When we were first learning about haskell, we learned that do-notation was just syntactic sugar:
thing1 x >>= \a -> thing2 a y >>= \b -> thing3 b z
As we progressed, we probably learned that Applicative is often a better abstraction because it offers less power
(in the same way that pure languages are better because they don’t offer the power of mutation).
However, it can get syntactically obnoxious working with applicative sometimes:
(\xFld yFld -> MyRecord{xFld,yFld}) <$> getX <*> getY
So we learned to enable ApplicativeDo and write
do
xFld <- getX
yFld <- getY
pure MyRecord{xFld,yFld}
Recently I found myself working with Functor (because less power → more flexible) in the same way I might work with Applicative.
There is no language extension to make functors look nice.
However, I remembered my >>=, and noticed the similarity with <&>:
foo x <&> \x' -> bar x' y
As it happened, I could start by writing the more familiar monadic algorithms.
Once I made sure it worked, I retrofitted it to the functorial case with very few keystrokes thanks to <&>.
So thank you <&>! You are a wonderful little tool!
P.S. There’s probably something to be said for your friend & as well…