A music player, but also a music platform...? I don't know, still figuring out what this is. And what the final name will be.
- Create the backend resources: terraform/README.md
- Index the music: shiva/README.md
- Run the client: client/README.md
This system is optimized for a personal music collection:
- A reasonable amount of artists/albums
- Low cost (and dependent on usage)
- Simplicity
What is "a reasonable amount of artists/albums"? Any amount that a person could hold in their music collection. Now if you plan on using this software to compete with a popular music platform, it won't scale. Once the number of artists goes from the hundreds (or even a few thousands) into the millions, the system will become unusable slow due to certain database operations.
The cost will depend on the usage of the system. And the usage is the sum of:
- Storage
- Requests
- Transfer
From those variables the only fixed cost is storage, even if you don't use the system, you will have a fixed cost from AWS just for holding your data. For the size of a normal personal music collection requests and transfer will most likely not transgress the limits of the free tier offered by Amazon. However, check what those limits are in your region and remember that they can change at any time.
Gawshi aims at being very simple to use, inspired by classic music players such as Amarok. Its focus is mostly on artists and albums, rather than tracks, and it won't provide many (if any) "social" capabilities (e.g. facebook or twitter integration). It's for personal use.
Installation was made as simple as deploying a serverless app in the cloud can be made. It's not a trivial task but it's as automated as possible. It boils down to:
- Manually create a user that will be used to create another user with very limited permissions (this is the only "manual" step)
- Creates the new, restricted user (from now on is all done from the cli)
- Create all gawshi resources and deploy the system (with the restricted user)
- Index your music
- Publish it
Indexing and publishing was intentionally split into 2 steps so you can mess around with the data before publishing. This could mean fixing artist names, removing some tracks (or even full albums) that you don't want uploaded, or just checking out what will be published. This requires some SQL knowledge though, because you will have to mess around with the sqlite database where we store the data for publishing.
You can use DB Browser to inspect and fix the data before publishing it.
- Simplicity: Work with very few concepts.
- Humbleness: Recognize your limitations.
- Excellence: Do few things but excel at them.
- Embrace the medium: Don't try to mimic a native experience on the web, make the most out of what you are given.