Amju Wolf

Algorithmic patents amount to patenting maths which, by very longstanding precedence, is not a thing, for good reason.

You absolutely can patent “math” (well, more like physics) IRL. What matters though is that the invention actually has to be novel and non-obvious, and IMO it should also be harder to patent if it’s in a segment like software where costs of development, iteration and “research” are generally extremely cheap. Like, it should have a way higher bar for the “novelness”.

And I would not allow any kind of software design patent (use copyright or trademark to protect that).

…and that’s how it still works.

Considering those scanners usually register as keyboards and just do a dumb input “typing” it shouldn’t even be hard, lol.

The worst part would probably be picking a barcode scheme where the inputs make sense…

That sounds very illegal, yeah. You can’t advertise a price and then charge something different. It doesn’t matter that the person didn’t notice it. At that point you might not have price tags at all (which is also illegal, just FYI).

It generates code and then you can use a call to some runtime execution API to run that code, completely separate from the neural network.

On the contrary, it’s the only comparison you can make, since they are literally the only options.

…and there is no way to do that, currently.

That’s not something that’d likely scale enough to bring any meaningful sum of money.

Even then it targets a tiny, tiny minority of their even current userbase, let alone if they want to approach more “average” users.

They’re two separate(ish) issues.

But it’s still a bad idea to use national TLDs for stuff that has nothing to do with that nation.

Granted, is ICANN wasn’t just a money-grabbing machine with no forward thinking they wouldn’t give nations clearly “generally desirable” gTLDs, but since they did already that doesn’t mean they should be misused.

I had a similar issue and in my case it ended up being some AMD crap (I think an updater or something) that probably didn’t install properly or something.

IIRC I just ended up disabling the scheduled task that was running it and that was the end of it.

That would give random strangers (at least partial) control over what is indexed and how and you’d have to trust them all. I’m not sure that’s a great idea.

Sounds like cruel and unusual punishment to me. I wonder if he could still fight it. Sure doesn’t sound too bad until you realize that it’s literally only to make an example out of someine while ruining his entire life for a crime with no victim.

For the people who do find out about it and it hooks them enough sure, it’s not really forgotten or underrated. But I still think it’s kinda obscure / not well known?

Where are my Outer Wilds boys at?

Firefox has a profile manager (the thing that’s also exposed to about:profiles). Run it like firefox -profilemanager and you’ll get a profile switcher.

Run firefox -profilemanager -no-remote if you want to open multiple different profiles at once (only the original one without “no-remote” will open new tabs when you click on links outside the browser). You’ll probably want to make a shortcut for different profiles though, not sure from memory what it is (but probably -profile ProfileName) and then you can easily use profiles.

The support is actually pretty decent, just kinda hidden. You don’t get a profile switcher because the browsers are completely separate, they don’t really know about each other.

As easy as it is to hate what he did to Twitter I wonder how much of that original value was just “VC hype” and now people see it for the failure it has always been like all social media in a way. Turns out it’s pretty hard making money out of non paying users.

Well a good reason could be that it brings federation to the masses. You know, like everyone who uses federated networks wants it to be. This isn’t some exclusive club and wider adoption is a good thing.

If only to prove that it can work.

Ehh it’s not that simple either way.

Like, platforms don’t actually own your data and usually explicitly state so; if for no reason other than not having liability for what you post.

If they did actually own the data (beyond having the very broad license to use it) they’d also have to curate 100% of it, otherwise they’d get sued to oblivion by copyright holders and whatnot.

You’re completely wrong.

This means that they will implement it, and then it’s only a tiny change to make it available everywhere if they decide to do so later.

The option alone also now also allows people to build stuff that will only work in those WebViews, rejecting to work without the integrity check, which is already a huge loss.

Do these actually work against HDCP? (Outside using a camera, obviously). I know it used to work decently well against most “ordinary” attacks like VMs and capture cards.