My son, which is 13, started programming on his own. I have to admit that I am blown away by his capacity to discover and tinker.
He is now writing C and asking me the usual questions: "how should I name this type", "how should I manage memory", "how can I create a linked list implementation".
But while answering, I found myself thinking "I would love a language that does X or Y".
I currently use Zig, Elixir and Rust, but I have been doing C programming for nearly three decades and there are still some things I like about C. Those things are more related to the ecosystem of C than the language itself. Like portability, library availability, "hackability"...
I am closely watching Zig progressing, and with the new Io interface I was like "yes, that's what I would have done". I have a very high respect for the people working on the Zig language, a lot of decisions feel right. And realizing that in the end, they followed what I had in mind for years gave me confidence. Maybe my ideas are not so bad after all.
Watching my son have a fresh new perspective on things I know revived something in me. The will to tinker and do silly things.
For those reasons, I have decided to try my hand at creating a programming language.
It should be as fast as C and as easy to write as Lua, have good memory safety, support embedded targets (no allocation) and be portable to the usual targets and WASM.
I have been thinking intensely about it and have decided to name it "ten", as 天 in Japanese which means heaven. Yeah, I am THAT presumptuous. But I am not mistaken, I know it has a 99% chance of failure, whether it's never used, or I cannot produce anything good, many reasons can make it not work. But as I said, something woke up in me, and I think it is time I do this stupid thing.