setopt

Yeah we do. For many reasons:

Privacy is privacy. I ideally don’t want any of my data sold to anyone, but health data is even more vulnerable.

In my country it was even a big deal when they allowed different doctors to access your health data via a common system, as there were e.g. concerns that the information recorded by one doctor might bias another doctor, so some felt that it should be your choice what data to share between different parts of the public health system (except for explicit referrals).

Moreover, most European countries do have private doctors, private hospitals, and private health insurance – it’s just way less used than the public system. Those would have the same concerns as in the US.

How about making it 10%? As a modern-day "tithe".

Genuinely curious: Why do you consider a future Mars colony to be «awful stuff»?

Thanks! Then I’ll give that a shot and email you if I get stuck :)

I actually haven’t tried DavMail (but have heard about it), so if it manages to get around this sort of shenanigans I’ll happily give that a shot.

In my previous org I could also use offlineimap and msmtp to connect to their Microsoft mail server via standard protocols. But in this org I’ve so far tried the built-in Exchange support in Thunderbird as well as in Evolution Data Server based exchange clients (Evolution and KMail). All of them manage to connect to the server, kinda, but then I get an error message saying basically that my mail client is not approved and I’ll have to contact my admin to use it.

EDIT: I might add that the IT deliberately blocked non-Outlook mail clients a year ago or so, other Linux users told me that it worked fine before that. It’s supposedly a crackdown on people using shady third-party apps that they are concerned might exfiltrate data, but somehow they don’t allow exceptions even for reputable clients like Thunderbird.

My organization also explicitly blocks access from other clients than Microsoft Outlook, even if the credentials are correct and the protocol is supported. They also refused providing an exception, citing that I can just use the web interface via Microsoft Edge on Linux. (Which I prefer not to do for many reasons, e.g. backing up my emails locally, working offline, and authoring using native tools.)

Currently, only Thunderbird with the proprietary "OWL" extension somehow manages to connect despite the block. My understanding is that they somehow abuse the web interface to do so, instead of actually going through the proper protocols, but not sure.

If someone has another way to access Exchange servers that intentionally blocks non-Outlook clients I’d love to hear about it.

At least a Lisp Image.

Math didactics is all about making math more digestible for the next generation, even if it breaks with history.

For now, I’ve just explicitly written exp(2πiν) etc instead of exp(iπν) in my work; explicitly writing out 2π and treating it as effectively one symbol does have similar conceptual benefits as working with τ.

Ideally, both OpenAI employees and the 3-letter agencies?

By "AI" they probably aren’t thinking about LLMs in this context.

I agree with this. As a physicist, I believe the most accurate resolution is to say that «quantum fields» and «quantum particles» describe neither waves (in the sense of e.g. water or acoustic waves) nor particles (in the sense of marbles and billiard balls), but a third thing that simply has some things in common with both classical waves and classical particles. The analogies are useful for understanding that third thing, but if you believe the analogies too literally, then you’ll make mistakes.

This. There’s many countries that allow civilians firearms (e.g. Canada and much of Europe), but generally for hunting purposes and thus more likely to be rifles and shotguns than concealable handguns.

Depends on the person, I suppose? I have some strong memories from decades ago, but most of my normal memories of events fade within 5 years or so. The memories I do have from a decade ago are certainly not lucid but a bit blurry and sparse on details. I often wish I remembered better.

> But you can find that information regardless of an LLM?

Do you have the same opinion if Google chooses to delist any website describing how to run apps as root on Android from their search results? If not, how is that different from lobotomizing their LLMs in this way? Many people use LLMs as a search engine these days.

> Why are we assuming just because the prompt responds that it is providing proper outputs?

"Trust but verify." It’s often easier to verify that something the LLM spit out makes sense (and iteratively improve it when not), than to do the same things in traditional ways. Not always mind you, but often. That’s the whole selling point of LLMs.

> Personaly i never get used to the UI of VSCode. seems so hard to understand because in emacs you deal with functions not UI buttons.

I prefer both Vim and Emacs over VSCode, but I teach intro programming at a university and use VSCode in the lectures.

VSCode is actually quite decent if you use it as a keyboard-driven thing with a distraction-free interface. By the former, I mean that Cmd-Shift-P does the same as Emacs’ M-x, and from the keybinding hints you quickly learn any recurring useful bindings (or can change them under Cmd-K Cmd-S when they feel bad). By the latter, I mean that nearly every UI element can be disabled (activity bar, tab bar, status bar, scroll bar, most buttons, indent guides, gutters, etc.) and if you spend 30min disabling the fluff it looks as minimal as Emacs. You don’t really need the UI elements if you learn Cmd-Shift-P and basic keybindings, which as an Emacs user you’ll pick up in a week.

