saying that because the UNIX compliance tests are thoroughly stuck in the past, a bunch of security measures (that users expect to exist on modern systems) have to be disabled to run the test suite.
SIP is not some irrelevant detail, nor is it an obviously good "modern security practice". A system that the hardware owner cannot modify is not really on the same wavelength as traditional UNIX.
The hardware owner can modify it all they want, disable SIP. For a large swath of average and non-average users that will not be modifying the closed-source OS they're running, it is obviously good.
Traditional UNIX is does not mean having an open source core you're intended to tinker with. Traditional UNIX was entirely proprietary, connected with well-defined text-based interfaces. This is entirely orthogonal to user control and manipulation of the OS.
Traditional UNIX was entirely proprietary, with source you're intended to tinker with.
AT&T was forbidden by its consent decree from productizing Unix. Licensees received the source code, on an as-is basis, with no official support from AT&T.
Sure, if by “you” you mean universities, Sun, DEC, IBM, etc.
As a non corporation end user with a sublicense from one of the above licensees (which is the apt analogy for an individual user in my opinion) you most certainly did not receive the source or have the rights to it.
I regard "traditional Unix" as the 1970s era. There were no sublicensees. You bought the source code from AT&T. That was the only option.
You then likely hacked the source code to make it work on your machine.
I can understand that others might include the commercialization of the 1980s and the Unix wars as part of "traditional Unix", but even then, my IRIX system ca. 1990 supported end-user kernel extensions, to handle new types of devices. Our imaging hardware was not "connected with well-defined text-based interfaces".
Sure, but that just means UNIX vendors had access to the source code.
It doesn't mean end users had access to the source code.
In fact, vendors would have been prohibited from distributing AT&T source code to customers who weren't themselves AT&T licensees, just as CSRG was prohibited from distributing BSD to customers who weren't licensees before 4.4BSD-Lite.
The same goes for most other proprietary source-available products. For example, game developers can't open source Unreal Engine projects even though Unreal Engine source licenses are available to end users for free.
You would absolutely normally get source to the vast majority of the unix components with a unix install.
At a bare minimum, you would almost have to have source to the vast majority of the kernel because there was no kernel modules. Sysgenning for a particular system involved compiling the kernel for that set of hardware.
I'm literally working on an early sun system emulator right now, heavily assisted by the standard sec tape that came with the standard software distribution of sunos, which has plenty of AT&T code.
No?
As jeremyjh wrote at https://news.ycombinator.com/item?id=50032733 , "A system that the hardware owner cannot modify is not really on the same wavelength as traditional UNIX.". Since traditional Unix was source-available, with the expectation that people would hack on the source, I fully agree with jeremyjh's characterization.
I think you are talking about a later era in Unix history? The early Unix licensees were educational institutions and commercial sites, not vendors or resellers.
Every one of those academic and commercial end users got the source code.
From https://en.wikipedia.org/wiki/History_of_Unix :
> "In 1973, AT&T released Version 5 Unix and licensed it to educational institutions, and licensed 1975's Version 6 to companies for the first time, ... Anyone could purchase a license, but the terms were very restrictive; licensees only received the source code, on an as-is basis. ... Unix was entirely unsupported by AT&T"
From the same link, "Perkin-Elmer began distributing Wollongong Unix for Interdata as Perkin-Elmer Edition 7, the first version supported by a computer company". That was in 1980.
So while yes, "Traditional UNIX was entirely proprietary" and could not be redistributed to non-licensees, it was also provided with "full user control and manipulation of the OS", and not merely "connected with well-defined text-based interfaces", quoting jdiff at https://news.ycombinator.com/item?id=50033080.
Also worth keeping in mind that macOS is intended for use with commercial third party software that’s repeatedly been proven untrustworthy, doing things like prodding system internals and installing dodgy kernel extensions where neither is warranted.
SIP is as much defense from the likes of Adobe and Google as it is from more traditional malware.
For now SIP can be disabled (on mac, not of course on iOS that is completely closed) but I wonder for how much. It's clear that the desire of Apple is to make macOS a closed system like iOS, installing third party software outside the AppStore is more and more difficult, with Gatekeeper that asks you to enter the system settings and warn about "potential insecure stuff" each time you want to run an application downloaded from the internet not signed by Apple (yes I know that if you disable SIP you can get rid of Gatekeeper as well).
I think that piece by piece Apple wants to go into direction of making macOS just a desktop version of iOS, that is closed and impossible to modify. They will use of course security as an excuse, the practical thing is that with iOS they succeeded because they went to the market with an OS that was already closed, but making users accept that in new version downloading a software from the internet and installing it it's no longer possible takes time. But they are getting there soon to me.
Yup. And why is UUCP still in there? I mean, yes I have used it, so far back that the docs were written in Quenya. But requiring it now?
Btw, MacOS 27 (the current version) still has UUCP.
Reminds me of (technically) POSIX compatible Windows.
"""POSIXLY_CORRECT (originally proposed as POSIX_ME_HARDER) is a historical environment variable used in GNU utilities to force strict compliance with the POSIX standard."""
ah, yes! https://en.wikipedia.org/wiki/Microsoft_POSIX_subsystem
the netbsd gallery (https://netbsd.org/gallery/) had a picture of somebody using that on windows xp (iirc) in order to compile pkgsrc packages in a "posix compliant" environment. the picture seem to be gone now :(
My first exposure to ‘compliance is not equivalence’.
The certification also was kind of bad at actually checking that the standard interfaces behaved as they should.
The certification documents read like the world's most abstruse bar trivia quiz...
"1.12 Threads
1.12.1 Cancellation Points
Question 31: Which C stdio functions have cancellation points that occur when a thread is executing?"
https://www.opengroup.org/csq/repository/noreferences=1&RID=...
If you'd try to certify any modern Linux distribution, you'd end up with a very similar list.
- The Linux kernel also has timer coalescing
- Linux also has lazy atime ('relatime') by default
- No Linux distro installs uucp by default, and certainly not suid
- Desktop distros will also run a file indexer of some sort
etc.
Isn’t the difference here that Linux was never trying to claim UNIX certification?
And famously, GNU's Not UNIX
Increasingly Linux is not gnu. Gcc as replaced in many places a while ago, Ubuntu has replaced coreutils.
This comment made me feel awkward for two reasons.
First, Linux was never GNU - the GNU ecosystem has always been a significant part, but if you were obstinate enough, you could get away without using any GNU tools (which is even easier today).
Second, you give just two examples. A few years ago we had a trend to rewrite everything in Rust with these newer tools usually not being part of the GNU project. Sure, you can replace everything now. However, most GNU tools are here to stay for a long time especially in enterprise distros where backward compatibility is important.
