Open CyberShadow opened 7 years ago
I haven't noticed any issues with haskell-mode
and magit-blame
; that said, I'm not using haskell-indentation-mode
(I prefer shm
).
Can you maybe try with haskell-indent-mode
and see if that solves anything?
I've done zero setup other than installing haskell-mode
as per https://wiki.haskell.org/Emacs.
If haskell-indent-mode
is the preferred option, then perhaps it should be made the default?
I keep forgetting which is the preferred option; I just know it's the other main one available and trying that could help narrow down whether it was the indentation mode at fault or some other configuration error.
It works with haskell-indent-mode
.
So the issue is definitely with haskell-indentation-mode
; at least we know it's not all of haskell-mode
that's broken ;-)
@CyberShadow: For mundane reasons haskell-indentation-mode
takes over TAB, Shift-TAB and RET. Mechanism provided by indent-line-function
works best for languages that have indentation derived from parentheses, i. e. non-semantic indentation. That is the explanation for current state.
Now, it would be much better if haskell-indentation-mode
did not intefere with anything. Anybody out there willing to take a look?
Running
magit-blame
on a file withhaskell-mode
causes the RET key, which should normally runmagit-show-commit
, to instead attempt to runhaskell-indentation-newline-and-indent
(which doesn't do what is expected and fails).Instead of changing the key binding of the RET key in
haskell-indentation-mode-map
, canhaskell-indentation-newline-and-indent
be implemented on top of the regularnewline-and-indent
by providing its customindent-line-function
?