Most active commenters
  • n8m8(3)

←back to thread

1369 points universesquid | 25 comments | | HN request time: 2.003s | source | bottom
Show context
joaomoreno ◴[] No.45170585[source]
From sindresorhus:

You can run the following to check if you have the malware in your dependency tree:

`rg -u --max-columns=80 _0x112fa8`

Requires ripgrep:

`brew install rg`

https://github.com/chalk/chalk/issues/656#issuecomment-32668...

replies(8): >>45171142 #>>45171275 #>>45171304 #>>45171841 #>>45172110 #>>45172189 #>>45174730 #>>45175821 #
1. yifanl ◴[] No.45172189[source]
Asking people to run random install scripts just feels very out of place given the context.
replies(2): >>45172767 #>>45174960 #
2. hunter2_ ◴[] No.45172767[source]
I would agree if this were one of those `curl | sh` scenarios, but don't we consider things like `brew` to be sufficiently low-risk, akin to `apt`, `dnf`, and the like?
replies(3): >>45172964 #>>45174003 #>>45174196 #
3. tripplyons ◴[] No.45172964[source]
Anyone can upload an NPM package without much review. For Homebrew, you at least have to submit a pull request.
replies(2): >>45177196 #>>45182290 #
4. dmitrygr ◴[] No.45174003[source]
> don't we consider things like `brew` to be sufficiently low-risk,

Like ... npm?

replies(2): >>45174514 #>>45175021 #
5. anthk ◴[] No.45174196[source]
APT repos for Debian, Trisquel, Ubuntu... require far more checkings and bureaucracy.
replies(1): >>45174375 #
6. socalgal2 ◴[] No.45174375{3}[source]
I'll bet they don't. There's way to much churn for it all to be checked
replies(2): >>45174994 #>>45175530 #
7. hunter2_ ◴[] No.45174514{3}[source]
I thought getting code into brew is blocked by some vetting (potentially insufficient, which could be argued for all supply chains), whereas getting code into npm involves no vetting whatsoever.
replies(1): >>45182278 #
8. justusthane ◴[] No.45174960[source]
ripgrep is quite well known. It’s not some obscure tool. Brew is a well-established package manager.

(I get that the same can be said for said for npm and the packages in question, but I don’t really see how the context of the thread matters in this case).

9. justusthane ◴[] No.45174994{4}[source]
No, they are extremely well vetted. Have you ever heard of a supply chain attack involving Red Hat, Debian or Ubuntu repos?
replies(1): >>45175191 #
10. fn-mote ◴[] No.45175021{3}[source]
Nah…

Everybody knows npm is a gaping security issue waiting to happen. Repeatedly.

It’s convenient, so it’s popular.

Many people also don’t vendor their own dependencies, which would slow down the spread at the price of not being instantly up to date.

replies(3): >>45175798 #>>45177580 #>>45186549 #
11. jonquest ◴[] No.45175191{5}[source]
Yes, the XZ attack affected Fedora nightly and Debian testing and unstable. Yes, it got caught before it made it into a stable distribution (this time).

https://www.redhat.com/en/blog/understanding-red-hats-respon...

https://lists.debian.org/debian-security-announce/2024/msg00...

replies(1): >>45181120 #
12. const_cast ◴[] No.45175530{4}[source]
Churn? On Debian?

It takes like 2 years to get up to date packages. This isn't NPM.

replies(1): >>45176926 #
13. dabockster ◴[] No.45175798{4}[source]
> Many people also don’t vendor their own dependencies, which would slow down the spread at the price of not being instantly up to date.

npm sold it really hard that you could rely on them and not have to vendor dependencies yourself. If I suggested that a decade ago in Seattle, I would have gotten booed out of the room.

replies(1): >>45176729 #
14. marcus_holmes ◴[] No.45176729{5}[source]
I have repeatedly been met with derision when pointing out what a gaping security nightmare the whole Open Source system is, especially npm and its ilk.

Yet here we are. And this is going to get massively worse, not better.

replies(1): >>45179724 #
15. SchemaLoad ◴[] No.45176926{5}[source]
The xscreensaver dev managed to very easily slip a timebomb in to the debian repos. Wasn't obscured in any way, the repo maintainers just don't review the code. It would be physically impossible for them to review all the changes in all the programs.
16. what ◴[] No.45177196{3}[source]
Homebrew has been compromised before. To think it’s immune is a bit naive.
replies(2): >>45182243 #>>45183548 #
17. albedoa ◴[] No.45177580{4}[source]
> Nah…

I mean, I believe you, but the person you are replying to obviously believes that they are similar. Could you explain the significant differences?

18. Intermernet ◴[] No.45179724{6}[source]
Nothing specific to open source is to blame in this instance. The author got phished. Open source software often has better code vetting and verification than closed source software. npm, however, does not.
19. goodpoint ◴[] No.45181120{6}[source]
So the attack was successfully stopped and you complain about it?
replies(1): >>45186470 #
20. n8m8 ◴[] No.45182243{4}[source]
Agreed that it's a bit funny given the context and no community-managed package manager should be 100% trusted.

That said, I think rg is pretty well known to linux daily-drivers and they just wanted to share something quickly for powerusers who want to check their workspaces quickly. Probably better to just instruct n00bs to use grep than install a whole cli tool for searching

Come to think of it, I wonder if a 2-phase attack could be planned by an attacker in the future: Inject malware into a package, flood guidance with instructions to install another popular tool that you also recently compromised... lol

21. n8m8 ◴[] No.45182278{4}[source]
Went and found the link: https://docs.brew.sh/Acceptable-Casks#apps-that-bundle-malwa...

> Unfortunately, in the world of software there are bad actors that bundle malware with their apps. Even so, Homebrew Cask has long decided it will not be an active gatekeeper (macOS already has one) and users are expected to know about the software they are installing. This means we will not always remove casks that link to these apps, in part because there is no clear line between useful app, potentially unwanted program, and the different shades of malware—what is useful to one user may be seen as malicious by another.

22. n8m8 ◴[] No.45182290{3}[source]
https://docs.brew.sh/Acceptable-Casks#apps-that-bundle-malwa...

> Unfortunately, in the world of software there are bad actors that bundle malware with their apps. Even so, Homebrew Cask has long decided it will not be an active gatekeeper (macOS already has one) and users are expected to know about the software they are installing. This means we will not always remove casks that link to these apps, in part because there is no clear line between useful app, potentially unwanted program, and the different shades of malware—what is useful to one user may be seen as malicious by another.

---

So there might be pull requests, but Brew's official stance is that they do not actively moderate casks for malware. I guess there's something built into the MacOS packaging step that help mitigate the risk, but I don't know much about it outside playing w/ app development in XCode.

23. tripplyons ◴[] No.45183548{4}[source]
I'm not saying its immune. I'm saying that NPM doesn't have as many protections, making NPM an easier target.
24. jonquest ◴[] No.45186470{7}[source]
I’m not complaining, I’m pointing out facts. If the facts offend you, that’s your problem. Ignore them if you wish.
25. johnisgood ◴[] No.45186549{4}[source]
Convenient, as in the barrier to entry is way too low. I am pretty much against it.