ghcid VS ihaskell

With a combination of ghcid, an hiedb filewatcher and the -fdefer-type-errors flag you can get pretty solid IDE behavior. Currently only ghc 9.4.4 is supported but happy to personally help people set this up if interested!

In addition, for Haskell, I usually have ghcid running, which likewise re-runs on every file change, but gives faster feedback about any type errors than the full compiler, and also is configured to evaluate

VS Code + Haskell extension is usually best, but ghcid is an alternative which is much simpler, easier to set up, less pretty and powerful but still pretty easy and effective to use. Here's a workflow:

you should also checkout the ghci vim plugin https://github.com/ndmitchell/ghcid/tree/master/plugins/nvim

I just went through the same thing. I settled on using stack and ghcid. All it does is recompile on any change to source code so you at least get lightning fast feedback. Both stack and ghcid have been easy to install and use so far.

Have you tried out ghcid? It basically just runs ghci on your program every time you save, and gives an updated list of errors and warnings. Not interactive in the sense that you don't manually test your functions with it, but like 95% of debugging in Haskell is just fixing errors at compilation time. I find it to be a very nice developer experience. Just need a text editor and a terminal with ghcid open and you get immediate feedback as you program.

https://github.com/ndmitchell/ghcid

Are you referring to IHaskell: https://github.com/IHaskell/IHaskell?

Note that we do have the IHaskell kernel for Jupyter, so we don't have to be that jealous.

More info: https://github.com/gibiansky/IHaskell/issues/1251

I'm stymied trying to import a package. It's Graphics.Matplotlib. (See here.) I'm wanting it for my IHaskell, but when I add it to the IHaskell stack.yaml

That looks like a generic front-end error for when the back-end is unavailable, the back-end error should be more informative, but I don't know where exactly you can find it. At this point it might make sense to open an issue on the issue tracker of IHaskell, they will be able to give you more useful answers.

​ curl -sSL https://get.haskellstack.org/ | sh git clone https://github.com/gibiansky/IHaskell cd IHaskell pip3 install -r requirements.txt stack install --fast ihaskell install --stack ​ jupyter labextension install jupyterlab-ihaskell # did not work ​ stack exec jupyter -- notebook

I haven't actually used IHaskell, however, the png file is presumably created in whatever the working directory is when the script is running. The IHaskell wiki says: