It's interesting to see more customisation tools appearing for macOS, as just a few years ago I was looking for ways to strip down the OS (CI related), and while such info was widely available in the Windows world, to the point that customised "distros" are available, it was nearly nonexistent for macOS; only the Hackintosh community had some useful articles on how things worked.
Could the rise of LLMs and vibe-coding have motivated people who otherwise wouldn't bother?
It’s hard to strip it down these days as the OS image and its core, immutable filesystem cannot be edited. Admittedly this helps keep idiots from destroying their filesystem and also blocks many malware attacks on the system, but, for example. I don’t believe you can delete the chess program.
Not a lot of good reasons to upgrade to 27. They removed Rosetta and you have to reinstall that if you want it. So is MacOS turning into something that more regular people are going to have to maintain in the future or end up with something like Windows 11?
I switched to MacOS 3 years ago because of Microsoft and the writing on the wall seems to say I got another year left before I'm forced into Linux. Because if I have to maintain my own OS then I might as well install Linux and do it once.
If a temporary tool being retired is the reason you switch, then wait until you find out how many LOC are being deleted from the linux kernel this year and next
That doesn't seem to do much in the `curl | bash` setting, given that you're not verifying the attestation in that case. You still need to download it separately and run `gh attestation verify` first.
(Note that the attestation does not appear to cover the shell script either, it only covers the script's final payload. The shell script is also referenced via `main`, so it's mutable even if the underlying payload is properly attested. That's not good!)
“You wouldn't run a stranger's code without reading it.” Yes I would. We all do it all the time. macOS itself is closed source, and even if it weren't, there’s way too much code to read.
Code from trusted repositories is an entirely different thing compared to running 'wget some_github_repo_shell_script | sh' . That said, the likes of Tailscale are setting a bad example.
You download a dmg and run it blindly? You download an exe and run it blindly. I wish it were in an rpm or deb coming from signed repos, but it's not so here we are
But then again, I'm a bit paranoid. At a minimum I would download the script and read it, and if it was too long or not written clearly enough then I would just drop it and find something better.
Great initiative. I recently got stung by an advert on reddit for "HBO Max for MacOS, 6 months free" from the official HBO user (don't get me started on how that slipped through). Front and center was a curl | bash copy to clipboard that obfuscated the payload source in base10. I knew better, but I think we've made this kind of thing way too acceptable. Of course it was malware and I realized the instant I pressed enter. Thankfully I didn't give it my password and immediately disconnected from the internet and killed the machine. I'm genuinely concerned these kind of attacks are going to become much more commonplace with AI, plus the ability to inject malicious code in to things that get run by trusted scripted installers.
What’s your suggested installation method instead? Unless it’s “download and read the source before running it” this is no worse than npm install, or pip install, or clicking “trust” on a git repo in VSCode
It is actually worse than those examples. Pip and npm may be insecure, and that is a fault of those tools, but most user expect secure package managers and should demand it
Telling users it’s fine to raw dog arbitrary commands directly into their shell is dangerous and lowers the bar for all security. In fact by even making this comparison you are communicating that you are complacent with pip and npm’s issues and why shouldn’t you just execute arbitrary commands without even a second glance? Security doesn’t matter!
And for the record, even with pip and npm being the way that they are, they are still better than a curl pipe because they are versioned. In the case I get a compromised deployment I understand immediately if I got hit by the affected package, and the entire repo can then be audited. Not the case when I’m just curling whatever the internet wants to send into my process space
Why is this even an app? It looks like it just generates a mobileconfig profile on the fly. Instead, seems like they could be offering a download of a pre-generated mobileconfig, which seems like it would be much safer than installing some app via curl|bash .... but then I guess there wouldn't be a chance to have an "app", get github stars, and do whatever else.
> Bash starts before the download finishes ... Drop the connection mid-transfer and you get partial execution: a command like rm -r /usr/share/program can truncate to rm -r /usr. Commands ran, cleanup didn’t.
curl | bash scripts all define a function and then call it on the last line. This is a non issue in the real world.
> The server knows you’re piping — and can lie
This `sleep` based trick is always a cool demo to show freinds yes, but the server can also sneak in malware in a multitude of other ways given you're downloading code and binaries from them.
