On 6/7/07, Aaron Bentley <aaron.bentley at utoronto.ca> wrote: > Does this, itself, need a test? I don't think the risk of it regressing is so high that we need one. -- Martin