inexcf

All this shilling and promoting a "software factory" and you decided to write it in javascript. I won't infect my repos with an npm mess.

Ah okay, so your pedantry about it being USDC and not Bitcoin was unnecessary and "it exists and you can use crypto to pay" is in a made-up fantasy world where you can in fact do that.

Like can you pay with crypto at a Stripe POS when the infrastructure goes down? Or was this just completely irrelevant?

The point is that the look is not inherent to software rendering. You're misinterpreting this as some kind of denial of historical association. You're own quote there identifies texture filtering as the difference. Explaining that you can get those sharp textures with hardware acceleration, isn’t “pointless nitpicking.”

This is ridiculous. He explained it perfectly and he was right stating that it has "nothing to do with software rendering". The fact that people specifically chose to do it differently on hardware makes nothing about what the article mentions specific to software rendering. Otherwise if you do the smoothing on the CPU you have to call it a hardware renderer now to be consistent.

Well seems like Bitwarden is dying. A clear move towards enshittification. I was fine with the premium subscription existing while i was self-hosting Vaultwarden, but now every step seems to make that worse. Now new features will be under the commercial license an everything else will be slowly neglected. Time to jump ship.

Ironic that this is hosted on a site with a blatantly illegal cookie popup. Not gonna read that shit.

Weird. I don't use the official harnesses by OpenAI or Anthropic but with all others i tried there wasn't a single one that hides any of that. What harnesses did you compare to?

Factually wrong. I mean how can 20x be about 5h usage when 5h limits were not part of the Pro plan for a long time.

That's missing the point, or supporting it. That is brand recognition/knowledge and not a factor in decision making. Knowing three tire brands doesn't tell you if i would buy any of them because of their ads. Personally i never bought tires, mine just are whatever "Mercedes Original" branded ones the car comes with. If i had the need to buy new ones i would get some recommendations. I would not go "I'll buy X because i saw the name a few times". It's hard to think of any buying decisions i make where ads could matter. When i buy a phone i decide based on features with the highest priority being that it runs GrapheneOS. When i go shopping for clothes i don't look at the brands at all, and i avoid any pieces with visible branding on it. I can't think of a point in my live where i picked some product because i saw it or the brand in an ad. The subconscious part is so hard to measure and disprove but it is just as hard to prove it works as advertised. Ads might make you remember a name, but i am questioning how much that affects my buying decisions, especially when it makes me trust the brand less when they have to fill every ad slot for people to even talk about them.

I doubt it is brand recognition and more an abstract kind of marketing you mean. Esoteric word of mouth.

And homeopathy works because people pay for it.

I specifically avoid products i see ads for, cause the product can't be that great if it needs obnoxious advertisements, so the effect is more than subconscious. Now, i don't see many ads cause i block all of them where i can and roadside advertisements are not allowed here in Germany, but even if i was, i am of the opinion that the effect of this subliminal manipulation is severely overstated by the people selling ad space. I think the idea that i would buy something just because a company writes their name everywhere is ridiculous and as far as i know studies show that the effect on buying decisions is basically not measurable. I think the entire value in advertisements is just more people in general knowing of a brand.

I did that exact thing with GLM-5.3 from Z.ai with a chinese IP Camera. And i did not have to trick it in any way.

Why does this look exactly like https://pairdrop.net/ ?

Is there poop in nature? Well, does a bear shit in the forest?

Nah, this one stops you while you are scrolling and only lets you move on if you stop for a moment.

God this site has atrocious scrolling behavior. Allowing me to scroll halfway to the next "page" but then forcing me to stop scrolling and start again to get to the next one made me close the page half way through.

What about egui and wgpu is not native? Those are both generally what is considered running something natively. Both even mention running "natively" in their descriptions.

I guess what you mean is "OS-provided controls". In that case, ye that is not what this is.

What requirements does it not follow?

>(root)

GrapheneOS is not rooted.

