12 May
2008
12 May
'08
8:05 p.m.
Matthew Naylor wrote:
I wonder to what extent deforestation supersedes such optimisations. This would be a good topic to raise with a Cleaner. The paper Neil mentions looks like a nice alternative to uniqueness typing -- and it appears that there will be a FitA talk about it, this Thursday.
What if you have, say, a giant array, and you run a loop that updates the entire array. Rather expensive to keep copying and collecting all that space. That's why most array-based code is explicitly in-place. But wouldn't it be nice if it didn't have to be? Similarly, there are recursion patterns for which fusion isn't very easy. (How would you fuse, say, a Gaussian elimination algorithm?)