I actually prefer building and synchronizing database from a Git repository rather than a directory. This allows more feature such as proper handling of removed resources, automatic parsing of creation/modification times, etc.
It also allows synchronizing databases by only performing changes, rather than rebuilding the entire website from scratch.
Flow:
- Go commit by commit since the last one (stored in vars.latest_one) and only perform changes sequentially.
- Properly handle files that were moved or removed, but do not leave them with a 30x or 40x status code, only the config file is in charge of that for now.
I actually prefer building and synchronizing database from a Git repository rather than a directory. This allows more feature such as proper handling of removed resources, automatic parsing of creation/modification times, etc.
It also allows synchronizing databases by only performing changes, rather than rebuilding the entire website from scratch.
Flow: