qwik

wow that took WAY too long uh, i’m going to tryhard on this devlog to hopefully make up for the 2nd 8 hour devlog

• did a bunch of reformatting pydantic models
• built the pydantic models for track-related endpoints
• found a neat thing to speed up development!
• set up and started working on all the tracks.py endpoints, these also integrate with the metadata extractors!
• aaaand that’s it!
• then a few small commits to fix some bugs :D then some testing and i was NOT locked the fuck in
  yeah uh i keep taking massive periods where i just don’t work and it hurts a lot, but oh well…? hopefully i can finish this before the end of ft…

anyways this marks the first time i was able to upload a track and handle it! it got all the extraction on that one, but if that didn’t work you can always patch details on it later <3 i might refactor that later but right now i just want to get it Working

now i have to work on being able to stream that track (transcoding oh god ffmpreg) and a few other small interactions, maybe content id, then it’s polish stuff! then i can, oh god i have to build a frontend in 2 weeks this is going to be terrible for my votes

ok i spent an hour getting pyright to shut up about the random stuff that was fine but it didn’t like, mostly updating models.py (the database schema) to use a slightly newer class that has type hinting, then went around and did some cleanup work with the remaining type mismatches!!

i know you’re not supposed to do code screenshots but uh ts is just too satisfying + there’s not much else to show

the next devlog I SWEAR FOR REAL THIS TIME will have some audio data being tossed around the api in a few hours but first i need to sleep

I had to spend like an hour trying to figure out how to fix my stupid python install, spent probably a bit too long chasing bugs that didn’t exist… but I also finished the import service, which should be super easy to link into an endpoint and then i can FINALLY start serving things with the api :3

i uhh don’t have much to show for the picture sorry </3 its just like 3 functions

wrote a quick test script for the new metadata extraction, and it Works now™️
this file doesn’t show anything on musicbrainz because it already has all the metadata, i’m too stupid to figure out how to modify files to get a partial match, so you’ll have to trust that part works, there’s also some file validation logic and stuff you can’t see here

now I can finally start implementing actually. importing tracks and files :3
breaking news from the future: broke the db and wasted 2 months of some 50 year old’s time lmao

Didn’t want to use a logging system so I just threw my own thing together lmao, also there are albums endpoints now

This devlog video uh Really sucks because i’m like halfway between features but you can see that it works (kinda)… ignore me not knowing how to call my own api (the docs page doesn’t let me authenticate for some reason, i should really build a devtool or something)

well two months have passed how did that happen
i dunno if i’ll ship this in the next Two Weeks but uh-
auth endpoints are working somehow??
hopefully now i can actually start handling media files

uh. Oh. 8 hours. Uh. Oops.
Yeah uh I wanted to have a frontend utility to have all of my devlogs not just be a bunch of code spewed out and then I decided I needed to do all the config stuff and spend way too long overthinking exactly how it will work… BUT GOOD NEWS next devlog I’ll be able to actually start the app and start moving some data about! First orders of business will probably be to figure out how to transcode files (jobs) then serve them to clients with direct Song ID requests, then I can start organizing them. I will uh. Try to make that Not 8 hours this time.

iiiiiii am in Pain trying to learn databases and SQLAlchemy at the same time, so this has taken far longer than it should have to set up the “model” files (basically a guide for the software to translate python objects later into SQL)

Set up a neat diagram for the tables to hopefully let me understand them better because I may be in a bit over my head, this may or may not change over time, but you can see it at https://dbdiagram.io/d/qwave-db-setup-695a456039fa3db27b0c1f73

(very minor mistake (you can’t change devlog attachments) in playlist_tracks there should not be a connection between playlist_id and track_id)

https://github.com/qwikster/qwave/blob/main/qwave-docs/endpoints.md

I’m really spending a long time setting up everything I need to do before I start writing code. This time around, I got all the endpoints (hopefully) that i’ll need to start writing routing files under FastAPI, and spent some time getting the package set up. if you’re just finding this, maybe join #qwave? :3c

Spent a probably unhealthy amount of time working on setting up files and directories as well as deciding exactly what I want the project to be. I’m taking a bit of a bold step not downloading album covers, but I think it will be a good idea in the end. The project will use FastAPI and SQLAlchemy to use data, and I’d like to make this possible to develop your own frontend for!