> There's no point in spending any more effort on estimation than is needed to support the planning required to make the business a success; any more effort tha

> There's no point in spending any more effort on estimation than is needed to support the planning required to make the business a success; any more effort than that is simply wasted. This is basically what I argued. Software businesses often have such fantastic margins that you can do any old thing and still rake it in.