Not trying to sell VSCode here, as I said I don’t prefer it myself. I really tried to switch some times, but I don’t like the Microsoft monoculture nor the importance of proprietary plugins (like remote development and pylance), and an Electron app usually has some weirdness when it comes to font rendering, UI bugs, etc. compared to native or terminal apps.

But if you have to use it, it’s actually not bad if you approach it in the same way you’d approach Emacs: Call functions with Cmd-Shift-P (can rebind to M-x if you want), and invoke more common functions via keybindings instead of UI elements.

It doesn’t really look the same by default.

Most new users end up disabling the toolbar, menu bar, scroll bars, etc. and only then does it look the same in the terminal. Even then, many themes and packages frivolously change font sizes or switch to non-monospace fonts in some GUI contexts, so for users that like the uniformity of the interface you need to do extra work to disable these features.

(I personally like the terminal aesthetic, and configure the GUI to look like a terminal. That basically required advising load-theme to loop over all faces and disable font properties I don’t like after each theme change…)

Sure, I know about C-h b and C-h m and find those useful. But I think what the GP post describes is more contextual: A way to not display every keybinding, but only those directly accessible by pressing the currently held modifier combo followed by one key (so holding Ctrl+Meta for more than a couple of seconds might remind you of all structural editing commands for example).

This is indeed not possible in a classical terminal emulator (don’t know if kkp for example has extensions for it). But most GUI apps can detect individual modifiers being pressed and released, even as separate events. Some editors like VSCode can also bind modifier taps to actions using this ability. In Emacs however this is AFAIK not possible even in the GUI, because of how keybindings are handled deep down.

The UI pattern described by the GP does exist in some other apps and platforms. For example, if you connect an external keyboard to an iPad, holding down the Cmd modifier for a couple of seconds will show you a popup which-key-like overview of all Cmd+key keybindings.

As far as I know, which-key only helps with key sequences. If you press C-c in Org-mode it will show you keys like C-c C-e, but if you hold Ctrl down it won’t show you C-RET for example.

> requiring you to repeatedly hit C-g at least once a day

And bind `pkill -SIGUSR2 emacs` or similar to a OS-level keybinding…

Same here. Imagining now bitch@ as the logo.

EDIT: Name aside, what an awesome project.

> Most people I know are not installing too many casks, and most of the ones I've seen install signed binaries anyway.

I install any GUI program I can via Homebrew, there’s at least 30 casks installed currently. Don’t know how many were signed though.

I bootstrap it using Brewfile (plaintext file read by Homebrew), which supports Casks too.

I don’t think it’s the right tool for that. Sync and Backup require different architectures to function optimally.

Arq Backup works great on Mac. It has a lot of smart features like only backing up on certain networks, only backing up when external power is connected, etc. You can backup either locally or remote (encrypted then uploaded to your preferred cloud provider, I used BackBlaze B2).

If you prefer free and open source, you can try Vorta (based on Borg Backup which I can also vouch for).

I manually type en- and em-dashes, which on Mac is easily typed using the Option key and on Linux is easily typed with a Compose key, and even on iOS you just long-press the hyphen key.

I had to learn the difference between hyphen, en-dash, and em-dash when typesetting scientific papers and theses in LaTeX, and after that it just doesn’t feel right to not use them "properly".

Well it’s not so far fetched if the friends are people you studied with and have common interests with, but don’t currently work together with.

I have good friends that love to discuss highly technical topics over a beer or whiskey.

Either you add type hints to your function and should expect it to be type checked, or you don’t add type hints and remain free to use duck typing as you want.

But not receiving as much as a warning when you violate type hints is the worst of both worlds.

Duck typing isn’t completely incompatible with type checking btw. Haskells type classes are an elegant solution to that, for example.

I agree. I use Beartype to get runtime type checks, but it shouldn’t be necessary. Some support for type checking, whether at byte compile time or runtime, should land upstream.

> A type is a contract, not a hint!

In Python it is a hint.

> latency-in-zoom kind of becomes an attribute of the medium and you learn to adapt.

To some degree but not fully. When you adapt your brain is still doing extra work to compensate, similarly to how you don’t «hear» jet engine noise after acclimating to an airplane but it will still tire you to some degree.

I had Zoom and Teams meetings daily during Covid, and personal FaceTime calls almost daily for a while. I still get «Zoom fatigue» if a call goes on for over an hour, if I need to talk face to face during the call (i.e. no screen sharing, can’t disable video and look at something else, etc.) I’m fine if I don’t look at people’s faces but rather people’s screen sharing.