arcterus

I still don’t fucking understand how people voted for someone who sounds they’re having a stroke 24/7.

It’s both amusing and a little disturbing that the term for killing the mice is “sacrifice.” I’m now imagining a bunch of researchers dancing around the mice while ritually decapitating them.

My dude, the chips aren’t manufactured in the US. If the tariffs don’t apply to the chips that are inherently imported from outside the US since basically only TSMC and Samsung make them at this point, then there is no tariff at all. Companies in the US import the chips, then use the imported chips as part of their products. All the companies in the US do is assemble the imported parts (and sometimes not even that).

EDIT: Ah, there was a miscommunication. I think we’re both saying the same thing at this point. Well, mostly the same, since this doesn’t really help US companies and just drives up prices for everything.

Even without the US being involved, this sounds kinda sketchy since they’re basically just dumping some chemical on your eyes that doesn’t have real long-term studies AFAIK. I wouldn’t use something like that until it hasn’t been shown to eat your eyes or something over the course of 40 years.

I mean up until recently they were all like, “haha just wait until the list comes out and all you libtards get yours.” Of course, now that trump has made it abundantly clear he’s on it (as opposed to it just being normally clear he was on it before), they’ve split into gullible idiots who feel betrayed and cultish idiots who now think pedophilia is actually not so bad somehow.

I’m convinced you’re a troll/bot. That is not in fact how tariffs work since the chips are not made in the US.

I’m building multiple patchsets on a laptop. How tf do you expect millions of lines of even somewhat optimized code to compile in a minute or two? The configuration by itself wastes like half of that, not to mention nix taking 2 minutes to evaluate because specializations are slow af. It in fact takes more like 2-3 hours for them to finish.

Looking into it, the US implementation goes down into the components, so yes. Except, I believe it’d be $50 chip @ 100%, other components at whatever tariff rates they may have, and then the 15% per-country/region tariff applies to all of it on top. So if the other components have no tariffs, it’d be $172.50. I’m now wondering how expensive everything would end up if you have tariffs on materials as well.

In any case though, it becomes ludicrously expensive no matter what because you’re at most dodging the 15%.

EDIT: You can also dodge some of the tariffs if some percentage of the product is made in the US. I wonder if you’d be able to dodge the chip tariff if the materials for it were partially sourced from the US. If possible, that’d probably be cheaper for companies than actually trying to manufacture chips here.

EDIT 2: Actually your calculation may be right, I’m having a hard time finding how they’re actually meant to be calculated. Admittedly it seems a bit weird to me that the rate would override the country-specific rate and thus be the same for chips from the EU and China, but I suppose none of this makes sense in the first place.

Pretty sure that’s their point. Say a product costs $100 dollars with no tariffs. If you import the product from the EU with a 15% tariff, it’s now $115 with tariffs (assuming no tariffs importing the chips into the EU). If you manufacture the product in the US, you need to pay 100% tariffs for all the chips. Obviously the impact depends on how much the chips cost relative to the entire product, but if the chips are half the cost ($50), then with a 100% tariff you’re now paying $150 for the product manufactured in the US.

I like how they’ve demonstrated that their verification is basically worthless.

They’re just using company email addresses to determine what tech stack they use. This is like saying “wow a lot of companies use gmail or outlook.” In other words, the sky is blue. This is ignoring that a large number of companies use multiple different tech stacks depending on the internal organization/subsidiary. Meanwhile, proton’s products are basically all half-baked and basically don’t match the equivalent Google/Microsoft suites at all.

Lol they want $400 billion in investment in the US too. However, aside from that, I sort of wonder how buying Intel would work out for TSMC. They’d basically be buying a research wing + existing US foundries. The most important thing for them would probably be ensuring that the foundries and so on in Taiwan are the most advanced to avoid the US ditching them if the PRC decides to invade.

IIRC it took them a little while to add support, but I was more thinking of stuff like KeePass. KeePassXC has passkey support, but AFAIK none of the Android apps do yet (although it sounds like KeePassDX is getting close, finally). Also, when I was using Bitwarden, I had issues with some services not liking its passkey implementation (despite being fine with Proton Pass for whatever reason). May be fixed now, but it was incredibly annoying at the time.

I’ve never used them, but if you want streaming, you can use Moonlight/Sunshine. It’d be very cool if a project integrated everything together, so you could choose whether to download the games or stream them from the server.

TFW u wake up and decide to kill a god

I’m planning to, I’m waiting for the kernel to finish building rn lol

EDIT: PR got merged BTW (https://github.com/NixOS/nixpkgs/pull/431115).

There are select places where it’s usable, but generally speaking most merchants don’t accept it. You can sometimes buy giftcards or something like that with crypto and just use the giftcard to pay, but honestly that’s kind of a pain. Also I’m only talking about Bitcoin and Monero. Most of the other currencies I probably wouldn’t bother using.

You’re supposed to be able to use lib.kernel.unset to unset them. In any case, that’s just one problem. The main issue is the entire option is ignored because of a typo in nixpkgs.

I’m gonna be honest, I use NixOS, but the docs fucking suck, and a number of things are just broken in nixpkgs. For instance, I recently discovered the structuredExtraConfig option for patching the kernel straight up does not work. This means you cannot unset any kernel options, which means some kernel patches won’t work unless you manually supply the entire kernel config.

EDIT: what’s even more annoying about it not working is that it fails to apply silently. In other words, your kernel tries to compile and then an hour later it fails because your config changes weren’t applied.

The goal is basically to prevent end users from using weak passwords and to make it much harder for phishing to occur, both of which IMO are kind of necessary. The vendor lock-in and the slow development of FOSS implementations are not great though. It’s also not great how passkey support on at least Android seems to require proprietary blobs.