So no, Linux has never been GNU, and whatever fluctuations we had over decades with innumerable fights like egcs vs gcc etc., the GNU ecosystem is and will long be part of Linux distros.
Over time I see fewer and fewer gnu tools in my standard mainstream Linux. The tooling I used 25 years ago was far more gnu based than today, and crucially the FSF view of what free software is, and by extension the gnu public license, seems to be constantly retreating.
25 years ago it was reasonable to say the typical Linux machine was “gnu/linux”. Today it’s not.
Huawei managed to get their RHEL fork (EulerOS) UNIX certified. However their certificate expired (probably due to sanctions?)
So legally at one point a Linux was Unix.
If I remember correctly, two Linux variants were on the list at one time.
They were:
Inspur K-UX (expired on 3 February 2019)
Huawei EulerOS (expired in September 2022)
https://en.wikipedia.org/wiki/Inspur_K-UX
https://en.wikipedia.org/wiki/EulerOS
These were both based on Red Hat.
The relevant text from the article, (the rest of this comment is a direct quote from the article):
So, if you want your installation of macOS 15.0 to pass the UNIX® 03 certification test suites, you need to disable System Integrity Protection, enable the root account, enable core file generation, disable timeout coalescing, mount any APFS partitions with the strictatime option, format your APFS partitions case-sensitive (by default, APFS is case-insensitive, so you’ll need to reinstall), disable Spotlight, copy the binaries uucp, uuname, uustat, and uux from /usr/bin to /usr/local/bin and the binaries uucico and uuxqt from /usr/sbin to /usr/local/bin, set the setuid bit on all of these binaries, add /usr/local/bin to your PATH before /usr/bin and /usr/sbin, enable the uucp service, and handle the mystery issues listed in the four Temporary Waivers.
Then, and only then, is your macOS 15.0 actually UNIX® 03-certified.
This is batshit insane. I can guarantee you with 100% certainly not a single macOS installation in the entire history of macOS – let alone when just counting macOS 15.0 – has implemented even half of these changes. I’m sure there is a small number of people who have System Integrity Protection disabled permanently, and an even smaller number of people who have enabled the root account, and an even smaller number of people who have done both of those things – but that’s it. All the other changes are far too obscure and specific to be of any use to anyone.
This uucico stuff is funny. Do you think there is anyone actually using UUCP on MacOS?
I had no idea there was a built-in way to make the filesystem case sensitive. My employer probably does not want me to spend time reinstalling my OS for this now but maybe next time I'm given a MacBook for a job I should do this...
Curiously, the iOS filesystem is case-sensitive. A bit of a trap for developers on macOS. I’d expect plenty of third-party stuff to break switching the system drive to case-sensitive.
Makes sense; the iOS filesystem isn't exposed to the user, but macOS' is.
It actually is since iOS 11.
You can make an apfs volume that’s case sensitive and mount it as your project directory/etc, no need to reinstall the whole OS.
If I only want one directory to be case sensitive, yeah. I like when all of them are though.
You think that you do, like I thought that I did, but I doubt that you want it in reality. Last I checked, which was probably under a decade ago(?), there were a bunch of things on macOS that misbehave on a case-sensitive FS, both first- and third-party. I reinstalled same-day.
That's frustrating. Yet another reason for me to prefer Linux for my personal devices and just to tolerate macos for work I guess.
That’s a bit of a weird article and I’m not 100% sure what its point is.
> This is batshit insane. I can guarantee you with 100% certainly not a single macOS installation in the entire history of macOS – let alone when just counting macOS 15.0 – has implemented even half of these changes.
Well, maybe that means it doesn’t really matter and especially not to the extent that the author seems (?) to be making it out to be. If no one disables System Integrity Protection in order to be UNIX® 03-compliant, maybe that means that to the extent that people care about macOS being “a UNIX”, they don’t care specifically about the parts of the specification that are incompatible with System Integrity Protection.
The explanation I remember seeing many years ago, no idea how true it is, was that this was entirely about purchasing.
There were (are?) companies and especially parts of the government where it was pretty difficult/impossible to purchase anything that wasn’t on an official pre-approved list. Trying to get anything new added was a massive problem.
One of the things that was on the list was something like “a computer running Unix“. Tada. Macs now run Unix so you can easily buy them.
Like other people here, I really don’t think true compliance mattered to basically anyone for their actual use. The idea that it was just box checking so they were easier to buy in some institutional environments makes much more sense to me.
I can see that. Before linux, there were lots of versions of unix from all kinds of vendors.
I'll bet all kinds of government contracts needed some way to compare vendors bidding on a contract
for example:
- These ran UTX-32, Gould's version of Unix based on a BSD 4.2 kernel developed by Gould's UNIX development centre in Urbana-Champaign (IL) to support multiprocessor systems
- AIX is based on UNIX System V with 4.3BSD-compatible extensions.
- DG/UX 1.00, released by Data General in March 1985, was based on UNIX System V Release 2 with additions from 4.1BSD.
- HP-UX (from "Hewlett Packard Unix") is a discontinued proprietary implementation of the Unix operating system developed by Hewlett Packard Enterprise
- Xenix is based on the operating system UNIX System V, which Microsoft licensed from AT&T Corporation in the late 1970s.
You know what, I think the story I’m remembering may have been about NeXT. It was one reason so many three letter government agencies bought them.
So OS X continuing that strategy early on would make perfect sense.
Tahoe was Unix 03 certified, and is still listed on that page if you filter by standard and choose Unix 03.
Perhaps Golden Gate simply hasn't yet passed certification.
Indeed. And, weirdly, there are items that show up by that 03 standard alongside macOS 26 (e.g. "HP-UX 11i V3 Release B.11.31") that also show on the front page. Is it a latest version kind of thing?
Yea — said as much in another comment. That looks like it could be the rule — for example your HP-UX version there appears to be the latest, final HP-UX.
Yep linked page is active certification rather than historical.
I swear I've seen these "[Latest macOS release] does not appear in UNIX certification site" threads every year since the release of Big Sur in 2020.
It takes several months for the certification to appear on the site. What would be newsworthy is if we see a release next year without this release getting certification.
If you want to start a new trend, you should do this with FIPS 140, which has to be recertified every year. "Apple crypto is no longer FIPS certified"
Even better, last year's 26 releases are still undergoing the process.
This is what flagging is for.
Yeah, I remember this happening before. Quick search turned up this from two years ago for Sequoia https://www.reddit.com/r/MacOS/comments/1fl035s/macos_sequoi...
