Currently, pipelight loads the first file it recursively finds.
I would like pipelight to load every file in can recursively find. So that a user can define pipelines in toml, yaml... side by side in the same directory.
Refactor configuration loader. Nothing too difficult
A bit more difficult.
- Need to refresh the screen every time a process output is pushed to kernel queue.
We need to use the kernel here because implementing our own queue could lead to read/write concurrency on process input/output.
Add a no-log cli parameter to run a pipeline silently without producing any logs on filesystem.