Talk:Sieve of Eratosthenes: Difference between revisions
→Recent edits to Haskell and my reversal of them
Line 258:
: also, "naive" version is ''not'' based on Euler's definition. The primes in "primesEQ" are not separate, but the same - this is by design; for the reader to see the difference when separate supply is used later, in the more efficient versions. Maybe we ought to add types as you did in all the versions, not as it is now in the most efficient tree-merging version only. Will think it over... -- [[User:WillNess|WillNess]] ([[User talk:WillNess|talk]]) 14:16, 10 April 2015 (UTC)
: also, writing "primesEulerQ ()" (with the ()) is not guaranteed to prevent memoization. A compiler could "optimize" even this code, recognizing it as "common subexpression", causing it to be memoized. Only "_Y" guarantees this (well, so far, as I could see, for various versions of GHC). [[User:WillNess|WillNess]] ([[User talk:WillNess|talk]]) 14:29, 10 April 2015 (UTC)
| |||