On Oct 1, 2013, at 6:03 PM, Joe Darcy <joe.da...@oracle.com> wrote: > On 10/1/2013 4:18 AM, Paul Sandoz wrote: >> On Sep 20, 2013, at 5:30 PM, Martin Buchholz <marti...@google.com> wrote: >>> Looks good. >>> >>> Random nitpicks: >>> >>> I prefer the old variable name DOUBLE_UNIT to DOUBLE_ULP; else you are >>> inventing a slightly different meaning for ULP. >>> >>> I think using hex floating point literals is preferred: >>> >>> - private static final double DOUBLE_ULP = 1.0 / (1L << 53); >>> + private static final double DOUBLE_ULP = 0x1.0p-53; >>> >> I don't have any strong preference on the above comments, Doug? > > Please use the hex literals in a case like this; it is exactly the situation > they were designed for. >
OK. Updated: - to use hex floating point literals for all the random impls (i retained the other way in comments, since devs may not be familiar with this). - reverted DOUBLE_ULP back to DOUBLE_UNIT. http://cr.openjdk.java.net/~psandoz/tl/JDK-8025136-SR-enhancements/webrev/ Paul.
