Made non-finite floats throw NOT_SUPPORTED_ERR

I’ve changed the spec now to require throwing NOT_SUPPORTED_ERR for
non-finite floats, and also required -0.0 to be treated the same as
+0.0, for ACTION-2082.

(Sorry for the member-only link.  Do we have a process yet for working
with the spec in the public CVS repository?)

I noticed though that the spec at the moment doesn’t say anything about
throwing TYPE_MISMATCH_ERR for setFloatTrait() when given a non-finite
float.  Andrew, were you saying that setFloatTrait() also must throw
this exception here, or was it just for getFloatTrait() and



Cameron McCormack ≝

Received on Friday, 4 July 2008 02:40:43 UTC