It’s the way to go, but too difficult for most users in my experience. They rather just install Docker Desktop and use git bash. Sad reality

Their entire offering is such a joke. I’m forced to use Docker Desktop for work, as we’re on Windows. Every time that piece of shit gets updated, it’s more useless garbage. Endless security snake oil features. Their installer even messes with your WSL home directory. They literally fuck with your AWS and Azure credentials to make it more “convenient” for you to use their cloud integrations. When they implemented that, they just deleted my AWS profile from my home directory, because they felt it should instead be a symlink to my Windows home directory. These people are not to be trusted with elevated privileges on your system. They actively abuse the privilege.

The only reason they exist is that they are holding the majority of images hostage on their registry. Their customers are similarly being held hostage, because they started to use Docker on Windows desktops and are now locked in. Nobody gives a shit about any of their benefits. Free technology and hosting was their setup, now they let everyone bleed who got caught. Prices will rise until they find their sweet spot. Thanks for the tech. Now die already.

They use Windows

Not having to install dependencies is a benefit of containers and their images. That’s a pretty big thing to miss. Maybe give it a closer look.

Your choice of container runtime has zero impact on the rate-limits of Docker Hub. They probably had a container image proxy already and just switched because Docker is a security nightmare and needlessly heavy.

I gave podman compose a fresh try just the other day and was happy to see that it “just worked”.

I’m personally pissed about aardvark-dns, which provides DNS for podman. The version that is still in Debian Stable sets a TTL of 24h on A record responses. This caused my entire service network to be disrupted whenever a pod restarted. The default behavior for similar resolvers is to set a TTL of 0. It’s like people who maintain it take it as an opportunity to rewrite existing solutions in Rust and implement all the bugs they can. Sometimes feels like someone just thought it would be a fun summer break project to implement DNS or network security.

A single malfunctioning service that restarts in a loop can exhaust the limit near instantly. And now you can’t bring up any of your services, because you’re blocked.

I’ve been there plenty of times. If you have to rely on docker.io, you better pay up. Running your own NexusRM or Harbor to proxy it can drastically improve your situation though.

Docker is a pile of shit. Steer clear entirely of any of their offerings if possible.

Just ask yourself, who is still posting on Facebook? Your friends? I hope not. The last time I hung out on that site, the groups seemed to be the only valuable section to participate in. But it’s ultimately just a circlejerk and you’re feeding content into a garbage platform stuffed with ads. Not a great way to spend time.

Marketing play to grab the money off of rich parents. There are still teachers, they are just proxied by “AI”. And there will also still be teachers monitoring. And there will still be teachers for certain topics.

So it’s teacherless, but with plenty of teachers.

Should I really give up my empty metric of 70K followers and move my communication and journalistic research to another echo chamber and advertising platform run by another billionaire?

It really is a tough one.

Oh boy, what a marvelous idea. This could save the tanking DJT stock and allow them to prolong the scam. It would allow Trump to close the Truth Social scam with a seemingly sensible move. Elon is supposed to be in his cabinet anyway. It’s perfect.

Following along with the style of your own post: YAML doesn’t suck, because I feel so.

Thanks for asking.

Right. Who needs expert opinions if an LLM can produce similarly convincing garbage?

I feel like you’re catching on.

So if I watch all Star Wars movies, and then get a crew together to make a couple of identical movies that were inspired by my earlier watching, and then sell the movies, then this is actually completely legal.

It doesn’t matter if they stole the source material. They are selling a machine that can create copyright infringements at a click of a button, and that’s a problem.

This is not the same as an artist looking at every single piece of art in the world and being able to replicate it to hang it in the living room. This is an army of artists that are enslaved by a single company to sell any copy of any artwork they want. That army works as long as you feed it electricity and free labor of actual artists.

Theft actually seems like a great word for what these scammers are doing.

If you run some open source model on your own machine, that’s a different story.

AI is great, what OpenAI does is blockchain-level idiocy.

Depends on the product. It’s just something to think about when signaling errors. There is information for the API client developer, there is information for the client code, and there’s information for the user of the client. Remembering these distinct concerns, and providing distinct solutions, helps. I don’t think there is a single approach that is always correct.

I don’t necessarily disagree, but I have spent considerable time on this subject and can see merit in decoupling your own error signaling from the HTTP layer.

No matter how you design your API, if you’re passing through additional layers, like load balancers and CDNs, you no longer have full control over all responses your clients receive. At this point it may be viable to always signal a successful backend connection with a 200, even if the process resulted in a failure.

Going further, your API may include partial success scenarios, think batch processing, then the result could be a mix of success and failure that doesn’t translate to HTTP status.

You could even argue that there is really no reason to couple your API so tightly with a concept of the transport layer it uses.

I really hate it when people call for impromptu meetings and are completely oblivious to what you mention. People are absolutely incapable of bridging mental gaps. Nobody explains common vocabulary. Nobody explains the expected goal of conversation. Nobody evens the playing field. Instead, you watch people confused and asking stupid questions, before they arrive at a constructive mental place, right before the meeting is over.

Communication is art and a skill. Just because someone is talking a lot, doesn’t mean they communicate well.

If you can efficiently enable a group of people to arrive in a mental context where they can contribute value to a decision or process, you are a valuable team member.

IMHO this always requires preparation. You can’t expect to have a valuable exchange if you yourself can’t fully imagine the mental context the other people are in. At every moment you have to understand what might be keeping them from understanding you, and then approaching the specific conflict. “Why don’t you understand me?” is something you should never have to ask yourself.

Also, yes, build more prototypes and actually watch some shit go instead of talking so fucking much. Pictures are a thousand words and a real thing is like thousands of pictures. Stop talking already!

Respect the Accept header from the client. If they need JSON, send JSON, otherwise don’t.

Repeating an HTTP status code in the body is redundant and error prone. Never do it.

Error codes are great. Ensure to prefix yours and keep them unique.

Error messages can be helpful, but often lead developers to just display them in the frontend, breaking i18n. Some people supply error messages in multiple languages, depending on the Accept-Language header.