sprinkly-dust

Every house there (that can afford the essential $200 or so) has a Reverse Osmosis water filter because the municipal water supply (if any) is not considered potable.

One paper that really hammers this point: Inherent Trade-Offs in Algorithmic Fairness [1]

The example they focus on is different but the general principles and takeaways are very powerful and applicable to all classification problems including content flagging.

[1]: https://dl.acm.org/doi/10.1145/3219617.3219634

F-Star and Dijkstra Monads (which are apparently not really monads?) might get you to what you're looking for

The OS-wide Emacs bindings are one of the main things still keeping me on macOS; in editors like Helix (or Zed with Helix mode) I invoke these really cursed combinations of the Helix(vim-like) shortcuts in the editor with the Emacs ones from the OS and it feels really natural as a means of moving around text.

I haven't put in enough effort to replicate the experience on GNOME but once I jump ship from Apple I'll have to.

The way macOS treats shortcuts thanks to the separation of Command and Control is great. It even makes Google Docs pretty painless to use because it's mostly like Emacs. Word still tries to hijack the shortcuts so that's suboptimal.

I have fond memories of fiddling with the wave simulations on this website during high school!

A couple hundred lines of regex for a full email validation might disagree with you there.

I hope to see these seemingly mundane unknown unknowns raised in space travel centered hard science fiction. I think The Martian and Seveneves almost captured these but not quite.

> the guy who makes one watch a month

That's the thing, though. Most people alive today will never be able to possess such an object, no matter how prized and beloved it is. Still, if people want to be able to tell the time from their wrist in a reliable fashion, there are _plenty_ of far cheaper options available to them. The craftsmanship does have inherent value, yes. That does not mean the practical solution is worthless.

There can be practices incorporated in the production of software, involving AI use in a responsible fashion (difficult, of course), that produces practical solutions to real world problems far faster than a group of industry-hardened veterans painstakingly polishing their codebase in pursuit of craftsmanship. Those who appreciate how it is made will pay for the crafstmanship. Those who cannot afford to do so, and only care about a solution working well enough for the tasks they want to accomplish, the production line is good enough.

There are still free as in freedom software hardliner folks out there. The idea that every piece of revoked source code is an affront to computing rights might be less applicable in Obsidian's case since the files are still portable, and the system may be sufficiently extensible through custom plugins (you can load anything you want through the developer plugins option) that source code itself is not necessary. Though perhaps one might want to re-assure themselves that there is nothing 'malicious' happening in the software, that's only achievable with auditing it oneself and using reproducible builds. Perhaps the freedom to fork is also not as thoroughly infringed since the files are portable and reverse engineering is not impeded.

Club Penguin always held a special place in my heart but I never reflected on why. I'm sure part of it is a yearning for simpler times and a childhood unawareness of the nature of the world.

However, playing this reminded me of exploring hidden parts of a map imbued with love, the joys of finding other souls just lounging around or moving excitedly, and the cute little games and things strewn about every corner.

Agreed, I really prefer using my reMarkable 2. I see it less like an iPad and more like a bottomless pit of scratch paper and printouts that I can carry more conveniently than individual dead tree products. This is probably furthered by my not using their cloud subscription and using a USB cable or SSH to transfer files instead.

A lot of lower division math and computer science courses now presuppose iPads or other digital pen devices for working through handouts during lecture. Printed handouts are often available at request, but not the expectation / default.

On the other hand, I've seen more professors — especially in the humanities, but also upper div CS — start banning devices in lecture partially or altogether. Complete distraction (scrolling Instagram, etc.) during lecture is extremely prevalent, and they keep citing noticeable improvements in engagement after banning devices. This also coincides with a shift back to less take-home assignments and more exam-style assessment since they want greater assurance people aren't completely offloading their cognition to LLMs.

David Huffman too, made some very complex paper folding models from relatively abstract mathematics, although apparently did not like them being likened to origami.

I would disagree that a smartphone without a working camera is perfectly usable. A lot of the world — especially in developing countries — runs on QR Codes for everything from restaurant menus to electronic payments. Without a camera, other stuff too, like KYC, just doesn't work. These are the sorts of changes that, as you mentioned, are forcing people to use smartphones. And they rely on the camera.

comma.ai is an autonomous driving add-on (fully open source software and hardware IIRC) for cars that don't have (full) autonomous driving but do have hardware like radar cruise control, etc. that can let software control the accelerator, brakes, steering, etc. so that the add-on can take control of the car.

The OP brings up testimony of someone other than himself who prefers when software drives their car rather than him.

Thank you, this was a fun rabbit hole to dive down. That blog also has a well-argued article about Zero Interest Rate Policy which relates to the doordash story: https://www.readmargins.com/p/zirp-explains-the-world

The remarkable 2, especially if you downgrade the OS to the older 2.x versions, is very hackable. It runs full Linux, I followed a blog post about setting up FDE on it using go and cryptfs [1]. You can make GUI applications, people have even run XFCE, and people have fed handwriting input to another API so routing it internally shouldn't be much harder.

So yes, theoretically you can turn the reMarkable 2 into a emacs lisp machine, although it would take a considerable amount of time.

I haven't read up on the reMarkable Paper Pro, but I think the dev ecosystem for that is very much alive too.

[1] https://blog.redteam-pentesting.de/2021/remarkable-encryptio...

Currently that feature is unsupported or I just can't figure out how to do it. With the latest compiler version 0.14 any .typ file I try to compile will incur warnings about skipping the equations (skipping the main reason I'd want to compile a Typst file to HTML...).

As per their GitHub they haven't included MathJax or KaTeX support yet as they were more focused on semantic and structural accuracy of HTML output with this release.

Seconding this, I've used alternatives like Amethyst, and I can't disable S.I.P for yabai, but Aerospace fills in that missing aspect from Linux when I am in want of it.

This might be (relatively) easily achieved by networking a couple of Raspberry Pi Zero 2's together under the Pi 500+ keyboard shell.

It might not be Ubuntu but Asahi Linux runs Fedora pretty well on M2 Pro and older Apple Silicon Mac Minis: https://asahilinux.org/fedora/#device-support

That is one of the possible outcomes right? Producers have an incentive to collude and not compete with one another. They could create a consortium to fix prices, and use tactics such as acquisitions or _dissuasion_ to prevent new, more efficient competitors from undercutting their prices, thus distorting a free market equilibrium.

The consortium creates an oligopoly which prevents mutually beneficial deals that would have otherwise taken place in a regulated competitive free market between consumers and producers.

'rsync' used via a terminal emulator like Termux or Android's own Linux VM is also good for resuming partial file transfers.

My biggest peeve with the forced push towards apps is that you're often forced to fork over permissions to access valuable data that they couldn't otherwise gleam glean from a browser. Chief amongst this is the Contacts permission.

If you are one to carefully curate your contacts book to contain addresses, emails, birthdays for your own convenience and productivity, you have now provided a veritable goldmine of information for these companies to plunder, and betray the confidence of your acquaintances. I really despise this, and I've been looking for a solution but none thus far have seemed satisfactory.

I know of apps (at least on Android) like Bouncer and alternatives from F-Droid that can temporarily grant access to certain permissions like Location for a few minutes at a time, while giving the apps the illusion of full access.

However, save for using a Private Space or different user profile (both similarly require provisioning basically an extra instance of Google Play and everything) I haven't yet found a way to feed some sort of dummy contacts book to these greedy apps. If anyone knows of such a solution that is more convenient than setting up a new profile then please enlighten us.

It is my experience that this is what Google does with their Pixel phones. It is really quite simple to unlock the bootloader and do whatever you want on a Google Pixel you own (i.e unlocked, no carrier). They even give you this really handy Android flash tool which uses WebUSB to fully restore your device when you mess up. Heck, custom ROMs like GrapheneOS and CalyxOS are even able to sign their own images and allow you to lock the bootloader with a non Google OS.

However, all this comes with the caveat that SafetyNet will flay you alive. The cat and mouse game with Magisk and other methods to maintain root undetected is moot when I've used apps these days that make a fuss when you have developer settings enabled. To be honest, that seems acceptable to me, I can do what I want with my device, software vendors like banks and the like have a say in how I choose to access their more convenient services. I can play nice with them if I want, even using a second phone perhaps, but I have a choice.

Isn't flutter (especially including the flutter rust bridge), achieving that? You do need Android Studio and XCode installed on your machine but you don't have to interact with them directly.

That feature is available, albeit clunky. The latest Pixel phones have a developer option "Force Desktop Mode" which gives you Desktop-esque Windowing on an external monitor.

From my understanding, they are somehow able to intercept DoH and block queries to restricted domains. Perhaps it is through unencrypted Server Name Indication, hard IP blocking, or otherwise.

If you have access to a cloud VPS server, I would suggest you try running an Outline VPN Server instance. https://getoutline.org. By far, it is the most resilient to interception tactics. It was designed for easy use by journalists to bypass censors. Shadowshocks is harder to detect them other VPN types like Wireguard because those don't try to hide themselves at all. They (the IT department) would need some serious kit to detect that kind of traffic and differentiate it from https.

Although, beware that using it 24/7 will still look suspicious to any IT Admin because then all your traffic is going to a single weird server.

If you feel this way about the works of Neal Stephenson, might I suggest Cixin Liu?

Though well-done in it's own right, the 3-Body Problem Netflix series does not quite do justice to the intricacy of Liu's writing in the Remembrance of Earth's Past trilogy. You might find the change of pace to be somewhat jarring in comparison to Stephenson but it fulfills a similar enjoyment of Science Fiction.

Having composed many pieces of coursework using Typst, I must say that it certainly makes academic writing more streamlined, engaging, and dare I say fun — though that might just be me. The functional nature of Typst's syntax ensures I don't have the erratic behaviour emblematic of modern day Microsoft Word and sometimes even Google Docs. Using a local IDE such as VSCode brings all the features one could like.

In comparison to LaTeX, overall document typesetting is far more straightforward. However, for long multi-page stretches of equations solving, I feel that LaTeX is easier to type than Typst because its syntax is not that of a functional programming language but more akin to markdown. Thus, one does not need to think as far in advance when typesetting equations with lots of functions, superscript, and subscript.