still can’t change text after applying a filter in $current_year

I’d prefer a list of projects that publicly abuse their users… which ironically still includes Linux and *BSD.

At least Haiku is actually daily drivable if your hardware is supported.

It’s on the list, and they probably used Linux to write the entire site.

How does a duress pin prevent them from taking the device into custody?

And if you forgot you had a duress pin, hence weren’t “knowingly” damaging anything, would that be a useful defense?

My point was that

That’s all fine and dandy, I just don’t agree with simultaneously saying “make it indistinguishable if you’re gonna use AI” in addition to “don’t use AI”, since as you say there is no effective difference if you can’t tell, so why should it still be disallowed? Surely we can come up with some middle ground to describe the intent of the rule better than silently implying “just don’t get caught”?

or just use SimpleX

It’s getting harder to tell the difference though. Slop artists can still prompt their agent to follow the codebase’s existing style/format.

Some projects are already quietly saying “just don’t tell people it’s AI” and to make it look indistinguishable, then get on with life.

Can’t say I have any good rebuttals against that tbh.

What I’d like to see is a worldwide network of fast and reliable TURN servers that any WebRTC app can safely use.

Me and everyone else I know who are behind CGNAT/symmetrical NAT always have problems using any WebRTC app due to lack of good TURN server availability and configuration options in the specific app.

Black and white generalizations might be the worst sin of all.

I think the JP fonts are thinner so they don’t take up as much space.

No mention of Laya, which is open-source, can actually be used today, offline, and there is a vision model as well.

Reminder that if you can produce evidence that they didn’t, you can sue them.

I never said they were doing it or that people should assume it, just that it’s possible. Whether or not that matters is up to each individual’s own threat model and should be weighed accordingly.

“Bot” there doesn’t necessarily imply a malicious spam/AI bot… legitimate B2B use is also included, which is a much higher percentage than illegitimate traffic. Same for AWS, other cloud providers, and even Cloudflare itself.

Bot traffic describes any non-human Internet traffic

https://radar.cloudflare.com/bots

I appreciate the sentiment, but I think the reality is the number of people who are so upset over this that they switch providers… is a rounding error.

They can… as I said they control the entire OS. They can see you’re running Signal with account# XYZ, that you clicked the pay button, that Signal generated the ZKP (which they can read in memory or network packets), and all the key clicks and numbers entered of the payment details for that ZKP, the entire process.

they can’t link the account

Google can, since it’s their OS and their payments.

But even if you’re not using Play Store to pay, the vast majority of people who are paying are going to be using the same device/IP/fingerprint/etc. where Signal is running, so it’s not impossible to put the two together to link an account to a specific payment, if say, a government forced them to do so in order to track someone. It’s just a bad idea for several different privacy reasons.

there would be if it allowed unrestricted free accounts with zero barriers to entry, same with anything else like that.