Live data from Hacker News

"Localhost tracking" explained. It could cost Meta €32B

zeropartydata.es

181–190 of 286 posts

Re: "Localhost tracking" explained. It could cost Meta €32B

#181
post #176
post #153

Earlier quoted context omitted.

They knew who was going to be president this year.

The American president doesn't really matter in this case. The EU is where they're going to get destroyed.

It’s quite possible that a different administration would punish FB for this. The current president being who he is, is the reason it doesn’t matter.

Re: "Localhost tracking" explained. It could cost Meta €32B

#182

Earlier quoted context omitted.

File mode bits prevent processes not running as root from reading much of the info in /proc.

I don’t know… with a stock Linux, the information a user can get from top (via /proc, I assume), is pretty thorough. You can at least get a list of running programs, which by itself could be valuable.

Good point. I withdraw my comment.

Re: "Localhost tracking" explained. It could cost Meta €32B

#183

So I am seeing two issues here. 1. Android allows apps to open ports without permissions. And apps to communicate with each other without permissions. 2. The browsers allow random domains to access services on the localhost. Without notifying the user. We have seen vulnerabilities in the past accessing dev services running on localhost. Something should be done there.

I'd split that first list into two: 1a. Arbitrary apps can listen on ports without permissions. 1b. Arbitrary apps can access local ports without permissions. I've recently been experimenting with running the browser (on my desktop) in a network namespace precisely because of these reasons. Random websites shouldn't be able to access services running on localhost.

uBlock Origin ships with a "Block Outsider Intrusion into LAN" filter that I believe is enabled by default. I don't know if it works on the neutered Chrome version, but on Firefox it works so well I've had to add a few whitelists for cases where I do want access to LAN or localhost.

Re: "Localhost tracking" explained. It could cost Meta €32B

#185
post #110

Earlier quoted context omitted.

Sounds like a typical blurring of responsibility through bureaucracy. "If Zak is a billionaire, then he is responsible, but since he essentially did nothing wrong, then no one will be held accountable." Total nonsense. There are specific crimes, and there are specific people who planned this crimes, specific peoples who ordered them to be carried out, and who carried them out. And these people should be held accounta…

> if your employees, without your knowledge, somehow decided that the only way they could reach their targets was to commit a crime, why should you be held responsible for that? Thats where "known or should have known " becomes relevant. It's your company, it's your responsiblity to know what they are doing.

No, what you are suggesting is a typical strategy of avoiding punishment and creating an opportunity to break the law. A very common strategy, used everywhere, especially in dictatorial and socialist regimes.

There is a substitution of one real crime, committed by real people, for a crime "they didn’t know, but should have" against other people, for which there is no real responsibility, while the real criminals are declared to be simply "cogs" in the system.

As a result, no one is held accountable for a crime for which dozens of people who directly committed it could go to prison for many years, because the person held responsible is a high-ranking manager who "should have known, but did not know," who himself issues "a severe reprimand" or assigns a tiny fine for it.

Thus, the entire system is drowning in crimes, the commission of crimes becomes a REQUIREMENT of the system and the commission of crimes becomes a guarantee of the loyalty to the system.

Re: "Localhost tracking" explained. It could cost Meta €32B

#186

This system was designed and implemented by engineers who committed code in a source control system with their name attached, and the changes were requested by product managers in tickets in the ticketing system with their name attached. Those engineers and product managers should be personally liable for an equivalent % of their annual salary as Facebook is liable for a % of its annual revenue.

Sounds like the modern version of the CS Lewis quote:

> The greatest evil is not now done in those sordid dens of crime that Dickens loved to paint. It is not done even in concentration camps and labour camps. In those we see its final result. But it is conceived and ordered (moved, seconded, carried, and minuted) in clean, carpeted, warmed and well-lighted offices, by quiet men with white collars and cut fingernails and smooth-shaven cheeks who do not need to raise their voices.

Re: "Localhost tracking" explained. It could cost Meta €32B

#187

Earlier quoted context omitted.

AND, whenever you suggest here that engineers should consider the morals or ethics of what they are being asked to work on, you often get lots of push back in the comments. "I just want to work on cool tech! It's my company's problem what they use it for!" and "Hey, I'm just a code monkey, don't blame me! If my manager tells me to build the Torment Nexus, I build the Torment Nexus!"

This is one of the main reasons I’m for licensing software engineers like civil engineers are. You know that without a license, you can’t work in the civilized world. So when your license requires you to not build the torment nexus, and some manager comes and says “build the torment nexus” then you tell them no, knowing that they can’t just fire you and hire someone else to do it. Yes, they might outsource it, but yo…

You don't want a licensing requirement in software engineering. That attempts to solve the problem in the wrong place entirely. The problem is that it's legal to build the Torment Nexus.

Licensing would raise your costs and restrict your choices, while having absolutely no effect on issues like what's being discussed here. You would just get a more expensive Torment Nexus that may or may not be slightly more secure.

Re: "Localhost tracking" explained. It could cost Meta €32B

#188

Earlier quoted context omitted.

AND, whenever you suggest here that engineers should consider the morals or ethics of what they are being asked to work on, you often get lots of push back in the comments. "I just want to work on cool tech! It's my company's problem what they use it for!" and "Hey, I'm just a code monkey, don't blame me! If my manager tells me to build the Torment Nexus, I build the Torment Nexus!"

Some time later on HN front page: > Why I left FB,GOOG,Whatever >> Author describes seemingly abhorrently unethical and immoral practices they were completely ignorant of, occurring right in front of them that they were a key participant in. >> Accepted a massive salary to be ignorant. >> Shocked as all fuck about ethics and implications. >> Returned 0 money, cashed out. >> 100% ethical now.

A tale as old as time…

Re: "Localhost tracking" explained. It could cost Meta €32B

#189

This system was designed and implemented by engineers who committed code in a source control system with their name attached, and the changes were requested by product managers in tickets in the ticketing system with their name attached. Those engineers and product managers should be personally liable for an equivalent % of their annual salary as Facebook is liable for a % of its annual revenue.

Let’s be real, the people who are culpable are truly culpable are the ones who gave them the ok to build this in the first place.

Re: "Localhost tracking" explained. It could cost Meta €32B

#190
post #92

Remember in 2014 when the Android Twitter app started sending a list of all your installed applications back to Twitter? https://news.bloomberglaw.com/privacy-and-data-security/twit... Ever since then I refused to install native versions of apps that could be used in a browser. I don't use Facebook or Instagram so I don't know if that works anymore, and I recall testing that they were intentionally crippling Facebook…

I felt a prude at the time but eschewed native apps for browser versions and haven't regretted. Didn't benefit from notification distraction anyway. Apple and Google just didn't get their houses in order to be taken seriously. If it ain't on F-Droid, I'll wait.

There is another can of worms hidden in plain sight right here, I feel like.

From the article:

  You’re not affected if (and only if)
  You access Facebook and Instagram via the web, without having the apps installed on your phone
This is only what's observably true of a particular app under the hood from straightforwardly jacking into it with Frida or performing any other deeper analysis.

What's to say Meta/Google/OtherAnalyticsCorp/OtherMegaCorp hasn't already, on a large scale, colluded with[bought out] app developers to simply share session data out-of-band as another tentacle?

Rather, is it even reasonable to assume they all haven't been doing this all this time? (Maybe these also fall squarely under what GDPR, DSA, and DMA were supposed to mitigate? I'm not an expert here.. just my cynicism kicking in.)

I too go through fairly great pains to try to minimize unneeded apps on my device.

Post reply on HN