evilchili / grooveondemand Goto Github PK

readme needs a section to shout out the tech this thing is built on -- bottle, typer, rich, prompt_toolkit, etc.

several of the older modules are still very sketching-by-class and need a pass for docs, type hinting, and general orderliness.

Need to add this to the template so sharing links generates nice previews.

Maybe?

Perf seems not great, but all we're doing is calling asyncio.create_task() to scan every audio source file. Can we make it faster with queues and a pool of workers or something?

Reference task: ~30k sources, via NFS from Synology DSJ411-slim (RAID5):

• Laptop (12th Gen Intel® Core™ i5-1235U × 12), wifi, m.2 SSD (PCIe gen4): ~44min
• NUC (Intel(R) Celeron(R) J4105 CPU @ 1.50GHz), 10Gb ethernet, some 7200rpm spinning rust: ~90min

Suspect it is largely I/O bound on database writes.

Currently just default log handlers. Need to move these to streams, enable/disable, wire logs (bytes moved, etc).

Since it doesn't have a subcommand associated with it I forgot to mention it >_<

Add the ability to start the server from the shell, so you can click on links when you're working on a playlist without having to tab over to another session.

Maybe? would allow for measuring total playlist length etc.

Guide the user through each configuration option and save the results to the env file. Prompt with defaults. Offer to scan when done.

Tracks should have links to bandcamp / qobuz / whatever searches.

Main README needs usage examples. a term capture would be helpful. talk about autocompletes etc.

Would be nice to transcode media to webm or something for maximal portability and maybe not serving lossless audio to randos on the internet.