As you know, I'm the main (and more or less, at the moment the sole) developer for EclipseFP. It's going reasonably well, I'm putting a few enhancements here and there, and there are a few big things that I'm planning to do that would be exciting to do and hopefully to use.
However, I can't shake off a nagging feeling that maybe my energy is not 100% useful to the community. The compelling argument of EclipseFP is that it allows developers that are familiar with Eclipse to develop in Haskell as easily as possible: they can concentrate on writing their Haskell code and not have to learn another IDE to do it. But I'm wondering if really there is an audience (apart for myself).
- Unix Hackers will use Vi/Emacs and will scoff at the very notion of using an IDE.
- People really into Haskell would probably prefer an IDE written in Haskell like Leksah, especially since such an IDE actually proves that developing GUI programs in Haskell is possible!
- People that are in the Eclipse/Java world and want to go into functional programming can do so in Scala, Clojure and of course now Frege. Frege now boasts an Eclipse plug-in as well, so really the incentive is great: you can use a Haskell-like language in Eclipse, with probably better integration that what EclipseFP can achieve since everything in Frege boils down to Java. I suppose you could even write the Eclipse plugin code in Frege.
I'm even tempted to make the jump myself. Maybe with Frege I'll get a ecosystem that's doesn't consider Windows as a second grade platform. But can it deliver, in terms of power, performance, reliability?
One thing is clear: Frege is moving forward, which is more that can be said, as far as I know, of others Haskell on the JVM efforts (even the Wiki page says that it'd be a hard job to do, obviously the Frege developers didn't let themselves be scared).
I'll be interested in hearing feedback on people's experience with Frege, and I think I'll take it for a spin myself soon. I'll post my experiences!