Hi Bulat, On Tue, Jan 29, 2008 at 07:17:44PM +0300, Bulat Ziganshin wrote:
Tuesday, January 29, 2008, 5:56:48 PM, you wrote:
I don't have much of an opinion about the particular function being discussed, but I'm strongly against freezing the base library at this point.
reasons?
I don't want to get into a long discussion about this, because I don't think we're going change our viewpoints, but the reason is because I think that freezing it would cause more pain overall. I would like to see some significant changes to the base package, e.g. in how exceptions are implemented (to avoid the gigantic import cycle that the current implementation causes, and also because I think Simon's proposed replacement is nicer to use).
when it should be freezed?
I'm not sure it ever should, but it would have to be an awful lot smaller first in my opinion.
as an example look at ghc itself - it's full of #ifdefs and borrowed libraries. is that organization is recommended haskell programming style?
GHC tries to be buildable with just a plain GHC install, no extra or upgraded libraries, going back as many GHC versions as possible. That's because of the hassle of bootstrapping it again if you fall too far behind. I wouldn't recommend for other Haskell code to go to such lengths. Thanks Ian