pandoc has HALF A GIGABYTE of dependencies. Choose better doc generators, dammit

@sir true, for end user software. It's insanely good at specific things, like building parsers (parsec), but yeah the bloat is to much to swallow for most situations.

@sir @jb55 yes that's also something you can make parsers with.

@sir @jb55 I don't think you can compare Haskell with yacc. The people at GitHub working on semantic explained why they use Haskell[1], and I doubt that they'd be better off with yacc+C.

Also, pandoc is a beast with support for almost every widely used document path. This is less a problem with Haskell, and more a problem with using a less featureful doc generator like haddock, which is actually the standard tool used for GHC and supported by Hoogle. Actually, pandoc isn't even a documentation generator but rather a document converter, so I'd imagine that haddock is still used to generate the docs and then pandoc to produce PDFs or something, in which case it *should* be an optional dependency.

[1] github.com/github/semantic/blo

@sir Closely related language Idris looks better than Haskell if we are talking about practicalities like executables size etc.
It may be somewhere between C and Go in this regard.
And it is striving to become systems language eventually.

I believe it is as capable of parsing tasks as Haskell without crazy overhead (MB-sized helloworld) of ghc.

Sign in to participate in the conversation
Mastodon

cmpwn.com is a private Mastodon instance for friends of SirCmpwn.