rappatic

As my friend likes to say, Michael Burry has predicted 20 of the last 1 crashes. I saw some apocryphal analysis that showed you'd be beating the S&P 500 (and by a decent margin) if you bought every time he predicted a market crash since the recession.

This looks really cool and I will doubtless play, if it ever comes out. Unfortunately I get Hytale vibes from this website. I don't understand the point of announcing such a project, not least with such an elaborate landing page, so early in development.

Yeah, right. And replace them with whom?

How does an operating system even function without localhost? This is absurd I can't believe this was caught by even a single one of the 5,000 devs working on Windows.

> the Boeing 737-800 had just 220kg of fuel left in its tanks... enough for just five or six minutes of flying

Maybe I'm just unaware, but it's crazy to me that these planes burn 40 kilograms of jet fuel per minute.

While this is another ridiculous attempt by the government to exert control, there is something a little odd about an ostensibly neutral encyclopedia being owned by a wealthy, openly politicized nonprofit. When Wikipedia begs for donations, only a fraction of the money goes towards the direct costs of hosting the site (a few million a year), with the rest going to the bloated Wikimedia Foundation, which has 550 employees and a $400 million war chest [1], while administrators and editors get paid nothing. Recall that Wikipedia ran for years with no staff and a shoestring budget.

[1]: https://unherd.com/newsroom/the-next-time-wikipedia-asks-for...

Intuitively it makes a lot of sense. Red is obviously associated with “bad/stop/destructive” and green with “positive/go” while blue is neutral in this respect. The other colors aren’t primary and they have their own problems: orange is too similar to red and yellow is too hard to make out against a white background. Purple is okay which is why it’s used for “visited” links but it’s much less ubiquitous than blue in everyday life which makes it gaudy/distinctive.

I think Topological would more properly be called a "CAD startup" than an "AI startup." I know the distinction is a little nebulous but it's far more descriptive of what the company actually does. The foundation models are merely means to Topological's true end, which is CAD optimization.

It might sound like I'm splitting hairs here but I'm trying to illustrate the problem of defining a startup as an "AI startup" simply due to the presence of AI somewhere in its product. It would be like calling early-days Facebook (the actual website, not the diversified conglomerate that Meta has become) an "internet startup" instead of a "social media startup." Sure, the former is technically true but it's a near-meaningless description, because every company in the 21st century is an "internet company" in some respect.

> Going through YC’s startup directory reveals that of the 170 startups in the most recent summer batch 154 are AI startups.

The author calculates this by searching for the term "AI" in the name and description of each startup's YC page. But presence of the term "AI" doesn't make a startup an "AI startup," so to speak. For example, I picked one startup at random, Topological, which is "developing physics-based foundation models for CAD optimization." Just because the company uses AI doesn't make it an "AI startup." AI is rapidly proving itself to be an extremely useful and workflow-changing tool, and many companies now have adopted it somewhere in their product without suddenly becoming AI companies.

That was my original title but I think mods changed it because it doesn’t reflect the website’s title.

I don’t know how these were advertised when originally sold, but I think products like these ought to clearly display a minimum sever lifetime. Something like “Nest connected thermostat, with 10 years of guaranteed cloud connectivity” displayed relatively prominently on the packaging. Otherwise, if the vendor can arbitrarily remotely yank access to a key feature, then the consumer is being sold a false bill of goods.

This is a shortsighted way of seeing things. The first issue, though surface-level, is using LOC as a measurement. If Company A’s million lines of code are cleaner, clearer, and better-documented than Company B’s 100k lines, then in that case Company A would be better off. What I’m getting at is that the author means to talk about complexity, and is using lines of code as rough measurement for complexity. Code itself is not debt, the complexity engendered by code is.

Code is an asset. It is the product of software companies. Having more assets certainly increases complexity, but this is almost definitionally true. Imagine saying “the US interstate highway system is debt, because it’s complex and difficult to maintain.” The premise is true, but the conclusion is such a one-dimensional way of seeing things.

The AI stuff aside, in light of the above, what is the author’s thesis here? “For the same code, all else being equal, it’s better to have less complexity than more complexity”? Sure, true, but that’s a pretty easy and obvious point.

It seems this entire article could have been profitably boiled down to “make sure your AI coding tools aren’t adding unnecessary complexity to your finished code.”

If Claude usage limits are this important to your life, it seems a little smarter to just bite the Claude Max bullet. Isn't an ordinary sleep schedule easily worth $180/month difference?

When are we getting Mac biometric support for extensions? I want to be able to use my Touch ID with my password manager on my Mac! Do any HNers have solutions for this (Dashlane)?

Does the author find it aesthetic or artistic to write like this? No caps, incomplete sentences, banal and faux-poetic prose? He's writing an opinion piece about software, it doesn't need to read like Tumblr poetry written by teenagers.

Denuvo definitely does attempt to hire from the scene. It's less that they necessarily need those specific peoples' skills, but rather because there are so few people who can crack Denuvo, so hiring a single such person dramatically reduces the number of people out there cracking your games. The most famous example being Empress, who was the only person in the world cracking Denuvo for a few years. I'm sure Denuvo would have loved to hire her, if only to stop her from releasing cracks.

They should TeX these up for some real flair ;)

> The idea of "type safety over the network" is a fiction

> When it comes down to it, what is being sent over the network is 1s and 0s

?

When it comes down to it, all of computing is 1s and 0s. This is not some feature that's particular to the wire.

Why on startup? Windows startup is already so painfully slow, especially compared to Apple silicon machines, and adding Office to it would only compound this problem. I think this problem can be avoided, while also still helping pre-load Office, if Windows just detects when resource utilization is low and loads Office in the background then.

