Difference between revisions of "Talk:Performance/Data types"

From HaskellWiki
Jump to navigation Jump to search
(newtype optimisation clarification request)
 
Line 1: Line 1:
 
* In the newtypes section, could the first paragraph be clearer about the caveat "in most cases"? That is, could it point when using a newtype ''will not'' be optimised away? I myself was (am) under the impression that newtype is always optimised away. When can't/won't it be?
 
* In the newtypes section, could the first paragraph be clearer about the caveat "in most cases"? That is, could it point when using a newtype ''will not'' be optimised away? I myself was (am) under the impression that newtype is always optimised away. When can't/won't it be?
  +
  +
** It is always optimised away, yes. The report requires that. [[User:DonStewart|dons]] 02:41, 7 May 2008 (UTC)

Latest revision as of 02:41, 7 May 2008

  • In the newtypes section, could the first paragraph be clearer about the caveat "in most cases"? That is, could it point when using a newtype will not be optimised away? I myself was (am) under the impression that newtype is always optimised away. When can't/won't it be?
    • It is always optimised away, yes. The report requires that. dons 02:41, 7 May 2008 (UTC)