A community effort to collect a curated set of analysis pipelines built using Nextflow.

nf-core

Hey everyone :wave:

I have a question about Nextflow best practices : when writing process scripts, what's the recommended approach: putting scripts in `bin/` vs `templates/` ?

I understand the basic difference (scripts in `bin/` are directly executable and accessible in `PATH`, while `templates/` are used with the `template` directive and support variable interpolation), but I'm trying to understand:
- In which cases do you prefer one over the other?
- Any major pitfalls to avoid with each approach?
- Is there a community consensus for nf-core pipelines in particular?

Thanks !

I would always prefer `bin` over `templates`, in part because the template functionality is somewhat fragile and recommended against by nextflow and in part because a bin script will necessarily be more portable and force you to write more generalised code 

I think generally within nf-core you would be encouraged to aim for a bin script and there would need to be some justification for why you needed to use interpolation (template) rather than passing values as arguments 