That’s my point, that it forces you to specify either true or false. A lot of people just make it default to true in Conn.Case but I don’t really love that either and prefer to always see the explicitly given option.
EDIT: Maybe that’s what you’re saying… better to just default to true and you then it’s clear that when you specify false it was intentional?
Super useful article! Just bookmarked it for future reference. Thanks for sharing.
Seems like there are some useful checks there. Currently I’m working solo, so it’s a bit easier to achieve consensus (mostly just a handful of demons are awake in my head at a time 
).
Actually that’s a way better idea to default to true and have people explicitly state it as false. Why did that never occur to me?
Ya, it’s definitely totally fine to do that. I just like the Credo rule as it’s the least surprising thing. When I worked on a codebase where async: true was the default, I was unaware of that fact for a bit so I would put it anyway. Not a huge deal, but it causes inconsistencies. It’s also possible (though unlikely) that someone could be writing a test that needs to not be async and they think it is. This is all very, very minor stuff and can be solved with good communication (though we all know that can go). In any event, being required to put it just takes anything like that right out of the equation.
is there any way that credo issue can be fixed automatically. like
what can be fixed.
Like fx instead of fx()
and ordering of aliases?