Apple originally certified Mac OS X as a UNIX because they used the word UNIX all over their website and the Open Group came knocking; they decided it was cheaper/better for goodwill to do the conformance and pay the license fee than to fight the case in court.
Looking now, I see almost no uses of the word UNIX on the website except the docs archive, so I guess they’ve decided to abandon both the certification and the use of the UNIX trademark in macOS.
Maybe. But I don't remember seeing them use the word for the last decade, and the last-but-one release was certified. My first guess would be that it's too new, and hasn't passed certification yet.
https://www.quora.com/What-goes-into-making-an-OS-to-be-Unix...
It is strange now, but Apple made a big deal of marketing it as UNIX back then. Like at the front-page of apple.com. UNIX had a good reputation as solid and serious.
Of course, it was either OS X or bust, back then.
The reason Apple went to the trouble was not really a marketing exercise, but rather because a lot of U.S. government contacts required it, or a waver. This is the same reason versions of Windows were also UNIX certified. In neither case did this certification actually make them more "standard" in the way that the rules originally envisioned.
So Apple's use of UNIX in marketing was to make sure they never ran into that as a disqualified on a government contract. My impression is that those contracts largely no longer require that (one small measure of common sense has caught up).
This comment should be at the top. The title implies a non-issue: macOS silently disappears because the poster didn’t try to change the page's badly designed filter. The latest version of macOS probably will appear in months.
If this is an intentional move, Apple is just acknowledging reality. UNIX(tm) is no longer so relevant compared to the early days of OS X, when Linux was still in its infancy and the Unix certification gave Apple street cred among developers. Most developers these days target Linux for production. That's why Microsoft developed WSL, not WSU.
> That's why Microsoft developed WSL, not WSU.
Developed WSL and discontinued SFU (Windows Services for UNIX), which they had since NT 4 days.
SFU 1.0 and 2.0 relied on the MKS Toolkit by MKS/PTC, and SFU 3.0 later adopted the Interix (OpenNT) subsystem, which Microsoft acquired from Softway Systems in 1999.
The original POSIX Subsystem (psxss.exe) was Cutler's team who implemented a minimal POSIX.1 subsystem server alongside Win32 and OS/2. This was done primarily to satisfy US Federal Government procurement requirements (FIPS 151-2).
True, though to Microsoft's credit, the existence of the POSIX subsystem, even as a proof-of-concept, ensured the underlying kernel design supported the development of products like Interix, just as Windows on Alpha et al. ensured NT remained reasonably portable to future platforms like ARM, unlike say OS/2 where the only non-Intel release (for PowerPC) was basically a prerelease lacking core OS features like networking.
You're mixing it up with NT's "personalities" of which OS/2 was the primary target at the time. UNIX was purely for FIPS contracts
I think they are saying that "Unlike OS/2 itself (as a competing MS and then IBM product) had less thought and design went into it as a portable system, whereas NT was designed portable ground up"
SFU ended in XP/2003. Vista/2008 had SUA
That's correct, I put them in the same bucket though, as SUA was also based on Interix. So I consider SUA more as a rebranding, not something completely different like WSL.
IIRC SUA required recompilation. Putting the motto "Windows NT is a UNIX" to life was challenging, and introduction of incompatible SUA added multiple versions maintenance to the challenge.
Linux has been effectively Unix™ for me ever since the days when Dennis Ritchie indicated that he used Linux for everything on his web page. Late 2000s at the latest, maybe early 2000s. Unix enough for Dennis should be Unix enough for anyone.
Caring about which systems are branded UNIX is like caring about which NBA player was paid to endorse a particular brand of athletic footwear.
Given the millions involved, plenty of people do.
I've been using Linux for 30+ years now and have always considered it free (as in speech) Unix. For my purposes, the difference was always more legal and philosophical than practical. Of course, I was a student back then and didn't need to run the commercial software of the time.
Well he and others moved away from UNIX, created Plan 9 and Inferno.
In these efforts they also tried to create successor languages to C, Alef and Limbo (for userspace), and never cared that much about anything past C89.
He has settled with Linux because that is the winner of UNIX wars, which predated Linux wars on Distrowatch.
As for the original UNIX,
"Using Unix is the computing equivalent of listening only to music by David Cassidy"
-- Rob Pike
Yes, Unix = Linux for all intents and purposes in 2026. When I started in the early 90s, there was a galaxy of Unices - SunOS/Solaris, HP-UX, AIX, and many others people people don't even remember (ConvexOS anyone?). Heck, back then if you wanted to run a particular software for your industry, you might have to buy and install a specific proprietary Unix.
Today? It's all Linux or Windows (of course, now also Web/SaaS). All these different Unices have fallen by the wayside in favor of Linux. Yes, some are still around but the mindshare has been in Linux for the last 20 years.
I suspect you could rip out a ton of autoconf rules and no one would notice.
There's still a lot of people on FreeBSD, OpenBSD, etc
There is plenty of mindshare on game consoles, and embedded OSes, in both cases, POSIX while not being GPL has a very high value.
Sure, I just hope macOS doesn't end up doing some custom thing rather than staying Unix/Linux like, which we've had with macOS/OS X till now.
I’m pretty sure Federal procurement is the big driver of relevance for this certification. They have always had a huge Fed business.
If they are in fact dropping it, that’s why. But… Apple often “finishes” the enterprise features with a point release.
WSLv1 worked via Unix Subsystem.
Proof? I couldnt find anything to that effect.
See my sibling comment about what it actually was.
It is hard to find, unless one knows the technical details.
It wasn’t the Unix subsystem, but was a subsystem… it implemented the various Linux syscalls and binary format directly and forwarded them to the NT kernel, similarly to how the various win32 syscalls call into the NT kernel. So WSL was a “peer” of the win32 subsystem in the same way SFU or SUA was. But it wasn’t implemented “via” SUA.
Then in WSL2 they just switched to using a VM, which is way more boring but probably less work to keep up with (I’m sure the Linux kernel has a lot of stuff that needs translating/implementing and it’s simpler to just use a hypervisor)
No it did not, it used Drawbridge OS research for Library OSes, which was integrated into Windows.
WSL 1.0 was a Drawbrige picoprocess.
Your confusion is natural among those that aren't into Windows programming.
Here are the links for the related information,
https://www.microsoft.com/en-us/research/project/drawbridge
WSL 1.0 architecture,
https://learn.microsoft.com/en-gb/archive/blogs/wsl/pico-pro...
If true, a better title would be "macOS is no longer UNIX certified".
