Running Gleam Tasks external to a gleam cluster

[dominicfollett]

Hi there, thank you for an awesome project.

I am facing a challenge in running Gleam tasks locally on my MacBook, which has a different architecture compared to the Alpine-based Kubernetes pods running the Gleam master and agent containers.

Also, running the Gleam tasks on my MacBook requires that I do port forwarding for the master and agents pods. I assume that this issue #118 might be related to that?

Currently the only way I can run a task with a K8s deployment is if I run it from a K8s alpine-based pod.

I imagine we'd want to support use cases where a gleam task is run in a serveless context (like AWS Lambda) and accesses a kubernetes cluster to perform the computation. These two issues seem to make that difficult.

However I'd really love to help out. Do you have any suggestions/insights I could explore to help solve these two issues?

[chrislusf]

Current binary sends itself to the remote gleam agent. To fix this, it needs to send a binary compiled for the target OS.

[dominicfollett]

Thanks chrislusf I'll work on preparing a PR for this :)