This started off as a single file in my private nix config, to see if I could get it working at all. In that initial part, some parts were indeed LLM generated (esp. testcases based on my existing intents and failures).

When I noticed that this might actually work and be useful not just for myself though, I moved everything out manually, refactored and cleaned it up, and everything since has just been myself. I guess you’re still right though. I’ll see about adding a disclaimer to the README until I’ve gotten the chance to properly rewrite everything.

Huh, interesting. Does that work with wildcards like “put x on my shopping list”? Also, what are you using for that, if I may ask?

Sorry, I don’t quite follow 😅

What’s the problematic response?

Yay, that’s fantastic to hear!

Also, how’s your experience been with the PE? Getting a readymade device in a nice shell is appealing for sure 😅

Oh, in the demo gif, that’s via a shortcut (holding power for half a second). Sorry, can’t help with wakeword there 😅

Very cool. I’ll definitely look into that, and let you know back here :D

Glad to be of service… 😄

did you consider metaphone matching?

I did not even know about this. Sounds super interesting. Though it seems to be very language specific?

My original intent was to not rely on language specifics. But maybe we could just define additional steps in the pipeline for specific languages. Hm. I’ll have to think about this some more, but it might definitely be a great idea for a future version, so thanks for telling me about it!!

Have fun, hope this works out for you! FYI: you can also use an LLM as an additional fallback (first closest-intent, then on failure, LLM). README mentions it further down on Github.

Neovim, configured entirely through nixvim. I always liked neovim, but it’s never been as incredibly stable as now with nixvim.

Main/only IDE both in private and at work. Can’t ever go back, muscle memory has ensured that.

That ks for sharing this, this is fascinating.

Maybe the underlying rule is: the more you know about something, the more you are aware of its flaws, making the alternatives you know less about more attractive?

I don’t really know, sorry :(

If you want to migrate, is going conduit - conduwuit - continuwuity (first version) - continuwuity (current version) maybe an option?

I went with continuwuity and am happy with it. Development happens at a steady pace, with sane priorities. The server is stable and I haven’t had any issues to speak of, despite one minor bug that got resolved very quickly after creating an issue.

If you use nixos, you basically have to know/learn/use day-to-day the nix language.

nixpkgs are written using nix the language, using concepts mostly familiar from just using nixos.

Basically everyone using nixos is capable of contributing packages.

Ah, sorry. https://repology.org/repositories/graphs

Just gonna leave this here

Yes

And even if you do live there.

You mean, spend 4-6 days tearing your hair out, before landing on a solution which evaluates to literally the same output as your current version, but is 10% cleaner and more elegant?

Of course you do, after all, that satisfies that itch. Well. For a while, anyways…

Oh please. Be real. Are you sure there’s nothing in your flake to refactor or modularize? :)

Huh - you’re right. I went back to Signal’s X3DH spec because I was sure I was right, but it seems I misremembered how the “prekey bundles” work: Users publish these to the server, allowing (in my original assumption) for the server to just swap them out for a server/attacker-controlled key bundle for each Alice and Bob.

However, when Alice wants to send Bob an initial message and she gets a forged prekey bundle, Bob will simply not be able to derive the same key and communication will fail, because Bob knows what his SPK private key is, while the server only knows the public key.

A compromised server would allow the server to man-in-the-middle all new connections (as in, if Alice and Bob have never talked to each other before, the Server/Eva can MITM the x3dh key exchange and all subsequent communication). That’s why verifying your contact’s signatures out-of-band is so important.

(And if you did verify signatures in this case, then the issue would immediately be apparent, yes.)

Edit: I was wrong. See below.