This is fine and I hope it's permanent. Unix certification never really captured what once made "OS X is a Unix" attractive or exciting to many.
What I care about, at least, are experiences of ease, convenience, compatibility, control, order, and freedom around a programmable system that invites the blurring of the lines between user and developer, between quotidian use and automation, and between decoration and reshaping. Such experiences still involve many of the same elements as they did two decades ago: a composable language of commands, commonly observed conventions, well-organized hierarchies of files, formal software freedom, powerful package management systems, great in-band/offline documentation, etc. But Unix certification isn't the center of gravity around which such experiences revolve, if it ever was. And for that matter, neither is macOS.
MacOS 26 is still listed: https://www.opengroup.org//openbrand/register/xy.htm
Was it delisted from a different “product standard”, or is this submission about macOS 27 not being listed?
While it may simply acknowledge reality, I’ve already moved my main development off macOS to a NixOS workstation months ago. Too much friction in an OS that is clearly trying to be more “consumer” than “creative” at this point.
Which is a damn shame, because AI dev is going to result in a massive influx of new software right when Apple makes it ever more difficult to do so on its platform.
I still make Mac builds, but it is more of an afterthought on a secondary Mac build machine now.
You know something has peaked when they become less open and more “clawbacky”; that is basically collecting interest on an investment (marketshare) instead of remaining more open (at the cost of some things). “Increased security” is always the rationale.
> but it is more of an afterthought
If the platform I use is an afterthought for you then your software becomes afterthought for me as well.
Shit. Gonna have to make MacOS the top priority so osjdosnxkwjdk centers my software in their mind.
An account created 5 hours ago might not buy my hopefully-very-useful software, I am shaking in my booties
It is still there.
For those that don't know how to use a filter drop down, check UNIX 03 version.
That’s historical. The linked page is active. Zzz
I haven't seen much mention of APFS yet.
APFS is mandatory now. I recently used OCLP to coerce macOS 15 "Sequoia" onto my now-elderly 2015 27" Retina iMac. It's a capable machine: 4 cores, 32GB RAM, 1TB SSD as the system drive and 8TB HDD for data.
I kept my home directory on the HFS+ HDD. So, fast OS, slower but very capacious personal drive. It was a very simple hack and everything automatically obeyed. Worked fine up to Monterey (macOS 13), the last supported version.
Not now.
Lots of apps wouldn't run and the OS couldn't save lots of settings. I had to go through the first-login wizard every single time.
In the end I moved my home directory to the system volume on the SSD, and then moved all the big folders back out to HFS+ HDD (Documents, Downloads, Pictures, Music, Videos, etc.) This took 2 days of work but it did the job. Everything works now and the machine is noticeably faster, but now, I'm using 2x more SSD space and less HDD space.
The point being that the POSIX texts require setting classic Unix values on HFS+ drives, and I am not sure all of those still work on APFS, and that will make passing harder.
I think they might do it anyway, and soon, macOS 27, on Arm64 only, might pass... but in an even more wildly nonstandard config.
"silently" doesn't occur in the title, so is clickbait drama editorializing.
LLMs love "silent" and "quiet."
Also the suggestion Apple was entirely removed is false; see https://www.opengroup.org/openbrand/register/apple.htm
it's "silently" never left. :)
indeed, some may even call it a quiet load bearing footgun haha
I see macOS 26 listed under the Unix 03 standard. Was macOS 27 listed and then removed? The submission title + site contents combined do not give enough info to assess what exactly was silently removed.
macOS was listed on the main register (the link in this submission).
It was Tahoe, it’s since been removed.
The UNIX03 page still lists it, but I believe that’s a mistake.
macOS has been removed from the main register/list.
The index page seems to only list the status of current OS versions.
Perhaps Golden Gate simply hasn't passed certification yet? It's less than a month old.
macOS Tahoe was certified on launch (September 14th 2025)
https://news.ycombinator.com/item?id=45238930
Yes, that is interesting and certainly lends credence to the OP’s interpretation.
The this is much ado about nothing. Every single entry on the page you linked goes to its product page with the exact Unix standard it is certified as. Once you select a filter, you cannot get back to unfiltered list without the back button on the browser.
This is just a site rendering issue. Rather than silently removed, it's more like some bug was introduced. macOS 26 is still Unix 03 certified.
macOS is still very much registered: https://www.opengroup.org/openbrand/register/apple.htm
HN thread, 150+ comments:
The story behind macOS UNIX certification.
https://news.ycombinator.com/item?id=29984016
Well Unix as a standard is completely dead. macOS as a Unix is completely dead. I expect another branch in the already complicated as hell tree of Unixes.
Indeed.
I just wonder if macOS will stop being UNIX like.
That is, whilst it still has home brew and a BSD core it’s great. If that goes away it’s no longer a viable alternative to Linux.
What would that going away look like?
Swift on Mach
yeah I guess there isn't really much to move to... a new custom Kernel and Userspace? extremely unlikely - I guess it's here to stay, despite not being formally Unix.
More like iPadOS for example.
I'm suspecting some forces at Apple would be perfectly content with ripping most of the BSD core and the assorted CLI out and just have Linux containers spawned in VMs, WSL2-style.
would be awesome
Its called colima and its excellent.
https://colima.run/
and they have a first-party one now as well
https://github.com/apple/container
I haven't tried it. colima also simplifies linux VMs which is more of an analog to WSL2 than just running containers. It has really nice integration with Incus as well, if you like persistent system containers.
I also use colima mostly, it‘s great! (the Apple container thing de facto requires macos 26+)
Of course they were, even NeXTSTEP was designed to be compatible with UNIX only to compete with UNIX workstations like those from Sun.
Neither Apple nor NeXT had a UNIX culture, and efforts like A/UX and mkLinux were never that much serious.
Had they acquired Be, BeOS was even less UNIXy than NeXTSTEP.
It's distinctly problematic there unfortunately. Particularly the moment you add any UI toolkits to it. I've had so many problems with stuff that uses Qt because Qt takes a while to catch up with things Apple broke inside their Cocoa internals. Very irritating platform to work on.
Edit: love how I get downvoted for facts?! They literally broke Cocoa so it was throwing hopeless useless exceptions at the bottom of the stack when they changed the native file open boxes to run in another process. Two bits of software I use couldn't raise an open file dialog window any more.
I've been a software engineer for 25 years, doing Linux stuff at startups and big companies (the biggest even). In all my time as a software engineer, I've seen maybe 3 people who didn't use a MacBook Pro for work. Apple makes the bulk of its money from tech companies and schools and they cater very well to both. If it stops being Unix-like, they lose the thing they make the most money from. They are smarter than that.
