I've run into problems with composer on stage and production.
I've tried clearing the cache multiple times to no avail.
I recently re-organised my package repo and added an ssl certificate. This works fine for dev.
When I try to run a deploy, it seems to be trying to load packages via the old http urls that are no longer in my updated composer.json. I've checked that the correct composer.json is in my remote git repo.
I can't understand where this old composer.json is coming from or how I can get rid of it.
I'm running a version of trellis from late March and ansible 18.104.22.168.
Do I need to upgrade trellis?