Why is your only solution to literally any problem to rewrite everything? Five months for non trivial auth is hilarious. It’s not like you used WindowsScript, that’s just C#.

lol

You can already do what you want. S3 with HTTP, XML + XSL for responsive / dynamic content.

Lossless Scaling is a very popular application for Windows

This has never been true and writing it in an article as if it were true doesn’t make it more true.

You can also avoid using the internet entirely. I personally reference my 30 volume encyclopedia collection rather than risk seeing an ad or an internet service being anything less than perfect.

God, shut up. There’s the “communist” dumbass instances which is just far right levels of stupid, and then literally the entire rest of this platform is just normal centrist europe views. “Leftist” is a term Americans use to describe giving immigrants healthcare and thinking homeless people should be given homes.

Literally nobody is doing this shit.

Same fucking way americans correct issues with SSA, DHS, and the IRS: paperwork. Come the fuck on with this slippery slope bullshit. Ever had a job? The I-9 employment authorization paperwork… just shut up

Are you stupid?

Literally everyone has been saying this the entire time.

I’ve realized how easy it is to just actually run a network rather than half ass it with tailscale. I recommend this, it’s fun.

That’s not really how this works

For anyone who can’t read that many words at once: URI means identifier and URL means locator.

Matrix does not mix the roles of its URIs beyond the ability to attempt to reach that homeserver through NOT ONLY DNS but also through routing between homeservers. Matrix, unlike this garbage, actually wants to support ephemeral homeservers and such as well.

Sadly these web technology fucks cannot get their heads even a centimeter further towards the exit of their anus to even risk beginning to care about how garbage their architectures are.

Matrix was influenced by the traditional URI schema approach however they also used an inverted URI to have the most significant segment be the first segmont. This is why it has a prepended segment followed by a : and then a URL.

The relationship the URI is describing is homeserver owns user and traditionally we might go homeserver:user I’d argue it is obvious to just invert that into user:homeserver. See Java, dotnet, etc reverse dns naming conventions.

A matrix room is not at a URI, it is on every homeserver that participates in the room. I am not talking about Lemmy and its garbage entirely incorrectly semantic URI scheme. Matrix rooms are globally uniquely identified, and so the room URI only describes the idea of the resource of the room. room:homeserver.

The prefixes are an obvious and neccesary evil for parsing them out of unstructured text. A requirement for most users.

ActivityPub and related went “fuck everything, fuck reason, the web is fucking amazing” and came up with their own flavor of stupidity. Emails use an ancient first attempt at a URL. A URL. What does the URL do? It is explicitly intended to tell you which server to contact. People are going “yeah but email!1!!” entirely moronically ignoring historical context.

They aren’t URLs they are URI schemas. A URL is a segment of the URI schema.

Mozilla doesn’t deserve a single bit of trust or goodwill from users any longer. We’re in a dark world with no good shepard of a useful browser engine. No I don’t give a shit about that dumb C++ one from the OS project.

There is no need, and you’re defeating the point of using tailscale. Use headscale if you cannot summit your anxiety around trusting tailscale.

Private internet access is nice

In principle it’s just “slimmer ARM”. RISC-V is also extremely dedicated to using memory mapped IO rather than older style IO x86_64 supports.

Think lots of registers, a fun zero register that is always zero, and memory mapped IO.

Signal? Just use matrix already!