[Haskell-cafe] Strange type behavior in GHCi 6.4.2

[Bulat Ziganshin]

Hello Kirsten,

Friday, December 29, 2006, 6:30:22 PM, you wrote:

> I suggest *not* using these pragmas unless a combination of profiling
> and reading intermediate code dumps suggests that foo -- and its
> un-specialized nature -- is truly a bottleneck.

it's a matter of taste - and experience. may be it's simpler to add a huge
number of INLINE pragmas than to profile and especially read code dumps?



-- 
Best regards,
 Bulat                            mailto:Bulat.Ziganshin at gmail.com