1. Record one album a month. An album must consist of no less than four songs. At least one song must be an original. I can only record on a 4-track cassette recorder.
Problem: I'm getting older and find myself nostalgic for the days when I was running my record label and playing in bands. A small, manageable project with no expectations or demands to scratch that itch and get me away from computers for a few hours a week is nice.
2. http://postgra.ph -- Almost ready... just a landing page to test an idea: A GraphQL Backend as a Service powered by PostgreSQL. I just have to add the SSL certs tonight when I get home.
The MVP is based on ideas from PostgREST/PostGraphQL -- generating the API from the public schema of the database. It'd be the bare-bones service that I could throw together in a couple of weeks.
If it takes off then I'd look into integrations, adding PipelineDB support, auto-scaling, etc.
Problem: I just wanted an API-as-a-Service that would give me full control of the data schema but didn't require me to write yet-another-web-service-in-dynamic-language-framework-foo. There are nice solutions out there for different folks but I'm a big postgres fan and wanted something that didn't require me to learn a new framework, interface, etc.
Authentication and especially authorization can be completely handled by PostgreSQL.
In front of it all sits OpenResty (nginx) so that is where you would add whatever headers you would need
jwt's are a touchy subject but was the well-trodden route I was planning to follow for authentication.
Integration with auth0 and other third-party services would be a roadmap thing for me.
Authorization can be handled by PostgreSQL: it has built-in facilities for role-based access control and row-level security. You can develop the authorization scheme that fits your application.
For a service like this to work, one thing needs to be solved, automating the code deployment (i am talking views/functions/roles/grants/RLS). As far as i know (and i've asked other people) this is not a solved problem. This is what i am working on now. The rest is done
I was with you on your #1 until I saw 4 tracks- I'm not sure I could even get myself down to that few. I know the Beatles did it, but they were far greater musicians than I. I haven't listened to a cassette tape since 1999- how's the quality?
I think I could maybe do 4 tracks if it recorded to an SD card. Do you find the length of a tape adds to limitations in a way that keeps you brief?
Good. I record on high-bias Type-A tapes. You still get some hiss and tracks bleed the tiniest bit. But I think it sounds pretty good for the kind of music I'm making.
I've been using the procedural drummer in Garageband as the drummer for some of my songs and with a decent amount of swing it sounds "authentic" on tape.
> Do you find the length of a tape adds to limitations in a way that keeps you brief?
You get about 60 mins of record time per tape so... not really.
I find I like recording this way for many of the same reasons I like writing on my typewriter: zero distractions, low friction between thought and recording.
The restriction of 4 tracks and that I only have a month to record 4 songs keeps me from "fidgeting" with a song. I'm not able to aim for "perfect." It's more ritualistic. I show up on the same evenings in the same space. I begin the ritual by opening my journal and listening to last weeks tracks. I record some ideas, experiment. I end the ritual with a glass of bourbon. I close the book. I've written one song I think is pretty decent so far. Try as I might though they each have little tiny, beautiful flaws. Only so much you can do.
1. Record one album a month. An album must consist of no less than four songs. At least one song must be an original. I can only record on a 4-track cassette recorder.
Problem: I'm getting older and find myself nostalgic for the days when I was running my record label and playing in bands. A small, manageable project with no expectations or demands to scratch that itch and get me away from computers for a few hours a week is nice.
2. http://postgra.ph -- Almost ready... just a landing page to test an idea: A GraphQL Backend as a Service powered by PostgreSQL. I just have to add the SSL certs tonight when I get home.
The MVP is based on ideas from PostgREST/PostGraphQL -- generating the API from the public schema of the database. It'd be the bare-bones service that I could throw together in a couple of weeks.
If it takes off then I'd look into integrations, adding PipelineDB support, auto-scaling, etc.
Problem: I just wanted an API-as-a-Service that would give me full control of the data schema but didn't require me to write yet-another-web-service-in-dynamic-language-framework-foo. There are nice solutions out there for different folks but I'm a big postgres fan and wanted something that didn't require me to learn a new framework, interface, etc.