I have been testing different environments for Elixir development & noticed something odd. When I run the Elixir formatter locally in my usual setup, it gives consistent results. But when I try formatting the same code inside Cursor AI; the output sometimes looks slightly different, especially with long pipeline chains and spacing around function arguments.
I’m wondering if anyone else has noticed this? Could it be related to Cursor AI applying its own formatting rules on top of Elixir’s defaults, or maybe some hidden config I’m not aware of? This makes it tricky when working in teams, since code styles should ideally stay the same across all environments.
I use Cursor and have it set up to format on save (via the formatter) and I haven’t noticed anything odd.
The behaviors closest to what you describe tend to be related to Cursor Tab. Cursor Tab autocomplete will sometimes sneak in when I’m not expecting it and make changes; this could go unnoticed if I was keyboarding around for other reasons, and this would also including it’s idea of formatting since it tries to add not just lines of code, but code in what it sees as a consistent format (consistent with what may vary).
I’ve found Cursor Tab to be kinda annoying and have it turned of and since then I really haven’t seen that sort of unnoticed strangeness, including in formatting. It’s a bit ironic because Cursor Tab was one of the bigger reasons I signed up for Cursor… but over time it became less helpful and more something I’d be continually contradicting.
But when I try formatting the same code inside Cursor AI; the output sometimes looks slightly different, especially with long pipeline chains and spacing around function arguments.
Could it be related to Cursor AI applying its own formatting rules on top of Elixir’s defaults, or maybe some hidden config I’m not aware of? 