> You trust DNS, TLS, the CDN, and the origin simultaneously. A compromised CDN or BGP hijack delivers malware silently.
Well yes, that's how the internet works. If TLS of the server is really compromised, then the attacker will replace the checksum as well as the signing key. In real scenarios, you are going to be reading the signing key and checksum from the same domain. [1]
> You can’t reproduce what ran
`| tee inspect.sh | bash`
> Add sudo and it’s game over
Most credentials and important files live in the home directory, root is a red herring. If you're running it on shared server, then well... don't add sudo.
[1] Yes of course there are legitimate usecases for signing software. Common example: linux distros which are mirrored at many domains, but the checksum and signature are hosted on the canonical domain. But if I am curlbashing uv's install.sh from `astral.sh`, then doing signature verification using the public key hosted on the same astral.sh isn't adding much.
I think that's missing the forest for trees. The problem with these curl-to-bash approaches is not that you are literally unable to intercept and inspect them with enough effort and planning.
The problem is that:
1. The effort and care needed to test is unnecessarily high. You've got to guard against way more tricks from an interactive source that can see you and choose what it's going to deliver and how.
2. With no "standard" artifact that can be exactly compared, that work cannot be shared.
In contrast, release_1.2.3.zip isn't going to mutate under you and everybody can agree on what its size/hash/bytes ought to be, and if it deviates from that it sets off alarm-bells.
> curl | bash scripts all define a function and then call it on the last line. This is a non issue in the real world.
Why would a convention often followed by good/careful actors bind what malicious/careless people create?
> If the project publishes a SHA-256 hash, use it. Non-negotiable on production machines.
They're pushing FUD around downloading a file but then suggest that we trust the same chain of complex things to display the right hash value? Integrity != authentication.
Oh, things are about to get worse with the new macOS "privacy/security" measures. They are going to curb agentic workflows even more. I don't know how Apple just finds new ways to annoy developers, but we're in a minority after all. Of 200 million Mac users, probably just up to 1 million are developers, and the rest are normies who can't tell when they should authorize or cancel the pop-up.
It’s not about privacy, it’s about kneecapping competitors, just like when they blocked the advertising ID but exempted themselves from this because “Apple is not a third-party, we’re a second-party”.
Apple is an advertising company and thus inherently untrustworthy.
They love the ones that buy Apple hardware to develop apps for iDevices, pay the dev subscription and store fees for apps, or simply because they wanted a shiny UNIX and don't consider BSD/Linux OEMs worth their money.
Really? GNOME is not what I would recommend to people.
I should applaud their efforts, and I get that much of it is voluntary, but their bugs are numerous, notable and the way they interact with the rest of the universe (both people with accessibility needs, and the wider developer ecosystem on linux) can most accurately be described as arrogant and hostile.
KDE is the bastion of maturity here, and I would agree that it is mature.
I’m not sure how the love for GNOME continues when KDE (while not my personal choice) has clearly been running circles around it since GNOME3 and the gap has only widened since that change too.
Both desktops are pretty mature. I've run both of their Wayland stacks, and Mutter/KWin both perform great these days. It really comes down to personal preference for most use-cases.
Nope. The only people who notice or care about any of this are those who can't accommodate the storage. Outside that, it all just works better now (especially Siri).
Also "those who can't accommodate the storage" is funny.
What's that? You didn't pay Apple's 1200% markup on storage, just so you can have enough room for your actual work after the OS fills your disk with a bunch of bloat? What are you, poor?
I'm sick of juggling disk space on my 1tb laptop AND I don't want an llm attack vector anywhere near my machine, this things getting nuked from orbit or i'm not updating to golden gate, ever.
What's going on at Apple product strategy?
Could the rise of LLMs and vibe-coding have motivated people who otherwise wouldn't bother?
I switched to MacOS 3 years ago because of Microsoft and the writing on the wall seems to say I got another year left before I'm forced into Linux. Because if I have to maintain my own OS then I might as well install Linux and do it once.
Huh cool. They're doing curl | bash properly.
(Note that the attestation does not appear to cover the shell script either, it only covers the script's final payload. The shell script is also referenced via `main`, so it's mutable even if the underlying payload is properly attested. That's not good!)
Does apple not realize there is an SSD crisis happening? Used to love apple (Apple IIc was my first computer). But now? Terrible.
https://nocurlbash.com/#en
If you've already decided you trust the author, what's the actual threat here?
But then again, I'm a bit paranoid. At a minimum I would download the script and read it, and if it was too long or not written clearly enough then I would just drop it and find something better.
Its a different threat model. You should not curl bash.
But I assumed the intended audience are home users with entry level macbooks/minis with 128 GB RAM where this patch actually helps them.
Telling users it’s fine to raw dog arbitrary commands directly into their shell is dangerous and lowers the bar for all security. In fact by even making this comparison you are communicating that you are complacent with pip and npm’s issues and why shouldn’t you just execute arbitrary commands without even a second glance? Security doesn’t matter!
And for the record, even with pip and npm being the way that they are, they are still better than a curl pipe because they are versioned. In the case I get a compromised deployment I understand immediately if I got hit by the affected package, and the entire repo can then be audited. Not the case when I’m just curling whatever the internet wants to send into my process space
Excuse me, they are TLS certs.
Thanks Arialdomartini, as I was saying… we need to renew the SSL Certs
curl | bash scripts all define a function and then call it on the last line. This is a non issue in the real world.
> The server knows you’re piping — and can lie
This `sleep` based trick is always a cool demo to show freinds yes, but the server can also sneak in malware in a multitude of other ways given you're downloading code and binaries from them.
> You trust DNS, TLS, the CDN, and the origin simultaneously. A compromised CDN or BGP hijack delivers malware silently.
Well yes, that's how the internet works. If TLS of the server is really compromised, then the attacker will replace the checksum as well as the signing key. In real scenarios, you are going to be reading the signing key and checksum from the same domain. [1]
> You can’t reproduce what ran
`| tee inspect.sh | bash`
> Add sudo and it’s game over
Most credentials and important files live in the home directory, root is a red herring. If you're running it on shared server, then well... don't add sudo.
[1] Yes of course there are legitimate usecases for signing software. Common example: linux distros which are mirrored at many domains, but the checksum and signature are hosted on the canonical domain. But if I am curlbashing uv's install.sh from `astral.sh`, then doing signature verification using the public key hosted on the same astral.sh isn't adding much.
The problem is that:
1. The effort and care needed to test is unnecessarily high. You've got to guard against way more tricks from an interactive source that can see you and choose what it's going to deliver and how.
2. With no "standard" artifact that can be exactly compared, that work cannot be shared.
In contrast, release_1.2.3.zip isn't going to mutate under you and everybody can agree on what its size/hash/bytes ought to be, and if it deviates from that it sets off alarm-bells.
> curl | bash scripts all define a function and then call it on the last line. This is a non issue in the real world.
Why would a convention often followed by good/careful actors bind what malicious/careless people create?
Please take a look at this before making any assertions... https://github.com/omlahore/RemoveMacAI/blob/main/install.sh
Like I get why it’s bad, but also homebrew package installation is a more organized version of this.
Hashes are cool but also in a lot of systems you’re trusting the hash to be provided by the same website you don’t trust the binaries from…
They're pushing FUD around downloading a file but then suggest that we trust the same chain of complex things to display the right hash value? Integrity != authentication.
Apple is an advertising company and thus inherently untrustworthy.
Personally, my limited experimentation with Apple AI has left me quite liking it.
The contents of the Exportable’Privacy Report’ are interesting to look through
GNOME has reached maturity and hasn't changed significantly in years, while Apple is busy destroying macOS.
I should applaud their efforts, and I get that much of it is voluntary, but their bugs are numerous, notable and the way they interact with the rest of the universe (both people with accessibility needs, and the wider developer ecosystem on linux) can most accurately be described as arrogant and hostile.
KDE is the bastion of maturity here, and I would agree that it is mature.
I’m not sure how the love for GNOME continues when KDE (while not my personal choice) has clearly been running circles around it since GNOME3 and the gap has only widened since that change too.
The problem is Apple intelligence is decent, but not worth 20% of your storage decent.
What's that? You didn't pay Apple's 1200% markup on storage, just so you can have enough room for your actual work after the OS fills your disk with a bunch of bloat? What are you, poor?