This reminded me of a hilarious passage from Chapter 9 of the novel Ulysses. Stephen Dedalus is pondering a £1 debt he owed from five months prior:

> Wait. Five months now. Molecules all change. I am other I now. Other I got pound.

Hierocles was a Stoic philosopher. I think his "circles of concern" idea is beautiful, and it perfectly expresses how we should treat one another: treat your neighbours and close friends like family members, treat your countrymen and fellow citizens like your neighbours, treat other members of the human race like fellow citizens and countrymen...

I think the Stoic philosophy has been hijacked somewhat by people who think of it as nothing more than "don't show emotion; be impassive and aloof." But I think Stoic philosophy inherently embodies universal love, which seems to be in short supply these days. Not everyone is our enemy.

Fair. I was exaggerating a little.

My point is this: I spend over 100,000 minutes per year listening to music. Any unnecessary friction in the experience is therefore extremely frustrating and has a disproportionate effect on my life. In all my experience self-hosting, it adds a certain level of friction that just isn't worth it considering how much time I spend listening.

I am not hostile to self-hosting in other cases. I am very satisfied with my self-hosted Plex server, for example. But I don't spend 30% of my waking hours watching TV.

> the vast majority of us don't have a requirement to be able to access our media from anywhere in the world

I don't travel internationally often at all. But when I do, access to my music becomes all the more important.

> Spotify can pull the plug on whatever your favorite song is and there is NOTHING you can do about it

There's plenty I can do about it. If Spotify decides to "pull the plug" on a song or artist I like, it's my prerogative to switch providers or decide to self-host. I'm not locked into a lifetime serfdom with them, and it's not like the files will magically vanish from the internet once Spotify decides not to host them.

> who will willfully consume the most crappy audio codec

I listen to 99% of my music through Bluetooth on AirPods, so 320kbps is perfectly fine for my use case. If I was someone who cared deeply about audio quality, I would switch to a competing service with HiFi.

> hundreds of millions of clueless subscribers, such as yourself

At least try to be respectful.

At least in the case of a music player, self-hosting simply isn't good enough for me. I'm not willing to accept a single second of added latency or buffering or downtime because I don't have multimillion dollar server farms. The fact is that the vast majority of us don't have the resources to self-host a Jellyfin instance that can provide near-instantaneous access anywhere in the world to every song ever made at 320kbps. And that's the bar for music. I can deal with a little added latency vs. Netflix on a Plex server or something. But I'm not willing to compromise with music.

This isn't even to mention the numerous features that Spotify has which are difficult or impossible to replicate on self-hosting. The "radio" feature, song recommendations, the DJ, AI playlists, stations, automatic playlist enhancement, social features, Canvas... the list goes on. And of course I never have to worry about managing a library of mp3 files. When an artist I like drops a new album, it'll be on Spotify at 12:00am exactly and work perfectly. This isn't possible with self-hosting.

When you look at it this way, the chance to pay 6 bucks a month to get all these extra features and ignore the headache of self-hosting is a no-brainer.

> One of the hardest things as a DM, in my experience, is creating a balanced encounter without fudging

Just fudge. I know it might seem dishonest, but I think I fudge at least a little bit in probably 80% of my encounters. The most important thing isn't accuracy from the DM's perspective, it's fun and accuracy from the players' perspective. Unless the players catch wind of the fudging, it literally only has upsides:

- You waste less prep time fine-tuning stat blocks and can spend more time on the interesting and material aspects of prep, like designing NPCs, dungeons, etc.

- It makes the combat encounters more interesting, because encounters that would be super one-sided in either direction are instead close and nail-biting

- It allows you to end a boring encounter quicker or prolong unexpectedly interesting encounters

- You can create cool moments where a monster fails an attack at a crucial moment or succeeds when the odds are stacked against them

...etc. I think well-executed fudging is a complete win-win situation, as long as your players don't find out.

> mass appeal might go down

I don't think Fermi questions have mass appeal anyway. I don't think most people know what "order of magnitude" means, let alone understand dimensional analysis (ie., the conceptual idea of what it means for units to cancel).

I don't think making the game a little more complex would really affect its appeal, because it's already targeted at people with a certain level of scientific understanding.

I use em dashes all the time in writing, but unfortunately ChatGPT and co. use the em dash frequently—and most people use the em dash infrequently, not knowing how to type it on a keyboard—so it's starting to make my writing look AI-generated sometimes. I fear it'll have to go the way of words like "tapestry."

FWIW, you can type an em dash on Mac with shift + option + hyphen.

I think:

Call the three numbers a, b, and c. This means c = a + b, but we still don’t know to which person each number belongs.

When person 1 (p1) is asked what his number is, he has no way to know whether he has a, b, or c, so he says he doesn’t know. Same goes for p2 and p3. Clearly p1 somehow gains information by p2 and p3 passing. Either he realizes that he must be either a or b, and such his number is the difference between p2 and p3’s numbers, or he realizes that he must be c and so his number is the sum of p2 and p3’s numbers.

That’s all I have so far. Anyone have other ideas?

I wondered that too.

If we only have two files, A and its duplicate B with some changes as a diff, this works pretty well. Even if the user deletes A, the OS could just apply the diff to the file on disk, unlink A, and assign B to that file.

But if we have A and two different diffs B1 and B2, then try to delete A, it gets a little murkier. Either you do the above process and recalculate the diff for B2 to make it a diff of B1; or you keep the original A floating around on disk, not linked to any file.

Similarly, if you try to modify A, you'd need to recalculate the diffs for all the duplicates. Alternatively, you could do version tracking and have the duplicate's diffs be on a specific version of A. Then every file would have a chain of diffs stretching back to the original content of the file. Complex but could be useful.

It's certainly an interesting concept but might be more trouble than it's worth.