FLOSS developer intentionally corrupts his libraries and has multiple depending applications print out garbage, stating that "I am no longer going to support Fortune 500s [...] with my free work."

bleepingcomputer.com/news/secu

#FLOSS #labor

Follow

@fcr If you don't want to support fortune 500s with your free work, don't publish your work under the MIT license

I can't fathom people in this thread are siding with him. This is a breach of trust in the open source world. The updates were purposefully malicious.

He was allegedly also making a bomb and set his house on fire:

abc7ny.com/suspicious-package-

· · Web · 7 · 3 · 15

@Gargron @fcr yeah, I am not siding with the developer. His actions were shitty.

I am underlining the fact that:
1. Microsoft GitHub will block your account if it doesn't like the changes you make to your own code;
2. AGPL is a way better choice of license if one doesn't want to support Big Tech.

@rysiek @fcr Regardless of if it's your code or not, if you upload malware into a widely used software package you deserve to have your account blocked.

@Gargron @fcr I do not see them as *malicious*. these were not cryptominers, no data stealing code, it just rendered the libraries unusable.

"Mischievous" is the word used in the original story, and I think that's a way more accurate description.

@rysiek @fcr It didn't just make the library output the wrong value, it introduced an infinite loop, which in my view constitutes a denial of service attack.

@Gargron @fcr I can see why you feel that way. Personally, to me it does not cross the "malicious" line -- partly because this is something that should be trivially caught in any pre-deployment testing.

We can agree that this is not an acceptable behavior for a FLOSS developer, and it is in fact irresponsible.

That said, I do think focusing on the developer's (shitty) action is less useful than focusing on the bigger problem of open-source software developers doing free work for Big Tech.

@rysiek it is malicious because the intent of the action was to harm whoever uses the project by affectively causing a DoS. There is no question in that. The motivation is what makes it malicious, it could have been a bug, if it was unintentional, but it wasn't.

@Gargron @fcr

@rysiek @Gargron @fcr, the problem isn't even big tech using the free work of others, the average JS developers don't even realize that the ecosystem is fragile, not even the managers of Big Tech projects:

github.com/facebook/react/issu

They brush over these issues as if they were a "misunderstanding" on the part of people reporting them.

I'm afraid that the Unix philosophy doesn't really work these days. You can't trust hundreds of developers and their code for the most basic JS project.

@walter @rysiek @Gargron @fcr NPM has little to do with Unix philosophy. Unix/Linux is maintained in distributions where the quality of the toolset is taken care by the core teams. See projects like Gnome or KDE. Nothing like this in NPM which resembles more a laissez-faire market.

@movonw,
True, but you do have big ecosystems that looks like de facto distributions. A default Angular project comes with 25+988 dependencies, where you get the basics, like: "zypper install-pattern foobar-desktop foobar-devel". With that you install many projects that "do one thing and do it well", including the "colors" package.

And this doesn't count projects that embed (i.e. statically link) their dependencies. And yes, strict versions are a thing and[...]
[1/2]

cc: @rysiek @Gargron @fcr

@movonw,
[...] it's not a perfect analogy, but on NPM you just get more flexibility. They don't lock you into their "repos", think DEB/RPM repos, not CVS repos. Thus, you get to have code from all other "repos", think of PackageKit+Alien, but for Angular and React "repos".

If one really wants, the per-distro repo approach can be achieved here; and things don't even have to change much. Then you're on your own if you want to "zypper addrepo" or add a new PPA.
[2/2]

cc: @rysiek @Gargron @fcr

@walter @movonw @Gargron @fcr this has nothing to do with distributions. In Debian or Fedora, or Arch, or any other Linux distro, the *packagers* are responsible for quality of the packages that are published in the distribution-specific repository.

In your example the Angular people just pull random crap from Teh Intertubes and hope for the best.

It's not even comparing apples to oranges, it's comparing apples to the number three. 🤷‍♀️

@rysiek,
yes, it has nothing to do with distros, but there could have been a resemblance.

It's wishful thinking on my part, but even NPM has the notion of registries, so it's not a long way from here to adding to packagers to the mix. The problem is that there doesn't seem to be much demand for this, and then there's npm Inc. in the mix.

deno.land is changing things, but they went with "install from src" + bigger StdLib. So... no packagers yet.

@movonw @Gargron @fcr

@walter @rysiek @Gargron @fcr completely agree that npm culture is: build awesome things! Don't care! Be careless!

@walter @rysiek @Gargron @fcr

*Wow*

It would take maybe a couple of hours to fork the repo and/or import the code into the React sources and they don't even seem to have thought of the idea.

I continue to fail to regret never getting into React.

@rysiek @Gargron @fcr

I can't say I'm surprised about nr 1. GitHub, like most of these big companies, have a "We can close your account and remove your content at any time for any reason" in their TOS. The code is still his but GitHub is not obligated to host it.

@espen @Gargron @fcr oh I agree and have recognized that for years.

My feeling, however, is that a lot of people miss that fact. And then act surprised.

@Gargron @fcr

IMO, it's not a matter of supporting or rejecting him.

It is about un(der)paid work and about King Ludd: the common misconception that Luddites were "technophobic" and against "the inevitable progress" whereas they were using sophisticated tools and attacked industrial machines to collectively bargain for better salaries.

There is no factory here -but the software industry- and no collective action. However, it is a story about "who controls technology" and about social inequality.

@Gargron The open source world could use more breaches of trust in that case, because it could just as well have been actively malicious instead, like the last three times, instead of just causing an infinite loop.

@ChlorideCull @Gargron getting angry at this guy is similar to getting mad at people who block traffic when protesting imo. His actions might not have been “right” but this is someone driven to what is arguably a mischievous act of protest by a system that is often exploitive. At some point it’s going to break down more than it already has and this is probably just a symptom of that.

@Gargron @fcr
It reads more to me like another victim of the pandemic-induced mental health crisis. Also Eugene, nice avatar, you look good!

@Gargron @fcr It was obviously a desperate move, but the developer was treated like a modern-day slave by github - something that would never happen if he had code in his private repo (gitlab, gitea, etc...). A strong argument to NOT keep your code on corporate servers.
@Gargron @fcr It's always our responsibility to audit the code. Always. No excuses.

Open source does not mean safe, it means auditable.

Also, unless you audit the code AND compile from source, it's not different to closed source.
Sign in to participate in the conversation
Mastodon

The original server operated by the Mastodon gGmbH non-profit