Thats pretty much how serialization works now (with a byte counting outputstream) Except that serialization doesn't count transient fields (which ofcourse also cost memory) And you kill some overhead also by doing this.
johan On 2/25/07, Eelco Hillenius <[EMAIL PROTECTED]> wrote:
Hi, What do you think of this approach http://www.javaspecialists.co.za/archive/Issue078.html to calculating object sizes opposed to doing it using serialization like we do now? Eelco