Nah, Cell Broadcast has a lot less overhead since it does not have to do any routing or delivery confirmation and such. The load does not scale based on recipients which makes it far more reliable during an emergency.

Pretty weird when the moon is as much between earth and sun as the earth is between the moon and the sun.

Point 1: I said "to an extent" but i concede that it is nowhere close to 10%, but isn't murder state law? The FBI does not step in for that or does it? And the homicide clearance rate has dropped to 50% (72% in 1980) in recent years, for half of the murder nobody is ever held accountable.

Point 2: I was thinking that the most heavily armed private citizenry counts as "private security". More guns than people. Also far more private security guards than cops. To me that is exactly what "private security everywhere" looks like.

Point 4: To be honest i think point 3 actually applies to any place anywhere. Someone who knows the local police gets away with far more shit in my small german village. But "doesn't apply to anyone who is white" clearly is a sign of lack of rule of law. Laws should be race-independent and statistics show that in the US it's anything but.

Point 5: No chaperone-level warnings, but the UK, Canada, Germany, Australia, Japan all issue standing travel advisories about US gun violencem, the UK literally tells citizens "try not to walk through quieter areas alone, especially at night." And there were recommendations i got against business travel to the US especially with company issued equipment.

I mean, man, currently the US is not looking good from my point of view. I have a bunch of american friends who say immigration and everything else is ruining Germany. But looking at the stats it's looking quite okay:

Germany vs US: - Murder: 0.91 vs 5.7 per 100k

- Rape: 14 vs 40 per 100k

- Violent crime overall: 253 vs 380 per 100k

- Murder/homicide clearance: ~95% vs 50%

- Private security guard to police ratio: 0.9:1 vs 1.9:1

- Police killings: 10 vs 1100 per year

- Incarceration rate: 67 vs 600 per 100k

The highest incarceration rate in the developed world, ~4% of the world's population, ~25% of its prisoners. Are Americans just incredibly prone to being criminals? No. The crime rates above show the violent crime gap is ~1.5x, not 9x, and property crime is comparable or even lower than Germany's. ~95% of convictions are plea bargains extracted under threat of a "trial penalty", hundreds of thousands sit in jail pre-trial because they can't afford cash bail, and sentences run far longer for the same offenses. The US locks up 9x more people because it chooses to, not because Americans are more criminal. A state that imprisons its own population at Cuba and Rwanda rates while half its murders go unsolved, sells bail to the highest bidder, and runs on guilty pleas instead of trials. I would not correct anyone if they told me the US was not governed by the rule of law.

Man while reading your points i was thinking your last sentence will be a sarcastic note that implies that these points all do actually apply to the US. Of course the US as a whole is not comparable to other parts of the world where these points absolutely apply but points 1,2,4 and 5 seem to apply to a significant extent.

"last country to abolish slavery" vs. "last country to practice slavery"

> The reason the 30% rate is sustained though IS the abuse of market power with the PMFN clause, which ensures that competitors cannot price lower.

Steam's official price parity policy applies only to Steam Keys, which are free to generate and cost Valve nothing. The actual written rule, from the Steamworks documentation, says: "It is important that you don't give Steam customers a worse deal than Steam Key purchasers." It even allows discounts at different times on other stores, as long as Steam gets a comparable offer within a reasonable period. There is no written policy preventing a developer from selling a separate, non-Steam-key version of their game cheaper elsewhere. The claim that Valve informally enforces broader price parity beyond Steam Keys is the unproven allegation at the center of the lawsuit.

And the broader claim, that lower commissions would lead to cheaper games for consumers, has already been tested and disproven. Epic exclusives like Borderlands 3, Control, Metro Exodus, Phoenix Point, and The Division 2 launched with no Steam version at all, complete pricing freedom, and a 12% commission. They were all priced at full retail. The hypothetical £35 game you describe has been possible for almost a decade. Nobody has produced it because developers and publishers set prices, and they choose to maximize their own margin regardless of what the platform takes.

