It’s also a lot harder to make money with open source. You would need to find something to offer on top for money. That can be hosting (nextcloud), support (redhat), consultation (hyper) or additional features(zeds AI offering).

In a side node: when building for geentoo it should be possible to build all rust libs as dylib. The compiler only offers a stable abi for the same version and all flags being equal. When building everything yourself that should work at the coat of updating the compiler requiring a complete rebuild of all libs.

Sourcehut duports git-email.

Step one: Download a C or CPP repository.

Step two: Replace all semicolons with a greek comma.

Step three: ??

Step four: Poison Copilot, so that it randomly insert greek comas that the compilers totally choke on.

I should have been more clear.

I meant for self hosting.

Though realistically, even if the service is provided for the public, you could just use an instance of keycloak or something similar with open registration. That’s what an association I’m close to is doing already.

By default, the Credentials provider does not persist data in the database. However, you can still create and save any data in your database, you just have to provide the necessary logic, eg. to encrypt passwords, add rate-limiting, add password reset functionality, etc.

That is exactly the complexity I wouldn’t want. With just SSO it is enough to send a redirect URL to the browser and on the callback set a cookie. No js needed. If your service gets compromised and someone leeks the credentials, just log everyone out.

If i created a service I would go in the opposit direction. Only offer SSO and no other option.

You loose quite a bit of complexity that way.

I mean, if you really want to use a different language, there are macros that enable that for you. Here is the example for Rost. Just Rust, but in German.

You say you are on a budget. Yet you talk about 128 Gigs of ram.

Maybe you should clarify what your budget is.

Oh, yes it can. The license only changes what other people than the owner may do. It’s the rights and conditions they give you.

For most projects that doesn’t matter because there are several owners of the code base. Every single person who contributed can enforce these rights on their part. However, to contribute to Zed you have to sign a cla. Signing away all rights and ownership of your contribution. So they have all the rights and can do whatever they want.

They could close source everything tomorrow without any consequence and sell you a feature you made yourself.

Its timing based. When piped a script, bash executes each line completly before taking the next line from the input. Curl has a limited output buffer.

1. Operation that takes a long time. Like a sleep, or if you want it less obvious. A download, an unzip operation, apt update, etc.
2. Fill the buffer with more bash commands.
3. Measure on the server if at some point curl stops downloading the script.
4. Serve a malicious payload.

They can even serve a different file for curl vs curl|bash

Has the same vibes as anthropic creating a C compiler which passes 99% of compiler tests.

That last percent is really important. At least that last percent are some really specific edge cases right?

Description:
When compiling the following code with CCC using -std=c23:

bool is_even(int number) {
   return number % 2 == 0;
}

the compiler fails to compile due to bool, true, and false being unrecognized. The same code compiles correctly with GCC and Clang in C23 mode.

Source

Well fuck.

Deref is for smart pointers and not for inheritance.

I think the Problem is the kernel dumping RAM to disk in crash. Potentially containing secrets.

I manage all my services with systems. Simple services like kanidm, that are just a single native executable run baremetal with a different user. More complex Setups like immich or anything that requires a pzthon venv runs from a docker compose file that gets managed by systemd. Each service has its own user and it’s own directory.

https://github.com/rust-lang/rfcs/pull/3243

Only the rfc is merged. The others need work including implementations.

Difftastic looks really cool.

First Problem is the Server nearly everyone uses is dritten in Python. There is no real way to migrate between implementations. That at least could be fixed.

Second Problem is the massive storage space usage. That one is by design and can’t be fixed.