"To be honest, the database nerd in me just wanted to turn all knobs up to 11 and see what breaks."
"And multiplayer 'just worked' because the database system, which handles all the nasty concurrency, is the source of truth," Vogel said."
"I set out to see if I could push Patrick's demo [DuckDB-DOOM] to an extreme: doing the entire rendering pipeline in SQL. And while it works, I have to admit that it is a pretty... bad idea? Fast enough, but horrible to maintain and debug."
DOOMQL implements a Doom-like shooter entirely in SQL, including the rendering pipeline and multiplayer functionality. The project uses CedarDB for database services and relies on a shell script to run the SQL file roughly 30 times per second and Python to handle input. The implementation aims for about 128×64 pixels at 30 frames per second, producing rudimentary visuals. Multiplayer functions by treating the database as the authoritative source, leveraging built-in concurrency handling. The approach proves technically feasible and reasonably performant but suffers from difficult maintenance and debugging, making it an impractical choice for serious development.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]