This is an idle thought, but for once I guess I’ll write it down.
I’m working (slowly, it’s a low-ish priority free time project) on a backend for luvsound which I’d like to eventually open source for other indies to use… the project so far is all go + sqlite.
This has been fine. I like go. It’s portable, relatively simple and easy to deploy.
SQLite3 is great. It’s extremely portable and given its age and stability I feel like the time I have invested in learning about it is well spent.
When I transitioned pippi from python 2 to python 3 I decided to move all of the C backend into cython, and then regretted that decision since it makes it basically impossible to embed anywhere outside of python. Since then I’ve mostly ported pippi back into a C library, which has already been useful in microcontrollers and other environments.
I’d like that kind of portability for the work I’m putting into the luvsound backend, too. I realized tonight that it would be worth exploring developing each major feature as its own module (download codes, large file uploads, etc) which could persist its state in sqlite and express its behavior in C.
There would be some additional work in developing interfaces for each environment it could be supported in, but I love the idea of maintaining the core C libraries and some example interfaces in python and other environments I end up using while keeping it all portable enough to allow potential contributions of other target environments.
Log June 2026
Log May 2026
Log April 2026
Log March 2026
Log February 2026
Log January 2026
Log December 2025
Log November 2025
Log October 2025
Log September 2025 and earlier