You’re in a bubble. Windows is still huge, especially in middling companies that dictate it and standard HP/Dell/Lenovo hardware for all 3k staff.
It’s insanely common.
His point isn’t that it’s common, it’s that a unix-like environment is a key selling point.
Given that Windows made WSL in an attempt to capture developer mindshare, that tends to support his comment
You still end up in uncanny valley often enough, with commands that are missing, or working unexpectedly differently. It often feels like macOS could be made so much better for developers, but Apple just doesn't care because people buy their devices anyway.
In deeply saddens me to think that there is a developer who has not installed gnutils and put them earlier in the PATH.
The remaining differences are related more to systems stuff - managing daemons with launchd for example. These aren't really a distraction day-to-day.
It really is that simple. No different from a BSD or even a non-GNU (busybox, etc.) Linux distribution. If you like GNU tools, use them.
It’s a computer. I can make a computer go beep if I want, that is not the issue.
The fact that someone who didn’t grow up at the time when open source was an idealistic concept instead of free support to corporations needs to learn about the gnutils and how to install them with homebrew is what makes me sad.
Why do we seriously not have a command-line package manager on macOS yet? Windows has one by now. It’s just sad to me that you need a manual checklist of post-setup steps to turn a Mac into a useful machine for software development.
I don't think you have to grow up in some idyllic era that never existed. Its more like you type 'ls', notice that its total crap, and google it.
The point I was making was that knowing what GNU means in the first place probably requires some cultural background that many people lack, especially younger ones.
Right, and the point I'm making is that you can learn this from Google once you notice there is a difference from how common commands work on Linux and MacOS. You don't need to know about GNU, you just have to wonder why its different.
The problem is that being a developer is not a synonymous with UNIX.
It’s not that. It’s that their lawyers won’t let the engineering side distribute gnu tools. Period.
At Apple, if you even use GPL tools in your workflow you have to seek approval, and most often there’s pushback of “why won’t this (alternative that isn’t GPL) work ?”
Their lawyers aren’t stupid, they understand the GPL and its limits, but they’re simply doing a risk assessment- which one is less risky, the GPL variant or the BSD. For a corporation, that’s generally BSD.
I don't think apple makes the bulk from professionals at all, the iphone is their flagship product and their laptop line has been expanding towards casual markets with budget models.
Right but in this thread it is pretty clear we are talking about the money they make from MacOS.
The money they make from MacOS is a footnote compared to the money they make from phone hardware, the app store transaction tax, and ads. To a first approximation, the only reason that Apple cares about MacOS is so that people can make iOS apps.
This is a poor approximation, as there would be no need for such an extensive Mac product lineup if this were true. The Mac mini and one laptop product line with two or three sizes would be more than sufficient, and continuing development of first-party Mac applications like Logic and FCP would serve little purpose.
They don't make any money on macOS anymore. You get it preinstalled on their computers, you can't buy it separately, you can't install it on other computers, and major upgrades are free.
It is a part of the computer that people pay for. This is like saying they don't get any money from Apple Silicon because its already installed on the motherboard.
They absolutely do not want their developers to mass migrate to another platform. Walk into any SF startup and you'll see 100% Mac laptops, except for the Finance guy who needs Excel on Windows, and that one dev who insists on Linux, and all those devs build webpages that work on Safari, and iOS apps before they (maybe) eventually get around to making an Android app.
Apple is highly incentivized to keep devs on Macs, even if it turned out to be a loss leader for them.
Backend developers are already conditioned to use Linux VMs when developing in Mac OS.
And so are developers on Linux. 99.5% of your backend stuff will run just fine on macOS without any modification, but it's been a good practice for a long time now to run stuff in containers (Docker/OrbStack/podman/etc.) in order to have a reproducible environment free of contamination. My development process in macOS is basically no different than it was on Linux.
I bet you are located in US.
It already feels a bit like it's two separate operating systems on one machine. It's still usable, but I am starting to notice odd behaviours. Files created in the terminal not showing up in the UI, there seems to be no notification from the filesystem to Finder, informing it that the terminal has created a file.
Developers and operations people are already starting to move to Linux. More and more I see colleague either switching to Linux, or have a mac laptop for administrative tasks and then a Linux workstation, or VMs, where they do their actual work.
> Developers and operations people are already starting to move to Linux.
I’ve been in this industry for a long time now. Linux always had more credibility among devs and ops folk. This “starting to move” view is weird.
My view is that OS X was acceptable by these teams because it was close enough to Linux and corporate could deal with it.
When I worked in corporate IT Linux was simply not an option for a primary workstation because there was no standard configuration or management tooling that made everyone happy. Macs were an easy compromise.
I personally don’t see macos anyhow competing with Linux or to be anyhow comparable to Linux in each others sweat points. linux dominated server side, so be its fair victory but it’s continually shit kind of clone of many things done in MacOs desktop experience. It just kept cloning Apple since forever to repeat a glimpse of that experience even during Next times with WindowMaker and then with many incarnations of modern desktop systems to compare with OSX and MacOs continually and in that direction. Linux is open, it’s different, that comes with value but still a little handicapped brother to MacOS desktop experience with continually f*up fundamentals even if you include wayland as the base. So is MacOs in servers but there Apple just resigned last year and stopped pretending they have servers. Both have only one thing in common, shell with base posix tools. Nothing else. And none are as open as original idea for Unix as this idea is long dead and linux tries to make Stallman’s wet dream about GNU system come true as his own kernel implementation did not move forward much it at all since mid 80s.. that’s the only UNIX reality that matters.. rest is bureaucracy and 5000 pages of POSIX standard nobody cares about as this is more about C libraries than kernel’s internals.
You may be interested to learn that Apple does in fact have servers internally.
https://wccftech.com/apples-private-cloud-compute-server-m5-...
I don’t think it makes business sense for them to mass produce these for a consumer market but a data center of M5 chips sounds amazing for the workloads they’re running.
I don’t think the open source model will ever make an Apple quality UI because that’s the result of highly opinionated executives like Steve Jobs who don’t need to compromise or allow customization.
I’m glad both exist. I use them both and my life is good
Game and app developers certainly not.
Microsoft just shipped a full-fledged first class Linux container runtime as part of Windows[1]. I don’t think it even made front page of HN, which shows how far windows has fallen as a developer environment, but it is absolutely something Apple should be offering as a first class capability on OS X. The compatibility of the host OS file system with being mounted into docker images makes it a better environment for running such things.
[1] https://blogs.windows.com/windowsdeveloper/2026/09/29/wsl-co...
https://github.com/apple/container
Well… that’s handy.
Has it really, though? Because outside the HN bubble, Windows still seems to be the most common developer environment for most things that don't specifically target Linux or Apple platforms, and a popular option even then (e.g., in game dev).
> which shows how far windows has fallen as a developer environment,
Be mindful of living in an echo chamber. HN is one. There is a startup/SV-oriented culture here, which selects for a specific niche of people that is not representative of the general population. Going by this forum, you'd also think Java and especially C# were dead languages by how many votes and comments their threads get. Yet in the real world, Windows dev environments dwarf Linux, and Java/C# are everywhere in enterprise.
> Files created in the terminal not showing up in the UI
You might wanna look into that because that’s not how it works at all. I frequently switch between terminal and GUI apps and have not once noticed Finder failing to catch up with whatever it is I’m doing in the terminal.
Strange, because I have the same issue on two different machines. It doesn't happen all the time, but frequently enough that I assumed that Finder does some re-scan every now and then (or when you open a new Finder window).
Finder windows subscribe to filesystem events, so the reaction is pretty much instantaneous.
In some alternative dimension Ternus decides to realize the full potential of the Mach kernel by embracing Hurd in place of XNU. Users are discouraged because the new system is not enough like GNU/Linux.
macOS as Unix is not dead – the compatibility with developer and unix infrastructure it has is an important reason why the company I work for uses MacBooks (at least I believe so).
We've found it to be a much worse experience than windows for dev stuff. It's a different OS and architecture to the target env (Linux). Aligning with that makes life considerably easier.
Linux doesn't have an "architecture".
You know what I mean. ARM vs x86-64 -and- macOS API vs Linux API.
And yes, they are different even at the user end on the shell.
All of our production Linux systems are running on Graviton these days, i.e. they’re arm64.
Your production is not going to be the same as everyone else's. It's also not going to be the ideal for everyone else's, or even possible for everyone else's.
The differences between ARM and x86 are probably less relevant for the majority of server-side development than API and admin interface differences.
In practice, anyone seriously developing Linux software on either macOS or Windows is using VMs or containers running the actual Linux kernel, at least for testing, and the difference is that Microsoft has put far more effort into integrating specific support for these things into the OS.
If you're using a language compiled to bytecode or interpreted, then the OS and arch don't matter a whole lot (so long as you don't have to deal with Window's weird filename conventions or other oddnesses). If you're using compiled-to-bytecode languages, there's a huge advantage in forcing your code to be cross-arch safe from day one, lest ye accidentally end up with something that can only build on the exact CPU in your devs' laptops or can only run on the flavor of WSL you're using that year.
Writing on Mx Mac and deploying on Intel Debian is a little harder in the beginning, but spares you oh, so many footguns down the road. Eventually you might want to deploy to a different OS or arch for cost or convenience reasons. That's a really bad time to find out that your project can't be made to work there without heroics.
Why not develop on the same env? I mean if I were you my macbook would just be an ssh client to the x86-64 dev box that emulates the target env. So many advantages to offshoring the compute to some server (my preference is your own hardware ofc but theres flexibility in cloud). For one, the laptop is not getting hot just being a dummy ssh client. I can spill a coffee on it during a build and nothing happens, build goes on and I am just out a thin client I can replace immediately with anything with ssh installed, my cellphone included if I am so desperate.
The relevance of Open Group and the UNIX brand has been a walking corpse for decades.
I think he means macOS as official, certified Unix is dead.
macOS as a more Linuxy-friendly place to work on Linuxy-type software, sure. Definitely true compared to Windows. But no one cares if macOS is blessed as Official UNIX.
How many people are writing software that they need to be able to run on Linux, Solaris, AIX, and HP-UX in 2026?
I would guess that your company isn’t actually interested in the UNIX compatibility specifically, more that they’re interested in laptops that can run a similar toolset.
They probably also buy MacBooks since they’re reliable, performant, high-endurance (battery life) general purpose business laptops that employees like using and see good productivity with.
Does your company care about compatibility with UNIX® infrastructure or that it's compatible with Linux/BSD infrastructure?
People care that macOS has a lot of compatibility with Linux/BSD, but most don't care that it's UNIX® - just as people don't care that Linux/BSD aren't UNIX® compatible. Yes, Linux/BSD are very similar to UNIX® and mostly compatible, but they don't conform to the specification.
You (probably) don't care about the UNIX® specification. You care that your MacBook has good compatibility with Linux/BSD infrastructure. Many people assume that Linux/BSD are UNIX® and just don't pay for certification. They aren't UNIX® compatible. They are extremely similar, but they don't follow the spec 100%.
Sorry for putting the ® everywhere, but people often use UNIX when they mean nix-like operating systems and when we're talking about whether people care about macOS as a UNIX® or not, it matters whether were saying "people want macOS to be UNIX®" vs "people want macOS to have good compatibility with nix-like operating systems such as Linux and BSD like it does today." People usually care a lot about the latter and not about the former.
I never quite understood what the point of this was beyond marketing. Linux is also not listed probably because no one ever paid to have it listed
Enterprise. Checklists.
And Gnu’s Not UNIX.
It would be marvelous to see GNU/Linux certified as a UNIX specifically with rms’s preferred GNU/Linux as the name.
They could even just call it GNU is now UNIX.
It was Stallman who came up with the name POSIX by the way.
IIRC there were some UNIX Linux distributions in the past but the certification wasn’t renewed
As described in the thread, EulerOS and Inspur K/UX were both certified in the past.
They were CentOS Linux derivatives.
It is specific Linux distributions that would be Unix-certified. Huawei had one; not sure if it has ever been on this page.
Apparently it was for legal reasons. Apple was sued by the Open Group for describing Mac OS X as Unix® so they decided it was cheaper to bring it up to spec rather than to fight the lawsuit.
https://retrocomputing.stackexchange.com/questions/27095/why...
https://www.quora.com/What-goes-into-making-an-OS-to-be-Unix...
It was a legal problem, and an interesting story: https://www.quora.com/What-goes-into-making-an-OS-to-be-Unix...
Enterpise espeically banking need standard and certification for reliability thats why , also certain Linux distro had certified before from China based on RHEL.
This is UNIX 03 certified OS with OSS alternatives from I gathered on internet :
- Apple OS X 10.7 - Apple macOS 26 , provide OSS Darwin 9 - Darwin 26
- Oracle Solaris 10 / 11 , provide OSS OpenSolais , then Illumos
- EulerOS 2.0 , based on RedHat Enterprise Linux / RHEL 7 and 8
- Inspur K-UX 2.0 , based on RedHat Enterprise Linux / RHEL 5
- Inspur K-UX 3.0 , based on RedHat Enterprise Linux / RHEL 7
Linux can’t be listed, because Linux is not an operating system. It’s just a kernel. Unix specifies a whole operating system, which includes things like libc, cat, sh etc.
