-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 9/2/10 20:11 , Alec Benzer wrote:
I guess I would then be concerned with why they didn't allow it in the standard (though I guess "well, it seemed to be a good idea at the time" answers that).
Keep in mind that type theory has advanced a *lot* since Haskell '98 was frozen. Quite a few things that are commonplace in modern GHC were considered prohibitively difficult or expensive to implement back then; others were adjudged "too confusing" (which got monad comprehensions removed and the monomorphism restriction added; the latter is almost universally considered a mistake).
I think I also would want to avoid doing things not in the language standard on principle, since my instinct would tell me that if only a particular compiler implements, I shouldn't use it because it'll produce non-standard code. Though this sort of comes comes from C/C++ where there are different compilers on different platforms, but I guess with haskell people pretty much use ghc everywhere?
More to the point, ghc is the language developers' playground, so other compilers (the few that there are) generally follow ghc's lead. - -- brandon s. allbery [linux,solaris,freebsd,perl] allbery@kf8nh.com system administrator [openafs,heimdal,too many hats] allbery@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.10 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkyATt0ACgkQIn7hlCsL25UFWACcDYjiefvBZl1tH96XR5iOuWt3 Y9gAnjAYZt1UjfOgthSxG72Norny//ng =8tc2 -----END PGP SIGNATURE-----