On Nov 19, 2007 1:21 PM, Michael Busch <[EMAIL PROTECTED]> wrote: > Yonik Seeley wrote: > > On Nov 19, 2007 11:38 AM, Michael McCandless <[EMAIL PROTECTED]> wrote: > >>> If we opt to treat payload like termBuffer and copy the bytes, then we > >>> need no offset member. > > I'd argue that the current approach (creating a very lightweight wrapper > object that is immutable for every payload) should yield a better > performance than a buffer+copy approach (admittedly without having > numbers that proof this).
Immutable implies that the user needs to do a new byte[] for every payload, yes? It seems like this would be slower if payloads were common and faster if very rare. -Yonik --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
