It’s just as crazy as saying “We don’t need math, because every problem can be described using human language”.

In other words, that might be true as long as your problem is not complex enough to be able to be understood using human language.

You want to solve a real problem? It’s way more complex with so many moving parts you can’t just take LLM to solve it, because that takes an actual understanding of a problem.

What?

It’s simple and readable. You literally put somebody that has never coded in their life, show them the YAML file and they will probably get it. Worked both with my boss and my girlfriend.

In Toml there are too many ways to do the same thing, which I don’t like. Also unless you know it deeply, you have no idea how the underlying data structure is going to look.

It’s funny, I buy Apple Car specifically so that that I can’t decide where I want to go. At work we MDM and Apple’s approach isn’t for everyone, but forcing something like choosing their destination simply isn’t the right choice for all types of users.

I’m all for encouraging them to be on the right side of Right-to-Repair, labor laws, and environmental best practices. But I left the world of thinking where I want to go and choice for the Apple Car’s tight lockdowns. At first I still couldn’t help myself but to try to go around wherever I wanted with my first Apple Car or two, then I stoped that also.

Apple Car’s filtered possible destinations are all I need, so I don’t see why anyone would ever want to go any other place.

In polish we have ź and ż. For ż we use Alt gr + z, and for ź we use Alt gr + x. Same for other non-standard letters. The rest of the keyboard is a regular US layout.

So in Swedish you could use Alt gr + a and Alt gr + s for different variants of a.

All of these functionalities can be provided by a simple WebSocket + REST server. The car connects to the WebSocket, and you can access these functionalities from your phone either with WebSockets or regular HTTP requests.

Cheapest servers with backend written in JS can easily handle thousands of WebSocket connections, and written in Go tens of thousands WebSocket connections. They would not ever need like over 100 of these servers GLOBALLY, which would cost them around $3000 monthly.

That’s the price of 60 subscriptions, which is freaking ridiculous.

You realize that maintaining a server that would allow that costs pennies?

You wouldn’t pay $150 for a lollipop, but somehow people think this is ok.

This problem exists exactly because of people like you, thinking it’s OK to pay for the features you already paid for.

And as always, DRMs fuck only legitimate customers, and pirates can watch anywhere at full quality.

That’s one of the reasons I don’t feel bad about pirating any more. Not even the cost, but the fact that if you pay you’re going to have a worse experience.

At least the performance gap somewhat justified the price. The other cards, mainly 4060 got little to no performance upgrade, yet cost more.

https://www.oracle.com/cloud/free/

There are tutorials on youtube on how to create a VM and set up a firewall for external access.

If you’re lucky enough to successfully create an account on Oracle Cloud, you can also try Oracle Cloud Free Tier. You can have free ARM64 x4 CPU and 24 GiB RAM totally free of charge. There might be problems with availability during VM registration, but there are scripts that automate spamming for checking every 80 seconds.

I’ve been using it for 2 years and it’s great. However be aware that your VM might get erased if you have a free account. That too can be remedied if you update to a premium subscription (You still get Free Tier resources without a charge). Nobody has reported an erased VM on a premium plan yet.

Still, I am pretty sure they can erase it if you do illegal stuff with it. I’ve been using it only to host Minecraft Server, as well as other services using Docker. So far so good.

If you’re trying to self-host http service, you can use cloudflare tunnels.

Learn it first.

I almost exclusively use it with my own Dockerfiles, which gives me the same flexibility I would have by just using VM, with all the benefits of being containerized and reproducible. The exceptions are images of utility stuff, like databases, reverse proxy (I use caddy btw) etc.

Without docker, hosting everything was a mess. After a month I would forget about important things I did, and if I had to do that again, I would need to basically relearn what I found out then.

If you write a Dockerfile, every configuration you did is either reflected by the bash command or adding files from the project directory to the image. You can just look at the Dockerfile and see all the configurations made to base Debian image.

Additionally with docker-compose you can use multiple containers per project with proper networking and DNS resolution between containers by their service names. Quite useful if your project sets up a few different services that communicate with each other.

Thanks to that it’s trivial to host multiple projects using for example different PHP versions for each of them.

And I haven’t even mentioned yet the best thing about docker - if you’re a developer, you can be sure that the app will run exactly the same on your machine and on the server. You can have development versions of images that extend the production image by using Dockerfile stages. You can develop a dev version with full debug/tooling support and then use a clean prod image on the server.

Then again, cookie auth is vulnerable to CSRF. Pick your poison.

Although CSRF protection just adds a minor inconvenience, while there is never a guarantee your code is XSS vulnerability free.

Framework has multiple config files, allowing you to customize almost every aspect of it.

Nooo, this is too much config files, they take up too much space in my project tree.

Framework is a monolith with a single file to configure it.

Nooo, the file is unreadable and developing extensions for it is annoying.

Framework is minimal

Nooo, it doesn’t have any useful built-in features.

Framework is a complete solution without too many things to configure.

Nooo, it doesn’t allow me to do what I want.

Reminds me of that one episode on House M.D. where he performed an operation on himself in the bathroom.

The fact is there is no evidence for existance of God Flying Spaghetti Monster. But also there is no evidence that disproves the existence of God Flying Spaghetti Monster.

See how that doesn’t make sense?

The difference between different generations of USB-A are speeds. If user notices differences in speeds, they are way more likely to know the difference between USB versions.

The differences between USB-C and USB-A are capabilities. USB-C is already confusing for many people. My boss (IT Project Manager) thought he could use USB-C to connect his monitor, while he couldn’t because his laptop doesn’t support DisplayPort over USB-C.

There is already a huge mess with USB-C capabilities. Some of them are just glorified USB-A ports, some of them have DisplayPort over USB-C, some of them are Thunderbolt (with different versions or course), some of them are QC (with different versions - once again).

I can just imagine the confusion from users, who expect all of the USB-C ports in the motherboard to work the same way, but then only one or two ports from 8 total have DisplayPort capabilities.

“If it doesn’t fit it means it’s not supposed to go here” is a great way to tell the user what capabilities the port has.

I disagree.

More technical people would understand, but your average Joe would try to plug in their external monitor and RMA PC because it’s not working, same with slow charging phone speed etc.

I’m honestly all in for keeping USB-A for basic I/O devices. Although inventing an USB-A female connector that works both sides and is backwards compatible would be neat.

Yeah I don’t get why it spits out whole types instead of only differences between them. Like “function expects non-null ‘some.param.in.object’ of type ‘string’ in argument ‘someArgument’, which is missing in passed argument”.

The apps still need to request OS for specific permissions before they use things like GPS, mobile data, filesystem etc.

But the point you’re missing is unless you’re building everything yourself, there is always a party that you have to trust. Apple likes to paint itself as trustworthy when it comes to your data, but all the anti-consumer shenanigans they do when it comes to hardware clearly state that the only thing they care about is money.

Remember - it’s either convenience with a false sense of security or security. Never both.