I'm starting to use rash, and having a lot of fun. One question that I have is around best practices when invoking racket code. Specifically, as you might imagine, I have a bunch of shell scripts that invoke racket. When using Rash, I have the ability to make these "internal"; that is, my shell is in fact Racket, so I can just call this code directly. This makes me a bit nervous, in that I somehow feel as though the operating system's process abstraction is going to protect my shell-running racket even when my invoked-process racket goes belly-up... but maybe that's not really a problem in practice.
Does anyone have experience they can share?
Here's a simple blob of rash code formatted as a rashrc.rkt, replacing a ccs shell script that connects these two invocations of racket with &&, to show what I mean (and also to invite criticism; I suspect that I'm making a lot of simple things harder...)