> We will give you free matchmaking - just not for players who have bought on another platform, so if players buy on another platform they can't play with their friends.

Steamworks matchmaking requires a Steam account and Steam does not force developers to use Steamworks. This is how every platform-specific networking layer works: Xbox Live requires an Xbox account, PlayStation Network requires a PSN account. They offer a free matchmaking service if you want other platforms use other platforms. But i guess it is because of:

> Everything else is an attempt to stay the primary shopfront.

Yeah this sounds a lot like:"How dare they deliver a better service in the pursuit of user retention."

> taking payments, serving games, having refunds(!) and adding a chat panel isn't big and complicated in 2026

This lists surface features and ignores the infrastructure: global CDN serving petabytes of downloads, payment processing across dozens of regional payment methods, fraud prevention, Workshop mod hosting, cloud saves, anti-cheat, regional pricing tooling, Proton development, Steam Deck compatibility, discovery algorithms. But the real test is simpler: developers are not forced to use Steam. If the value were as thin as claimed, the lower-priced competitors would be winning. They are not.

1) This is a legit point although i don't see Valve as a big problem in that area. They invented lootboxes but refused to be as bad as others who followed them. Today with valve these things are restricted purely to cosmetics.

2) is weak. The 30% rate was set in 2003 when Steam had zero market power and was identical to the rate used by Apple, PlayStation, Xbox, and Nintendo. Valve later added tiered reductions (25% above $10M, 20% above $50M), bringing the effective average to ~24%. The rate moved downward while the platform added massive infrastructure: free multiplayer matchmaking, cloud saves, Workshop, Proton, anti-cheat, global CDN, refunds, and community tools. The 30% buys far more today than it did two decades ago.

Developers can also generate unlimited(or i think now limited to some ratio of steam sales) Steam keys and sell them anywhere else with Valve taking 0% commission. If valve were extracting monopoly rents, this escape route would not exist.

The actual lawsuit targets the PMFN price parity clause not the commission itself. And on PC, which is an open platform where Epic's 12% store gained roughly 3% share in seven years despite hundreds of millions in investment, the monopoly framing falls apart. That 12% is also based on EPIC using lots of anti-competitor and anti-consumers tactics and using Fortnite money to prop up the store.

>As I said in a different thread[0], the fact that some subscriptions are predatory doesn't mean subscriptions are necessarily predatory.

"Some" is doing a lot of work in that sentence. Subscriptions are structurally designed to extract payment for non-use. the whole point of the model is to be predatory.

You pre-pay for some allotment of resources but get nothing back for unused resources. And the main profit comes from the gap between what you paid for and the actual usage. And that's just in the best case scenario there the company is not running at a loss just to squeeze you later even harder. The entire model only works if people throw their money into the void.

If a provider genuinely wanted to cover the costs my usage imposes on them, they would bill me for usage. That's the honest version: you used X, here's the bill for X. Instead, subscriptions deliberately decouple what I pay from what I consume, and they always decouple it in the vendor's favor, never mine. I never (well i think Kagi[0] actually is different, one outlier) get a refund for the month I didn't watch anything or the API calls I didn't make.

[0]: https://getlago.substack.com/p/why-kagi-launched-no-use-no-p...

By that definition, every write() call that doesn't check for EAGAIN is a concurrency bug you're racing the disk controller. The term stops meaning anything.

Yeah, the one most languages (except for Rust)* decided was not a language problem and did not fix.

*should clarify, Node.js, PHP, and Haskell did ship patches. Python, Ruby, Erlang, and Go opted for documentation updates; Java went "won't fix."

Isn't that like saying there can never be a language with safe concurrency since the code could interact with C code that segfaults? I dunno this kinda reminds me of the 10/10 Rust CVE that turned out to be cmd.exe on Windows not sanitizing inputs and languages like Java just labeled it "won't fix".