Jinx VS libriscv

Check out jinx https://jamesboer.github.io/Jinx/

Don't get me wrong, I'm a big fan of automated tests where it makes sense. I wrote a scripting language that I use for my personal game projects, and I never would have been able to do it if it weren't for the battery of tests for every feature, error, and corner case I could think of. But games are rarely like other software, with hard rules about what is "correct" or "incorrect". And it would be a nightmare to try to keep up with designers, constantly tweaking and tuning, so what's "correct" is literally a day to day, constantly moving target.

A second trial for you might be Jinx. Depending on your definition of procedural, Jinx is 100% only procedural. https://jamesboer.github.io/Jinx/

Jinx (embeddable scripting language) works as following:

Feel free to look at my own interpreter, written in modern C++. You're welcome to ask me if you have any questions.

I use my own scripting language more like content, especially for things like one-off events and behaviors. Example: scripting special behaviors for a boss fight, or a room with a unique trap in it, or any other sort of one-off behavior that would be overkill for C++, but too complex for most other types of data-driven content. These days, visual scripting also helps to fill in these gaps between content and procedure.

You're welcome to look at my scripting language Jinx, written in modern C++. Just let me know if you have specific questions. Data flow is JxLexer.cpp -> JxParser.cpp -> JxScript.cpp. Most everything else is implementation details. Also, note the parer is pretty complex, mostly because Jinx has a crazily flexible syntax for functions.

Feel free to look at Jinx if you want an example of what I consider a fairly easy-to-use and integrate scripting system. Obviously, I'm a bit biased since I wrote it.

If you're working in UTF-8 internally, you could just write your own UTF-16 to UTF-8 conversion functions to convert strings at API boundaries. I did this in my scripting language because I didn't want to bring in dependencies.

Seeking: https://github.com/fwsGonzo/libriscv

This is a C++ RISC-V emulator that focuses on isolating a single process, aka userspace emulation. I am currently working mostly on binary translation, and recently I have made a push to move it from experimental state to fully supported. Another experimental feature is embedding libtcc and using that for binary translation. It is fairly fast to compile, and gives decent speedups. The challenge is what to do now that (perhaps) some low hanging fruits have been picked.

I definitely recommend people to consider the base ISA of RISC-V if they want to try to implement a CPU or even full-system emulation. I understand that implementing a GameBoy emulator might be more attractive because you are working towards something graphical, but you can definitely get something similar with RISC-V, eg. Doom (SDL example: https://github.com/fwsGonzo/libriscv/tree/master/examples/do...)

This is the example project: https://github.com/fwsGonzo/libriscv/tree/master/examples/msvc

Pretty much two days of work: https://github.com/fwsGonzo/libriscv/commits/master

Show HN: Libriscv – RISC-V userspace emulator library\ (7 comments)

There is a doom emulation demo here now: https://github.com/fwsGonzo/libriscv/tree/master/emulator/do...

You will need to add the shareware doom1.wad yourself. :)