At best a distro could be certified. But that certification would not automatically extend to other distros.
This is incorrect.
More than 1 Linux distributions have been tested, passed, and listed in the past... But you need to pay again every year, and recertify every new version, and they didn't bother keeping it up.
I described them and what this meant in some depth in this 2023 article:
https://www.theregister.com/software/2023/01/17/unix-is-dead...
Tahoe:
https://www.opengroup.org//openbrand/register/xy.htm
I imagine that the cert will be listed for Golden Gate soon.
Why is the client OS important?
We just ssh/http to the real Linux machine that runs our things anyway. These days with commoditized VPN and symmetric residential Gigabit internet, that Linux box does not have to be in the cloud either (specially with LLMs guiding the sysadmin tasks).
Any thin client with a browser, an ssh client and an internet connection will do the job.
Less things that can go wrong.
Sometimes just having a local machine that you can do everything on is what you need
Why would anyone care about what the Open Group has to say in this day and age? Is anyone actually still using SCO or AIX any more?
It could be that macOS 27 never got Unix certified. Wayback Machine on Oct 7th only shows macOS 26 being Unix certified.
https://web.archive.org/web/20261007103902/https://www.openg...
27 wasn’t certified and 26 was listed but is now removed.
From that Wayback page, it looks like only one version of macOS is certified at any one time: the latest version. Since macOS Golden Gate was only recently released, perhaps it's a process blip and we'll see Golden Gate listed soon.
Nope, Sequoia and Tahoe both remain registered: https://www.opengroup.org/openbrand/register/apple.htm
Interesting - when I visit the original page from this post and choose "Filter by Standard" I can see Tahoe under UNIX 03, but only AIX is listed under UNIX V7.
This suggests that different parts of the website aren't necessarily in sync with other parts, information wise...
26 is in UNIX 03, if you use the pulldown
I didn't realize macOS was actually certified UNIX since 2007, thought it was just marketing.
Is this Apple just not wanting to pay for the certification anymore or does it actually change anything for us devs?
There's a long story behind it, but Apple started saying that Mac OS X was UNIX, but UNIX was a trademark. So Apple decided that they either needed to alter OS X to conform to UNIX standards or buy the trademark. They chose the former.
> Is this Apple just not wanting to pay for the certification anymore or does it actually change anything for us devs?
It's probably not about paying and more about tiny differences. Will it change anything for you? Almost certainly not. In fact, it might make things better for you. If anything, Apple's adherence to the UNIX spec can sometimes mean subtle differences in tools between macOS and BSD/Linux. Do you want things like grep or sed that conform to the UNIX spec or that work like grep and sed work in the real world?
And these differences are often small edge cases that don't come up much, but a spec is a spec.
The money is inconsequential for Apple, but at this point it probably makes more sense to just stop calling it UNIX and not worry about shipping a version of a tool exactly meets the UNIX spec - it'd be more useful for their users if it matched the behavior on Linux/BSD.
Thanks for explaining.
That grep/sed difference on mac caused me issues so many times, always thought it was just Apple doing their own thing. I didn't realize it was a spec thing.
It's neither, really. OS X takes its command line utilities from BSD world, while Linux's mostly come from GNU. They're independent implementations, and GNU utilities are as a rule much more featureful than their BSD equivalents. Or much more bloated, depending on your point of view.
Both aim to conform to the POSIX, but POSIX is a very spartan spec, people routinely use non-POSIX features without realizing it.
Tahoe and Sequoia are still in the apple's apple's list(https://www.opengroup.org/openbrand/register/apple.htm)
Tahoe, Sequoia, Ventura, and Monterey were in main list until October 7.
Can't figure out the reason.
Perhaps Apple decided to get out of the way and let IBM and SCO duke it out on that page ;-)
It appears here https://www.opengroup.org//openbrand/register/xy.htm
I could see this being corporate incompetence or a genuine deviation. I wonder which one it is.
"Silently removed" meaning Apple probably did not bother going for certification of macOS 27.
And I'm not sure why it would be non-silently/loudly removed: is there an expectation of a press release or weblog post about non-certification?
They wanted mo' money from Apple than it was willing to give them.
Anyway, who cares. FreeBSD is Unix no matter what some standards organization says.
Better for both sides. They clearly have different visions for what an OS should do.
What? The Open Group has no vision for what an OS should do, it simply writes standards describing what implementations are already doing.
And does this have any actual impact on either the OS itself or its users?
well not specifically, but directionally could mean less adherence to a UNIX like environment.
Maybe it’s time to come out with a new standard that supersedes POSIX and Unix cert?
Standards are useful and the fact that these have lasted this long is IMHO evidence of that
It saddens me to think of all the CPU cycles spent on hosting online discussions that mention the fact that MacOS officially licenses the UNIX brand.
The registry of remaining products is a list of irrelevant operating systems hardly anyone uses anymore.
While strict UNIX compliance only exists with a few esoteric operating systems these days, the POSIX standard still has value. To wit, next time I get in an argument with the GCC or clang devs that the newest version of their compiler no longer compiles my open source software, or likewise if the userspace of a given Linux distribution no longer compiles my open source software, we can resolve whether it’s a bug with my software or their software by looking at the POSIX (or C99) spec and hashing it out.
For example, I once got in to an argument about how to handle code like this:
Whether this should return i if we rotate by 0 bits (like it does in GCC) was a discussion I had with the clang developers, and, lo and behold, we get this:
If the value of the right [shift] operand is negative or is greater than or equal to the width of the promoted left operand, the behavior is undefined.
As an aside, this works:
Because, as per the C99 standard:
If the quotient a/b is representable, the expression (a/b)*b + a%b shall equal a
And:
a result that cannot be represented by the resulting unsigned integer type is reduced modulo the number that is one greater than the largest value that can be represented by the resulting type
Likewise, POSIX also means that every system call my open source code does, with the exception of chroot() and setgroups(), is defined in POSIX [1].
The entire environment to make my open source software only depends on POSIX utilities to build, and, indeed, it can build in both a Busybox based and a GNU based environment. For example, none of the filenames in my project have a `-` (dash) in them because, until POSIX 2024, one could not have a `-` or `/` in a filename used by Make, and yes my code builds with a strictly POSIX Make.
