Although the Industrial Haskell Group has yet to garner its first industrial member since its inception almost two years ago, they have managed the impressive feat of getting paid to remove a feature from Haskell. Specifically, to make it easier to build programs written in Haskell that do not rely upon the GNU Multiprecision library for arbitrary-precision arithmetic (bignums).
We made this interesting observation when considering adding bignums using GMP as a primitive type for HLVM. Apparently, having bignums in the language is not very useful beyond irrelevant microbenchmarks like computing the digits of π.
The slides here also criticize the CAML Consortium (which has garnered 11 members) for charging too little and states that the IHG aimed to garner five members each paying £12k per annum. Why has this target not yet been reached? Our guess is insufficient sales and marketing directed at decision makers in industry who could benefit from using Haskell. As an aside, we believe this same mistake is why the founders of Stack Overflow found it so difficult to monetize despite having millions of non-paying users. In contrast, Rich Hickey managed to garner funding from a whopping 427 people and several companies for his own language, Clojure.
Regardless, the fact that they are trying to build a business around the development of Haskell itself is admirable and should at least prompt more professionals to take a look at what is on offer.