3 Nov
2014
3 Nov
'14
12:40 p.m.
On Mon, 3 Nov 2014, Niklas Hambüchen wrote:
On 03/11/14 18:16, Henning Thielemann wrote:
If 'seq' is faster than 'deepseq' it leaves things unevaluated and thus builds up unevaluated expressions, right?
Yes.
What I mean is you could write a function function (using repeated foldl' inside) that is written such that its return value is fully evaluated (by carefully using just as many seqs as necessary for that).
If foldl' did a full deepseq on its return value, then this function would seq the same data more times than necessary.
I see.