I totally agree with you. This article was really a response to a lot of hype around async web servers in Python.

I kind of knew what to expect, but wanted to throw real numbers against it. I was surprised to see a 10x slowdown with the async switch in Django.

Only reliable web server is an Erlang web server.

No connection proxy in this case. The pooled sync test uses client side pooling which shows better performance. Using a proxy would have the same effect, just moves the pooling to server side.

It is using the async driver. I am using FastAPI’s thin wrapper around SQLAlchemy which also does some slight tuning for it to work better with FastAPI in an async mode.

This is running with concurrent requests. 64 workers firing request to be exact.

I will take a “No AI” label so I can filter down to the games I want to buy.

Do ya’ll just not use linters?

If Microsoft knows how to do one thing well, it’s killing a successful product.

A lot of companies make their most senior devs engineering managers, and expect them to stay technical. I assume this was the case here.

That’s the project’s goal and they have 100% comparability across quite a few of the tools. Definitely still a ways to go before they can fully replace all of coreutils, but Ubuntu’s goal is to replace the tools peace meal with the once that are ready.

If you go with NixOS or guix, both of which are declaratively configured, you can swap 99.9% seamlessly. I run the same NixOS config on 2 machine and a VM. Only thing I had to modify between them are some driver specific settings, but those are easy to isolate.

That looks really well done. And a lot of stuff would be condensed by having viduals.

Doesn’t look like my preferred style… Maybe that’s why I couldn’t get into the book either 😅

War and Peace. Heard so many good things about it. Despite everything, went in not having super high expectations.

The whole book turned out like a reality tv show. All the characters had some petty drama that they blew out of proportion. Hundreds of pages where nothing really happens, people just complain or bad mouth other characters.

I had to stop half way through.

Totally get that. Just saying that different people want different things out of their jobs, and it’s a good thing that there are places where all of them can fit.

Isn’t that the whole point of hiring people that fit the company culture? I’ve worked at both types of places in different stages of my life. Both can feel good or bad depending on where you’re at. Don’t try to change the job to fit your needs. Find a different one.

Your summary of the language is spot on. I still hope that more distros take inspiration from the declarative config and try to move in the direction, or nix supports a better language in the future. I think that ultimately that’s what the average linux user would want. The ability to still customize in a safe manner. Silverblue, and others, are and will remain a great option for the new or indifferent user.

On your point about the transient phase, nix actually does that by default already. It installs everything at a separate path and then flips over in one go. You can even pick the mode, either try to do a live switch as you describe, or on boot. I don’t know if I see many benefits to images there.

I am at a second place now that uses NixOS in a corporate setting, and it is much easier than maintaining the CoreOS images, or similar. I’ve had some many broken builds of CoreOS images because something goes wrong between the custom packages and the base CoreOS images, I would rather just run an Ansible script at this point. Also, you end up using the exact same test suite for NixOS images as for your other images, so the same guarantees end up being met.

Definitely. That’s a great way to run different option together.

I was just using the DE as an example to demonstrate how cleanly NixOS can add and remove packages. The clean removal of packages with lots of configs is something that most distros struggle with.

No, I fully understand it. But if you build the whole system where every package is isolated, none of the packages interfere with each other, and every package is tested across a wide array of architectures, you can just as safely put together your ideal OS setup and don’t have to deal with being locked into very simple and bare system.

The right place for immutable OSes is if you’re using it as a server for container workloads, where you will never customize the base system. Or if you never want to customize your system. Yes, you can customize the system image, but it breaks all the guarantees that the images gives you because the packages themselves are not isolated and by bumping a wrong dependency for a custom packages you can still break the whole system.

Already has that. And if you use flakes, you can fully lock down your package versions that way the install is 100% identical on every machine no matter when you run it.

My favorite example of how idempotent NixOS is has to do with the DE. If you’ve ever looked at switching from gnome to KDE, or the other way around, most distros suggest to just re-install because each DE leaves so much cruft around and it’s so hard to remove everything in a safe manner.

With NixOS, you just change one line in your config, and the DE is cleanly swapped.