[Well-Typed Blog] Funding GHC, Cabal and HLS maintenance



We are all grateful for how engaged well-typed is. Definitely an excellent citizen in the Haskell ecosystem.

Also great seeing Zubin doing relentless release work on HLS.


Thank you all for all the effort you guys put into GHC!

A question I have as a Haskell-using startup sponsoring the Haskell Foundation: will funding the HF subsume all other ways of funding Haskell and its ecosystem development or is it not that simple? In other words: do we, by virtue of funding the HF, fund all the amazing work you guys do?


The HF has received a generous donation from GitHub that is earmarked for support of the Well-Typed GHC team. Additionally, we have hired them to perform additional GHC work that supports HF initiatives (such as adding error codes to GHC in support of errors.haskell.org, and I’d like to do more of that. They have provided excellent service at a discounted rate for things that help the whole Haskell ecosystem.

Sending them money directly is a good way to get it in their hands for use at their discretion for maintaining GHC, Cabal, and HLS. The HF has broader priorities, and also works on other things (like the GHC CI infrastructure, the error message index, etc). I’m happy to have a chat any time about our specific priorities.

Thanks for sponsoring us!


Yes, we’re very happy to be working with the HF and are supportive of its mission. Indeed, Well-Typed are sponsors of the HF too! As David says, they have broader priorities and are driving forward improvements and raising the profile of Haskell in all sorts of different areas, which is great.

But precisely because the HF have broader priorities, it’s also valuable if Well-Typed also have committed funding that goes directly to pay for the time of the engineers maintaining GHC and related tooling. So it’s good to support both (but of course I’d say that). :grin:

One other aspect is that Well-Typed can offer consultancy/support and work on specific features or fixes, which for some companies is easier to justify than general sponsorship. Recently that has included quite a bit of work on debugging/profiling tooling (e.g. ghc-debug, late cost centre profiling) and compile-time performance of GHC/HLS, all of which I think will be great for the community as a whole.