handsclean

“So you’re pro assault when somebody’s broken into your home at night, but suddenly anti assault when I want to punch your grandma?” Exaggerated but the same idea. Though people often communicate and maybe even internalize it in simplified “copyright bad” form, actual beliefs are much more contextual. The piracy debate would look a lot different if it weren’t literally millionaires demanding money from children.

Before the rest of these abuses, Honey was blatantly tracking users and selling that data, which I think is a good example of how privacy abuse is often a canary of generally immoral behavior.

People aren’t responding to their own assumption that AGI is necessary, they’re responding to OpenAI and the chorus constantly and loudly singing hymns to AGI.

> You can’t say everybody’s a crackpot.

Sure can. Whole world’s done this every time there’s a major advance in algorithms. We do this with other major advances, too, like how the industrial revolution was going to usher in utopia and GMO was going to end world hunger. Whenever we can’t see the end of something, only half the world figures it’s a vision problem, while the other half figures the end must not exist.

> It's hopelessly behind on many modern Web standards, too.

Let’s put this narrative to rest, please. Web standards that Safari skips are usually PWA stuff that Google unilaterally declared a “standard”. On actual web features, Safari is behind only about as often as Chrome is, and both (and Firefox) are so far from bad it’s ridiculous to complain about. WEBP was the exception that was just very late, but in hindsight it wasn’t so good after all, which Apple might have known, and anyway at present JPEG XL is shipped by Safari, held back by Chrome.

What I do think is real, though, is developers living their lives in Chrome then only testing in Safari, which makes wherever Safari lags bite them in the ass, and wherever Chrome lags not even noticed.

I’d love to see these comments about Safari lagging give specific examples. Every time I’ve seen specifics, it’s either only interesting to progressive web apps, or blatantly user hostile tracking/nagging “features”. In my personal experience as a web dev, Safari is often the first to implement new features, and otherwise lags literally just a few months behind, according to their release cycle. WEBP was the exception that was both a real feature and very delayed, but now with JPEG XL it’s Safari first by a mile and Chrome holding everyone back.

People occasionally post numbers and IIRC the front page peaks at ballpark 1 req/s.

They sold to Canva, which is strictly subscription based. As far as I know there haven’t been negative changes yet, but we’re still in the period before we’d see them.

Man, I don’t think this will be good for users.

The biggest shortcoming of Pixelmator is its lack of Windows support. This rules out use in most of the professional world, not because one must run Windows, but because one must collaborate with others. Pixelmator has long been Apple-centric, but while previously I’d hoped that, in the right situation, they might expand their strategy, now I can’t imagine I will ever be able to use Pixelmator for work.

Its second biggest shortcoming is the plugin ecosystem’s apathy towards it. Apple doesn’t have it in their DNA to fix this. Apple’s developer relations strategy is to own a lucrative enough audience that developers will endure anything for access to them. Apple doesn’t own the audience for professional image editor plugins, and I can’t imagine them suddenly learning a whole new mode of interacting with developers.

Additionally, when a company acquires a much smaller one, they really don’t care at all about the smaller one’s business, they care about how their existing business is affected. For example, when Apple acquired Dark Sky, they transplanted the features that fit into their existing strategy, but they weren’t interested in crowd sourced data or Android weather apps, so they just deleted it, and now the world’s weather forecasts are worse. Maybe, hopefully, Apple believes their walled garden’s value will be increased by the addition of a Pixelmator-like product. But I fear it’s more likely they just want to stick layers in Photos, delete the rest, lose every Pixelmator customer, and cry a fraction of a tear equal to Pixelmator’s profits divided by their own.

Affinity sold out, too. I don’t know where to go at this point.

The first boot after a macOS system update has long been bugged out. It launches a bunch of apps you didn’t even have open before updating, seems to be the 5-10 most recent apps you quit. Yes they were fully quit, yes I have the “resume” setting off. It also doesn’t do a resume, it launches them, i.e. tells them to create new windows, and it launches them before it finishes mounting disks, resulting in every update being followed by all my most used apps appearing out of nowhere and telling me all my config and data is gone. It doesn’t really matter, you just reboot again and you’re good, it’s just careless and makes the OS feel unstable. Maybe the firewall thing is unrelated, maybe it finally forces Apple to fix the bug, we’ll see.

Ok, how would you use Playwright and Nut.js to discover the OP’s “splinter”? Note I’m specifically asking about discovery, not testing for it once you already know what to look for.

I’m not sure I read this right, but I think this feature is headed for the $10/mo plan, and currently exclusive to the $25/mo plan only as part of that plan’s early access to new features.

Hire patsies. Motivate them by telling them they’re entrepreneurs. Every time somebody notices, act really shocked and offended, smack your patsy around (just the one they know about), coddle the customer (just the one who noticed). Bribe politicians to look the other way.

I’ve previously noticed that Claude is far less apologetic and more assertive when refusing requests compared to other AIs. I think the answer is as simple as being ok with just making it more that way, not completely that way. The section on pretending not to recognize faces implies they’d take a much more extensive approach if really aiming to make something never happen.

Normalizing to CPI gives you a decent comparison of cost, normalizing to industry gives you a decent comparison of competitiveness. CPI is the correct choice, industry-normalized would fail to show changes like the whole industry jacking up prices.

