I'm making a tmux session manager which allows you to pre-configure your sessions in a nice language and then switch between them (sort of like tmuxifier), but it will also allow you to create a config for a session based on a repository found on the system (meaning the session root would be the path to the repo).
I also want it to be modular and provide the ability to fuzzy find sessions and repos during creation of the session config.
Due to technicalities I must choose one of these approaches to create a config for a session from a fuzzily picked repo but they both have tradeoffs:
Approach 1:
session-manager new-session "$(session-manager find-repos | fzf --tmux)"
This way the user can easily do what they want with the output, although this could be considered more complex. The original plan was to display shortened names(meaning starting from the top level the shortest "path" that would be unique) but with this I can't do that, so it would look like this:
I think it looks a little bit noisy (but some may consider this more practical), compared to the shortened version (see Approach 2). Alternatively combining these two is possible, meaning: [short name] -- [path]
Approach 2:
session-manager pick-repo
And then it would be configured something like this:
filter_command = { program = "fzf", args = ["--tmux"] }
It's not composable and I think it could be considered less flexible for the user but it's simpler. But with this we can have our shortened paths:
Which of those would offer a better user experience?
