Fork me on GitHub
#fulcro
<
2023-04-16
>
Brian Marco20:04:38

I feel like I am missing something fundamental when it comes to getting a live coding environment in place. Each piece of the environment takes a step or two and as you add more features the number of scripts that need to be running increases. If I have a shadow-cljs server, and datomic, and a framework like tailwindcss, etc there are a bunch of processes that need to be spun up and down when switching between projects. Should I be doing this manually and leaving it running for long periods of time, writing shell scripts to do these tasks, integrating with my editor/IDE in some way, make a bunch of daemons, get CI setup and use that instead of a live coding environment, something else?

tony.kay20:04:56

Why run tailwind? Just use the full one while developing. I use docker for containers for services (e.g. redis, etc.). On Datomic, if you’re using on-prem them you could try to dockerize it as well. I typically have docker, shadow-cljs, IntelliJ and a REPL or two. But I just start them at the beginning of my work day. Not a big deal.

Brian Marco21:04:42

Tailwind is new to me, I’m still figuring out how it works (the quick start instructions have it running). Getting yet another tool set up prompted my question. I skipped over Docker to learn Guix/Nix instead, but Docker has a lot more support so I think I’ll add it to my workflow rather than write a bunch of Shepherd services from scratch. Thanks for the advice

thenonameguy06:04:39

I recommend using https://github.com/F1bonacc1/process-compose to solve the process orchestration/log tailing problem

Brian Marco14:04:30

Thanks @URZNNB6KX this will give me a smoother transition into containerization if I go that route and allow for more flexibility if I chose to pursue functional package management.

🚀 2