Not defending the default behavior of zed, but it is possible to disable this. Setting `"lsp.<server>.binary.path": <some-path>` will stop zed from trying to install that server.
On my machines, the "languages"/"node" directories for zed are empty and owned by root and the lsp servers are provided by nix. But you could also pin known good versions with npm.
As far as I know Vscode has no equivalent way to do this.
It costs several times what it would cost a small team of engineers, even assuming you gave the engineers more time to do it. I'm guessing (wildly) this was around 0.5M USD in compute time. You do get the result quicker, though.
> I'm guessing (wildly) this was around 0.5M USD in compute time.
That seems like an especially wild guess. If you take e.g. Opus 4.7 prices, and make the assumption that you are consuming roughly $30 for every million tokens of output (this comes from just summing the $25 per million tokens of output and $5 per million tokens of input and assuming that caching basically makes all that work out), and assume an output rate of 80 tokens per second (which seems like a high estimate based on online searching), it would take you about 2411 days of non-stop Opus 4.7 usage to hit 500k in API spend.
The only way you could possibly run that amount of usage in 6 days is if you were running ~400 instances in parallel. From personal experience, that seems crazy high for this project.
I think you are off by at least an order of magnitude (potentially even 2 depending on how the person is managing agents, but I could see something like dozens of agents 24/7, so I'm way less confident in 2, but I think it's still more likely to be closer to 10-20k in API spend).
From the leaked internal prompts, Opus 4.7 vs 4.6 recomputes several times over before returning the result. For heavy use like this, it costs Anthropic far more than you're paying as a consumer. They rely on the light users to offset the whales, and they're still at a significant net loss. If you tried this as an end-user, they might cut you off (though I understand their data centers are underutilized, so that wouldn't be for logistic reasons). Being part of the company and directly sanctioned, the author has unlimited access.
~7x overcompute * ~7x real cost to Anthropic * your 10-20k estimate for consumer use is my thought for actual total cost. If the honeymoon period runs out and they're still in business, this is what everyone will pay.
But usually companies are much more careful before even spending that half a million. (And most companies don't have that money sitting around.) They would do small PoCs, do comprehensive benchmarks and evaluations of those PoCs, and decide whether to actually go ahead, and, more importantly, stick to it.
Being able to afford half a million doesn't mean you do it on a whim, or just throw all of that away if things don't go well.
But what do I know. I am nothing compared to our AI overlords like Anthropic.
45 million lines would get to ~$1.125 mil for the linux kernel.
950k lines for Bun would get to $23,750
use whatever math you like ofc.
Does an Anthropic/employee pay that, no. Even if it's at a loss in terms of company revenue, it's worth burning the private capital for all kinds of other reasons.
This unhides the menu instead of hiding everything behind a burger. This should be the default. The defaults are awful in many ways and they've only gotten worse with the recent panel rearrangement.
I think this is a mostly fair criticism of nixos. Nixos has a lot of powerful tools, but if you don't need them, they can get in the way. Some assorted notes:
> the constant cycle of rebuild → fix → rebuild → fix → rebuild
I've found this useful to eliminate the rebuild loop:
https://kokada.dev/blog/quick-bits-realise-nix-symlinks/
It lets you make the config of the program you choose a regular mutable file instead of a symlink so you can quickly iterate and test changes.
> In contrast, Arch Linux simply downloads prebuilt binaries via pacman or an AUR helper
If a binary exists. A lot of AUR packages I used to rely on didn't have a binary package (or the binary package was out of date) and would have to build from source. On nixos my machines are set up to use distributed builds (https://wiki.nixos.org/wiki/Distributed_build). Packages that do need built from source get built on my server downstairs. The server also runs a reverse proxy cache so I only need to download packages once per version.
Distributed AUR builds are possible on arch, but they require a lot of setup and are still fragile like regular AUR builds, your only choice of dependencies are what's currently available in the repos.
> On my machine, regular maintenance updates without proper caching easily take 4–5+ hours
It sounds like the author may be running the unstable release channel and/or using some heavy unstable packages. Which might explain a lot of other problems the author is having too.
Back when I used arch, I found that as time went on, my system would sort of accumulate packages. I would install $package, then in the next version of $package a dependency would be added on $dep. When I updated, $dep would be installed, then eventually $package would drop the dependency on $dep, but $dep would remain installed. I would periodically have to run pacman -R $(pacman -Qtqd | tr '\n' ' ') to clear out packages that were no longer required.
In using your server as a proxy cache, do you just include the server as a nix cache substitutor, or simply use a MITM approach using something like squid proxy?
If the former, via substitutor (or if also using a remote builder), how do you manage when moving portable clients outside your LAN? E.g. traveling with your laptop? Do you tunnel back home, have a toggle to change substitutor priorities?
I find it the default timeout for unresponsive substituters excessively long, as well as the repeated retries for each requested derivation annoying, rather than it recalling and skipping unresponsive substituters for subsequent derivations in the same nix switch/build invocation.
The server is configured as a caching reverse proxy (just nginx with cache.nixos.org as upstream) which I think is similar to squid proxy.
Outside my LAN I do have the ability to tunnel home, but depending on the connection and the updates I just deal with the timeouts, or just wait until I get home.
I think technically you can override substituters using cli options but its not ideal. There are several proposals for configuring timeouts per substituter but none of them are merged yet.
Can you expand on how you use Zellij? I tried it and I understand you can use it for splits, and tabs similar to tmux. But I might revisit it if it allows an IDE like workflow with Helix.
Asahi only supports M1, M2, and alpha support for M3.
Not that I blame them for the lag, there's a lot of reversing work because apple doesn't document this stuff.