On Thu, Feb 5, 2009 at 4:58 PM, Don Stewart
A comparative analysis of the 10+ Haskell web frameworks would be awesome.
happstack, wash, fastcgi.., turbinado, perpubplat, riviera, salvia, kibro, ella, what was that one launched yesterday? *ah, yesod...
Wow, that's a lot more frameworks than I realized... Anyway, I'm the guy who just started that new one, yesod. I don't think it's a good thing that there are so many of these unfinished web frameworks out there, and starting more is probably not that helpful. I did so because, as far as I can tell, none of the other frameworks out there are even remotely interested in the same goals as me. Happs(tack) seems to have thrown all convention about web frameworks out the window, which is fine, but not exactly usable in many conventional scenarios. Wash, fastcgi, and kibro all seem very low-level to me. Wash and fastcgi seem to correlate pretty well to ASP and FCGI.pm to me, which frankly don't provide much more than basic help in creating web sites. Turbinado seems to be the closest to what I would imagine as a mainstream framework. However, it is ignoring some features currently that I consider vital (correct me if I'm wrong): * Can run on a shared host. This means something like FastCGI, not its own server. * It's not aiming for any high-level Javascript/AJAX integration. I frankly think the generating plain HTML on the backend style is a little outdated at this point, and think we need to start exploring more advanced web development techniques. That said, if there's an existing web framework that has any interest in those two goals, I'd be happy to contribute to that instead of fracturing the community any more. It would be great if we could get some serious support behind a single framework. Obviously, I have more opinions than the two I just listed (I previously linked to a 14-point blog entry on the subject), but the others I would be willing to forego to achieve unity. Anyway, that's my plea. Sorry if it hijacked this thread too much. Michael