Soon:

“Your honor, it may be true that my client’s driving speed in combination with the thick fog prevented him from reacting to obstacles, and that his car then struck and violently killed this man while he used the crosswalk. However, it was not the fog or my client’s speed that caused standard crash avoidance safety mechanisms to fail, but the crash-ee’s negligent decision to go outside without a phone with a functioning and active location beacon.”

Right: you retain ownership, and grant X certain rights to the content. Whether those rights include training AI on the data is legally and morally in dispute. X claims that right in its ToS, but a ToS isn’t law and may be legally invalid, and besides that the ToS system is famously broken in the US. Morally, I think it’s pretty clear that reasonable users did consent to their content being published as a tweet, and did not consent to X recreating the content as their own and taking credit for it.

> He owns Twitter, why wouldn't it be okay for him to train on Tweets?

Because he doesn’t own the tweets. Can you imagine if posting a photo you took to Twitter meant it’s not your photo anymore? Totally ridiculous.

That’s just redshift. Redshift alone wouldn’t be evidence of expansion, just relative speed, when people say redshift evidence they mean the relationship between redshift and brightness of a standard candle. And regardless of whether you call it redshift or the relationship between redshift and something else, it would be impacted by a change to redshift.

It’s not just phlogiston, it’s the lifecycle of all scientific theories that they’re used for as long as they make accurate predictions, then we start seeing things they mis-predict, then they’re revised or replaced. You seem to think the expanding universe theory can still be saved by some data artifact or parameter tweaking, but that’s been hunted for years and we’re still at “we just can’t make it match everything we’re seeing”. Historically, that’s what precedes significant revision or replacement.

> Redshift is not the only observation we have.

What else is there?

Doubt that’s it, they would make more disclosing the affiliate links as a way to donate or requesting donations directly. I’d guess it’s just normal for them, or maybe inserted automatically, or maybe somebody was amused by it.

The difference is whether it can be fixed. An expired domain is hard because we don’t actually know the attacker isn’t legitimate until they misuse the domain. In the case of hosting hijacking, though, the domain still has a legitimate owner, and the hosting provider is giving somebody else access without that owner’s permission. I’d blame the domain owner if they pointed their domain off at some malicious host, but the point here is that it’s hosting providers we consider trusted and legitimate that are doing this, and they can stop doing it, they’re just choosing not to.

Timestamp? I’ve watched the video twice now and read the whole article and can’t find one mention of it. The closest is one statement of “it keeps all that force still there”, which is vague, definitely not the same thing as CFM, and sounds like the sort of statement that’s compatible with the aforementioned nozzle lines.

There’s some other “quiet leaf blower” on the market, and all they’re doing is reducing power and telling you special nozzles make up for it, which of course they don’t. This article omitting performance measures is a red flag for me. What’s the CFM impact? It’s 40% quieter, but is it also 40% worse at blowing leaves?

I can’t tell what you’re criticizing. Yes, computers get faster over time, and future computers will be faster than the M4. If release cycles are offset by six months then it makes sense that leads only last six months in a neck-and-neck race. I’d assume after Arrow Lake and Strix Point the lead will then go back to M5 in six months, then Intel and AMD’s whatever in another six, etc. I guess that’s disappointing if you expected a multi-year leap ahead like the M1, but that’s just a bad expectation, it never happens and nobody predicted or claimed it.

No, I’m quick to come to their defense because this law is wrong and immoral.

“One has a moral duty to disobey unjust laws.” —MLKJ

All that international trade and banking you talk about refers to it as USD or US$, not $.

To 95% of the world, the relevance and influence of USD isn’t remotely close to their own currency’s. Sure, it’s relevant to some industries and to the larger economy – that makes it like CNY or EUR are to you, nothing more.

I do think a USD assumption is reasonable, but only because the statement was made on HN, which is predominantly American. USD does not own the $ symbol outside of the US.

After a quick chat with support, Spaceship also supports neither DMARC/DKIM/SPF on website-sent emails, nor automatic backups. They recommended using a third party plugin to connect the web hosting part to the personal email hosting part, but even then you just get SPF and DKIM, not DMARC.

I used and was impressed by Namecheap EasyWP, which seems to be part of what they’ve folded into Spaceship. They really delivered on high performance, serverless, cheap WordPress. Moreover, they proved that Namecheap doing greenfield produces starkly better results than what I’d come to expect from their main website and last-gen hosting.

However, EasyWP also ultimately lost me because they just never implemented DMARC/DKIM/SPF or automatic backups. I’m on board with necessary cuts to legacy and ecosystem compat, but missing these features isn’t that, it’s just an incomplete website – and solving it with plugins and manually synced external services significantly detracts from the value prop. And I’m not seeing either feature explicitly advertised on spaceship.com.

I highly doubt mass delusion is real. What other illness is always diagnosed not with supporting evidence, but lack of evidence for alternatives? It’s a bizarre leap given the known incompleteness of medicine, that only gets more bizarre when you dig into specific examples: it happened before the discovery of pathogens, yet we assume doctors didn’t miss a pathogenic cause? Everybody who drinks a certain batch of Coca-Cola gets sick, but if it’s not a contaminant that was tested for, it must not be a contaminant? It spreads more between people who spend more time together, and this is considered evidence is spreads via psychological closeness?

No, this is the same old bullshit: doctors, like politicians and businessmen, depend on an image of expertise, and an assertion of mass psychogenic illness is a non-disprovable way to preserve it when evidence-based diagnosis can’t.