Some suggestions for the deploy command

[nckrtl]

• Read the .env file if there is one. This would make it easier to set certain variables on a per project basis.
• Make reading the configuration files dynamic. This way you don't need a .infrastructure folder if you haven't defined any config key. Only caveat is that it will use the file name to generate the environment variable. So

  configs:
  traefik:
  name: 'traefik-${SPIN_TRAEFIK_CONFIG_MD5_HASH}.yml'
  file: ./.infrastructure/conf/traefik/prod/traefik.yml

  Would become:

  configs:
  traefik:
  name: 'traefik-${SPIN_TRAEFIK_YML_CONFIG_MD5_HASH}.yml'
  file: ./.infrastructure/conf/traefik/prod/traefik.yml

This would also work:

configs:
  traefik:
    name: 'traefik-${SPIN_TRAEFIK_YML_CONFIG_MD5_HASH}.yml'
    file: ./.infrastructure/conf/traefik/prod/traefik.yml
  redis:
    name: 'redis-${SPIN_REDIS_CONF_CONFIG_MD5_HASH}.yml'
    file: ./.infrastructure/conf/redis/prod/redis.conf

Maybe we could drop the CONFIG part, so it becomes {SPIN_TRAEFIK_YML_MD5_HASH}

• Remove the requirement for a Docker image. This would make it easier to run specific docker stacks with only redis or traefik
• Added some additional echo statement before and after setting up the tunnel.

[nckrtl]

An few examples of what the stack templates could look like with these changes: redis: https://github.com/nckrtl/spin-redis traefik: https://github.com/nckrtl/spin-traefik

[jaydrogers]

Wow, this is AWESOME!! Thanks @nckrtl!

I give you a lot of credit for going through this, because even I had a hard time reading my code 😆

Thanks for the clean up efforts on this. I really like where this is going. I appreciate your help!

[jaydrogers]

I was so excited to thank you, I forgot to mention...

Changes I made

• Variables follow SPIN_MD5_HASH_{{filename}}

Thoughts?

[nckrtl]

Yeah I like that. Glad you like the changes :)