Some context… a new mutation that accepts declarative configuration is more involved than I thought because of the way we create releases. Right now, changing the image, config, secrets, regions, vm size, etc each individually alter the app spec and deploy. Merging those into one operation technically works, but breaks release commands in the process. Good news is we’re revamping the release code and app config in order to support multiple process types per app, like web and worker. The new mutation I mentioned is how we’ll deploy all apps going forward once launched. We’re currently working on this and should have an update in a few weeks.
I’ve been trying to get this working by hand with the existing graphql API and I’m doing something very wrong. I keep wedging the entire application beyond repair. I probably need to rethink my entire approach. This set of changes isn’t almost ready is it?
It’s not. Don’t tell, because this isn’t public yet, but we’re working to replace Nomad (we’re very far out of Nomad’s sweet spot these days). We opted to wait on some of these things because we don’t want to build them twice.
Please do post your GraphQL adventures in another topic, though, we can absolutely help you get them working.
but I’m dying inside to know what you’re building/using instead 