It's a 48 point scale with a sliding base value that changes on a daily basis. With enough time and data the base values could be figured out, but for all real purposes it's obviscated. Using a pure time delta would not work, most reports involve three or more accounts and deltas just don't really work for that. As it stands the time values are shifted an unknown value. Any other method would make it either useless or would require a massive rewrite of code
On Friday, April 1, 2011, Daniel Schwen lists@schwen.de wrote:
You could instead just output a correlation-factor for the two edittime curves (integrate the product for example). That would reduce the amount visible data drastically, making it very hard to tell anything about the editing habits of each individual. Daniel
Toolserver-l mailing list (Toolserver-l@lists.wikimedia.org) https://lists.wikimedia.org/mailman/listinfo/toolserver-l Posting guidelines for this list: https://wiki.toolserver.org/view/Mailing_list_etiquette