Hi folks. I just read a rather interesting paper about a fork of GHC that performs "optimistic evaluation". This shows big wins in some cases. The authors claim to have implemented this in a fork of GHC and promised that it would be integrated into the production compiler "in the near future". Curios, I investigated the GHC wiki... Well, let's see now. The front page has links to "out release plans for 6.8.3" and "what will be in 6.10". The latter page, http://hackage.haskell.org/trac/ghc/wiki/Status/Releases mentions that a beta has just been released... and seems oblivious to the fact that 6.10.1 is production now. Clearly, any hopes I might have had of getting a handle on the current status of GHC from this wiki were dashed somewhat. ;-) Does anybody have any suggestions for a more reliable way of figuring out what the current activities and plans for GHC are?
andrewcoppin:
Hi folks.
I just read a rather interesting paper about a fork of GHC that performs "optimistic evaluation". This shows big wins in some cases.
The authors claim to have implemented this in a fork of GHC and promised that it would be integrated into the production compiler "in the near future". Curios, I investigated the GHC wiki...
Well, let's see now. The front page has links to "out release plans for 6.8.3" and "what will be in 6.10". The latter page,
http://hackage.haskell.org/trac/ghc/wiki/Status/Releases
mentions that a beta has just been released... and seems oblivious to the fact that 6.10.1 is production now. Clearly, any hopes I might have had of getting a handle on the current status of GHC from this wiki were dashed somewhat. ;-)
Does anybody have any suggestions for a more reliable way of figuring out what the current activities and plans for GHC are?
I'm assuming you're talking about the 'eager evaluation' papers? (There were a few). I think the verdict was that the runtime machinery was too complex for the performance gain. -- Don
participants (2)
-
Andrew Coppin -
Don Stewart