I no longer use Python for any of my open source because Python doesn’t have a stable standard to develop against; namely the Debian team got on my case I was using a Python2 script in 2020 when Python2 was made unsupported and I’m not going to play catch up should Python3 (or Perl) ever encounter the same fate Python2 has.
Instead, I write my scripts in Lua 5.1 these days. While Lua 5.1 isn’t a formal standard, the original implementation of Lua 5.1 (or a later version of Lua) has been implemented many times in many languages. There are two independent C implementations of Lua 5.1 in C, and there are also implementations in Go, C#, Javascript, Rust, and so on.
[1] /dev/urandom is also not defined in POSIX, but this is a runtime, not compile time issue, and writing a separate utility to put random bits in a normal file called /dev/urandom is one way to solve the issue.
This doesn't mean anything. The official Unix registry is completely irrelevant to the modern world. Nobody uses OSX pretending it's compatible with Solaris. Like Microsoft, Apple only cared about this certification because it used to be a checkbox on certain government procurement tenders.
unix registry silently removed from relevance.
linux compatibility is way more relevant now.
Solaris as well, apparently
They used to marketing Mac OS X around Unix, but not anymore.
At that time they were trying to lure people from other Unixen such as Solaris for workstation use. Not any more.
Unix? What's this?
This is inevitable, there is no escaping it.
When you're small, you need the certificates and badges others give you to be seen and be validated. I really don't think anyone needs to become familiar with Apple or macOS and that any of these certificates now matter a tiny amount to a multi trillion dollar corporation.
Given the amount and volume of software that mac has, the fact that almost all new interesting stuff are released for mac first or even mac alone, I also don't think it needs to attach itself to bigger names for platforms.
I think the financial burden of Ubuntu Mobile, Unity 8, and Mir; and the immediate rise of docker and WSL made canonical fully shift to server, so much so that their website no longer is about a better os for people, but what can it do for your cloud. Other distros also followed its path and Linus became all in on servers and IoT. I see windows as the biggest pile of trash in the world. Infinitely inconsistent, buggy, patched worked, visionless and rotting. Honestly that is one piece of software that is worse that AI slop.
And so what is there for Apple to compete with honestly? Haiku? Solaris?
> I see windows as the biggest pile of trash in the world.
I mean I don't disagree with that, but you're in dire need of a reality check wrt Apple's dominance
It's not even the most used OS. It's userbase has certainly grown over the last 20 years, but it's still at <30% on desktop...
There are however some demographics in which it's more represented, and that's probably just such a bubble where you're coming from
Usage and quality are not particularly correlated, though.
I never said they were?
And if you're honest, macOS has sadly dropped in quality too...
But the reason why I wrote my comment was just that you gave off the feeling that you consider macOS to be the only relevant desktop operating system, which seemed like a very strange view to have given the reality of it's userbase.
How is SCO still a thing?
It now looks like a museum of obsolete operating systems.
Xnu’s Not Unix!
Who is running Xinu today?
It'd matter if this certification came with an extensive test suite guaranteeing some degree of compliance. Instead, you have macOS defaulting to an antediluvian bash as /bin/sh full of ridiculous bugs like this:
Note that they even ship dash as a much better /bin/sh...
These days it defaults to zsh instead of bash. The shipped bash is ancient due to (I believe) being last version released as GPL-2.0.
I'm not speaking about the default login shell, which is indeed zsh, but what /bin/sh points to.
In many ways, macOS is kind of driving towards irrelevance.
Apple wants it to be more like iOS (so they can control what you do on it) while users want a normal computer they can do computer things on without it spewing ads or applying political pressure when someone tries to hold Apple accountable (e.g. the EU trying to make Apple stop abusing their market position).
> In many ways, macOS is kind of driving towards irrelevance.
You mean the second most used desktop OS with about 15% of the global consumer OS market? About 30% of consumer desktops in the US alone? That “irrelevant” macOS?
Indeed it is one in the same. The middle layer of MacOS has been rotting for years, the introduction of gatekeeper and SIP has made downloading and running programs or observing your system a pain. Software developers are being pushed toward App models and many other examples exist. For people who simply need a computer that works MacOS is moving futher away from that goal with every release.
It’s not about market share, is it? Also people generally don’t choose macOS because macOS, it’s just the thing that comes with their Mac.
I won't be surprised if you will have to boot into recovery, disable SIP, come back to user-space and run those unsafe 'ls' and 'cat' commands you were trying to live dangerously with on '~/Library', and then go back to recovery again to renable SIP to stay safe. You may also have to do that to install every software you want to install and use on your (is it?) mac other than the ones that are from companies that had tied up with and signed Apple's USA certificaiton (i.e User Safety Alliance™; involving non-publicly disclosed deals), or from the smaller devs who were not small enough and could pay a few thousand dollars annually and sent Apple 3 kidneys annually to publish softwares with very reduced scope and access on the device/OS. The data transfer between your mac and those app endpoints will be so protected that even your own self-destructive hands and insufficient intellect won't be able to gather what is happening beneath the Apple's hardware hardcoded encryption. All for safety beyond the UNIX way. So not all surprising.
Except you don't.
oh no! security measures! scary!
They are, when used against the user.
Is this “make shit up” day, or something ?
> make shit up
When it's about the fruit company, is there a difference? Unless it's for the eyes of the beholden i.e., le iFans.
Have you ever considered, oh maybe using a Mac rather than making up complete bollocks(*)? Or perhaps looking up the well-documented and fully supported procedure for installing unsigned software? Or maybe noticing that the procedure is usually only needed for applications like FreeTube where the developer has a reason to remain anonymous? Yes, you can disable SIP. But you will struggle to find a need to do so.
(*) My lawyers have asked me to point out that the case of Regina vs Sex Pistols (1977) established that this term refers to a priest in Old English (https://en.wikipedia.org/wiki/Never_Mind_the_Bollocks,_Here%...).
Ah, the triggering at mere mention of an iFan.
> oh maybe using a Mac
If that’s what you get reading that comment then maybe ask your lawyer to pump some grey grease into the upper few inches next time you two are at a petrol err gas station.
> usually only needed for applications like FreeTube
No, looks like that won’t be sufficient. Pumping grey won’t be sufficient. Ask the lawyer to also do a “recent top posts on hn” search for you, assuming their marbles don’t yet melt at first sight of fruit company criticism/rant.
> you will struggle to find a need to do so.
Right. A member of the “you are holding it wrong” + “this is how it’s supposed to be used” genus spotted. Not your fault then. Here you go: yeah, you are right. Consider yours truly enlightened and in your immense intellectual debt.