That may not be such a bad idea, we have a controlled space to operate in and we are targeting a finite set of beans representing a well defined spec, all of which should eliminate much of the cpu and gc traffic.

The effort involved getting json lib configured was not exactly lightweight.



Ian

On 10 Feb 2009, at 19:43, Kevin Brown wrote:

We could also write our own parser. It's not a terribly complex format.

Reply via email to