[bitc-dev] deeply immutable functions and Haskell's ST
Jonathan S. Shapiro
shap at eros-os.org
Thu Aug 25 14:49:05 EDT 2005
On Thu, 2005-08-25 at 03:54 -0700, Richard Uhtenwoldt wrote:
Thanks for telling us about ST!
> now, I realize that you guys are not planning to embrace monads,
> and you are probably a little turned off by the thought that this
> solution depends on a monad, so please give me a few paragraphs
> to try to set your minds at ease.
Hey, if it were purely a matter of my own preferences, I'ld be okay with
monads. I've done enough dataflow stuff and enough SSA stuff that it
doesn't really confuse me. My concern with monads is usability for other
[ Example with EQL and SomeGlobal omitted ]
I need to digest your discussion when I am slightly more awake, but my
summary response is that I'm having trouble understanding the example
because I don't know enough about Haskell to understand what readIORef
and readSTRef do.
Also, I may have been overly hasty about (deeply-immutable T). We had a
discussion today concerning (val T) that was revealing, but it is
definitely not something I want to go back into right now -- we need a
More information about the bitc-dev