We work on a fairly large ASP.Net MVC portal project that is about 300k lines of code. It has 20 or so related projects that use its API and design language that are somewhere north of another 500k lines of code. We are coming up on 2 years since its release this summer and have been planning for the next UI refresh. The current version has a mobile/responsive design, though it was done as an afterthought last year and was not originally designed with responsive in mind. Shawn King along with feedback from the rest of the team designed an incredible new UI that was created “mobile first” that we were planning on releasing this summer. I am sure he will write a much more detailed post about the design process soon so I won’t go into the specifics here.
Though we practice continuous deployment, we can’t make major interface changes during the school year. We have one short window in late May before the summer semester starts to roll out drastic changes. It is acceptable to deploy a new tool or add a feature here and there but a full on UI change mid semester would bring about an uprising similar to what Facebook experiences when they make an interface change.
I want to be realistic about what we can accomplish and avoid a death march. After reviewing what we have in the pipeline for the May release and considering the time needed for testing, I realized that it isn’t going to be possible to implement this UI refresh with any level of confidence. That means it will be another 16 months before our users get to see it. I think the testers will be happy, and the developers are relieved, but it makes me sad.