Not sure if its better, did both and i havent noticed much difference. Forgejo runs codeberg though and is maintained by them. Also has CI with forgejo actions.

My brother in christ, unix was all rights reserved. There was a non-compete agreement prohibiting at&t from selling their OS, hence why it was more or less given to universities. Later, the BSD’s did a theseus ship, and at&t still tried to claim ownership through legal methods. For them, the license symbolizes this independence from at&t, which is why it doesnt lay claims on user protection.

Look, im not even going to respond the first part. I love the bsd’s as well, from a technical standpoint. From a licensing standpoint, not so much (i see the value in a short license, though).

Im not concerned by what these companies use or do not use. Im concerned about protecting my, and other ‘common good’ software with a license that strictly prohibits user exploatation. The GPL does this perfectly.

The isc/bsd2 license does not protect the user

Yes, but this comes with restrictions on distribution of your binary/code/artifacts.

I see the value in these restrictions, but i also see why these libraries are avoided in commercial settings. These terms often come as a suprise from my understanding.

The EUPL solves this by only making claims of the actual modifications to the EUPL licensed components, not any third party user code.

This license (EUPL) was designed with cooperarion as the primary motive, and this is very valuable in my opinion.

I believe the reason we see so much permissive code is because of said suprises with the GPL’s, it defeats the utility of the license itself. I say this as an avid GPL lover, but i have also seen projects like libopencm3, which desperately needs EUPL.

On the other hand we have projects like Linux and VESC, where we absolutely positively need to kill user-exploatation dead in its tracks, mostly since it is an end-user product. The GPL serves its purpose perfectly here.

Also, you might note that the MPL is a valid choice here, but it does not offer the same protection in the case of third party extension of the licensed code, since it is file-based, in essence.

Ive actually spent a good amount of time looking into licenses, would love to hear more of your thoughts.

Here is a discussion and Here is the original author (i think) of the EUPL.

This is somewhat concerning, as im a big fan of working for free, as long as it benefits the users. I have also been looking at the EUPL as a happy middleground (it permits static linking, while any changes to the acual code is copyleft). Copyleft is important, and needs to be talked about.

Absolutely correct, its an insult to everyone involved. Open source would do, or just call it free without the ‘ware’

Its about the type of operations the compiler allow you to do, more or less. Like sharing mutable references, that can be independently changed in a ‘hard to keep track of’- manner. Other factors the compiler tries eliminate include buffer overruns and int overflows e.t.c.

Rust for example sometimes makes trivial things a royal pain, see linked lists for example. It also has a gaping microdependency/supply chain attack prone ecosystem, and the compiler interface is also not stable (afaik, caused some issues in linux). There is also no spec.

I have experience of both, and i love both, but C is my fav. Its often trivial to imagine the codegen with C, and there are no shortage of quality compilers. The language is also small enough that implementing a compiler is actually feasible.

Man kate is such a banger.

Dishonest and misleading. gdb ./main.elf, break 45. Learn your tools. Optimize for learning. Select tools that generalize. Avoid lock-in.

Absolutely this. It almost seems like a controversial opinion sometimes, but microdependencies is a code smell imo. This could largely be improved by providing a more extended standard lib, at the cost of innovation and velocity maybe. I found this interesting: https://blessed.rs/crates

deleted by creator

People like you should be in leadership positions. The landscape rewards quick solutions, and quick solutions are rarely good solutions. “Whatever works” might still be a bad solution, just look at electron and that entire ecosystem.

Interesting. Im curious, what are some key areas of math that you think is the most interesting/useful for software engineering (that you would personally recommend learning)?

I will likely have some spare time in the following months and i currently plan to spend it on deepening my senses related to linear algebra and analysis.

I need that story

For those wondering, this seems to be MIT licensed. I didnt check all components.

I think you are misguided. Given your level of experience, you are not in a position to spew hot takes on software architecture.

There is something else underneath this. Did you try to use linux or something?

General advice when it comes to software is to just start. There are always different paths, depending on what type of programmer you are. My opiniom: choose simple, not easy.

This is terrible advice. Communication is the solution.

Very cool! This is the kind of stuff i want to see. Is AVR still the best choice for this? Any Risc-V ICs that can fill this type of role yet?

Of these, only zig compiles to binary code. The others are usually interpreted.