On 11/23/13, 11:31 PM, Sean Owen wrote: > On Sun, Nov 24, 2013 at 7:22 AM, Phil Steitz <phil.ste...@gmail.com> wrote: >>> Are there any objections to changing the API sigs from foo(double[] values) >>> to foo(double... values) ? >> Yes. I am -0 for adding new methods that take varargs; -1 for >> removing the double[] versions. > Phil note that foo(double... values) actually has the same signature, > foo(double[]), in the byte code. This would not amount to adding or > removing methods, or even a signature change.
I really prefer to keep a set of methods that take double[] and double[], int, int cleanly and simply with no ambiguity about what is being passed / what is accepted. If we really want to support varargs, lets use other method names please. > > The downside is the inconsistency between declaring both > foo(double...) and foo(double[], int, int) although you could argue > that in the latter case, you're definitely meaning to operate on an > array and that's why it's different. Yes, the methods are designed, like many others in [math], to operate on arrays. Phil > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
