synthmeat

I'm not sure about the paper and claims on the whole but the Hanoi part has received some shade here https://x.com/scaling01/status/1931783050511126954

Supra-aural headphones with a spherical camera rig would be really powerful ambient for first big AI-first product.

To voice support for current state - do what you're already doing, and I agree on focusing on opening up the format.

Don't hesitate to charge for v2, if improvements pile up and you have affordances to do so. Will gladly pay.

It's pretty great already as-is. Thank you.

You can read more about it here https://www.mongodb.com/docs/manual/changeStreams/

My feeling is that database needs to be built with replication in mind from the get-go to have something like this work well.

MongoDB has had ChangeStreams for a long time now.

Schema-less is a bit wrong term, since you always end up with one but, indeed, schema-on-read vs schema-on-write discussion for individual use cases is far from a settled thing, even though the zeitgeist is that schema-on-write won sometime in the 70s.

You say "prototyping", but I would generalize that to "faster evolving" in the long term. Of course, not without tradeoffs.

I personally have zero issues writing a custom marshaller when needed for any the schema-less document collections I have. Constraints your application has on the data are a superset of db schema anyways.

Just to add to sibling comments, one killer feature for me is ChangeStreams[1]. It's miles ahead of what Postgres[2] offers, and it enables really interesting use cases. Some of my services built around ChangeStreams end up not doing a single query to the DB. Data is right there in the program memory, indexed how I need it to be, and program is immediately reactive to any changes in the DB.

[1] https://www.mongodb.com/docs/manual/changeStreams/

[2] https://blog.sequin.io/all-the-ways-to-capture-changes-in-po...

OpenResty with Redis on the same system - both Lua-powered - is a god-like power.

Ability to do basically control and rewrite all your ingress AND egress and have some state via Redis over UNIX sockets.

What's not to love? After all, it's what powers the bulk of internet.

Granted, it sounds dreamier than it is, since a lot of the contraptions you could come up with might be better placed at application layer. But having operational ability to do these things ad-hoc, security advantages of manipulating at your own infra edge or performance boosts, sure does come in handy!

Also, if they're not keeping old domain (thus being able to create any (new or old) email account), they're doing a really bad job at sysadmining.

Yeah. I'm not 100% even sure what is this "GitHub" part of, but I'd be thankful anyone is helping me in any way to recover my account.

IVMED-80 should come out this year as well.

You can even do live resharding since 5.0.

Safari for general browsing (and default) because it's the most energy efficient one.

Firefox for work. The best implementation of pinned tabs. Shame it's a resource hog.

Chrome for reference and research type stuff. No reason why Chrome, I just want that separate from other contexts.

Chrome Canary for development. Still a Chrome, the best browser for development, but I get to separate all the reference stuff nicely from actual Chrome

Brave for streaming, because it's still a Chrome (so supports all the things) and blocks video ads really well.

Tor for... nice try, FBI. :)

I set up iTerm to open up reference type links I care about in Chrome, and all the other ones in Safari.

Boolean capabilities of meter per every entry? Shouldn't impact size that much, easy to index. As a bonus, you can track failure of a sensor per each meter, and know when it first failed.

It has already been said, but it bears repeating:

"Any sufficiently advanced business model is indistinguishable from a scam."

https://news.ycombinator.com/item?id=8227941

What was it that broke it? Usage of unsafe? Security and bug fixes in the compiler/tooling? Just interested.

Go is yet to prove its stability over timespans similar to those languages you've mentioned, but it doesn't have many of noted (and similar) quirks.

All you've listed is the things you did for yourself. If that's all you did - do for the others. It might be more fulfilling than you think.

> What do you mean by using the key cluster module? I don't have it, but how does that help you when you need to use a laptop?

Well, I put a lot of mappings in the cluster module (tmux prefix, tmux macros, another alt, still playing with what's optimal there), but idea is to focus onto the cluster module what'll not be available on common/laptop keyboard so when you do have to use it, your muscle memory won't press the wrong buttons, since they aren't there. I don't think it completely removes the issue of muscle memory and changing keyboard, but for me it certainly goes a long way.

Similar thing with trackpoint module, when I use trackpad, my hand still starts moving in right direction and prepping a finger to move a cursor.

As a result of those two things, I really rarely experience dissonance using my different keyboards. The only thing I do notice is how it's terrible to use non-split, so I just try not to do it as often as I can.

> How hard is it?

Just keep both network interfaces fired up, and sort them by priority in Network preferences panel (I assume ethernet, then wifi for you use case). Then you can forget about everything and do work, tethered or not.

Turn the bluetooth on as well while you're at it.

This is the way.

My RSI (though there are confounding variables) disappeared after about 4 months using it. Few notes:

- Focusing heavy customization around the key cluster module very much helps with rare occasions I have to use laptop keyboards of mine.

- With trackpoint module, I whip out the mouse so rarely it's in a drawer.

- There are minor mechanical defects that don't affect functionality very much (or at all). Its build quality is oversold, but I don't mind. You will have to buy spare parts some day though.

- split is the best, because your shoulders are the key to posture

From my vantage point, that's perfectly fine. Block ads, change the fonts, color the site orange if you want. As you say, your display, your rules. I communicate to my users that I myself use an ad blocker on my own property. There's no dissonance there.

My mom asked me to disable ad blocking I've set up on her devices, as she finds them useful.

The "evil ads" argument doesn't even look really thought through to me. The way I look at it, there's 3 options:

1. Do everything for free. Rarely tenable.

2. Charge for things. This necessarily puts focus of business on monetization, or at the very least, it frequently pits providing and extracting value at odds.

3. Delegate monetization (a.k.a. ads). Then you can just focus on providing maximum value to your customers and money scales proportionally.

As a builder of things who's not greedy, #3 sounds like positively best option, if it's financially viable and product type makes it possible.

We're querying dozens, if not into hundreds, of GraphQL APIs for a couple of years now. Terrible DX, terrible performance, terrible uptimes. Everyone on the team, with no exception, hates them. Even a lot of those who produce them cobble together a bad REST implementation for parts of their own products.

Agreed even at rate limits comment - frequently, probably in moments of desperation, they rate limit according to Host header to keep their own products up and working.

Would batteries in the bands be A Really Bad Idea?™

> I will be happy to pay $.

Buy rpi400 (or a similar device with dominantly non-consumptive ecosystem around it) to your kid... and all his friends or a class.

I don't see Hetzner dedicated vCPU's on there? I did mention "dcpu" ones. But thank you for your benchmark, it's very useful for shared-cpu comparison.

How did I come to it? Measuring performance of my node.js service I need to run on it, iperf, speedtest.

Yeah, heavy sustained, and DO came up on top (even your link says so). Admittedly, I haven't tested across the pond. I've tested dcpu ones only AND only for my load (node.js with a worker). That's why "benchmarked" is in quotes. :)

Linode/Vultr/DigitalOcean/Hetzner/Terrahost

I can easily imagine cost savings on Hetzner could benefit certain loads, of course. It's always "depends".

For non-dedicated in the US? I can think of at least 2 reasons why not:

- crappy network (advertised 300-500mbps, usually lower)

- crappy cpus ("benchmarked" several providers, hetzner was lower end on cpu-bound loads)

Dedicated ones are baller though, unfortunately no US.

Honestly, as someone who worked on and off with both FB and TW APIs for ages now, FB developer (and business) API experience was actually superb (I know, I'm shocked as well).

TW is the one who's borderline hostile to developers and businesses building on it.