More road blocks and learnings

I've decided to start documenting my learning's as I go more regularly, so hopefully I actually do learn from them and won't make the same mistakes over. I went through a second attempt pretty quickly. I looked over my requirements docs and thought they looked good enough, so I started designing the database again. I started using Visio as I said I would, but then it started looking extremely like the database designer diagrams I built before, so thought why have that extra step. So I did it all in the SQL Server database diagram. It went really quick because I was able to use most of the previously designed database.

I then quickly created the data access layer by drag/dropping the tables onto a dataset. It created all of the data table adapters for me quite quickly. With that my intention was work on the business logic layer by creating a class 'on top' of each table adapter, but I soon found the job quite daunting. With the great code generators Visual Studio has with the dataset designer and class designer I was hoping I could find a way to generate the business logic layer some, like I did with the data access layer. Searching the web I didn't find much, but I did come across a document in the Jumping Into ASP.NET series that outlined the "Application Planning and Designer". I thought that maybe with the road blocks I kept hitting I would look at this stage again so I read it. It makes reference to using User Scenarios, which I have done some of, but after reading this section I realized I hadn't done nearly enough. So I have begun creating these and have recognized several other major road blocks and hurdles that I would have hit on the path I was headed. There were complete other systems within the system I am attempting to build that I hadn't considered completely enough.

So now my current objective is to complete the User Scenarios so there is a scenario for every interaction the user may encounter with the system. I don't know if this is over doing it or not, but I think it will help to ensure every aspect has been thought about. I have started a list of User Scenarios I am going to write and already there are about 20. In the article I read the author had written just a couple examples of a fairly simple application, but mentioned that it would probably take about 20. So if that simple application could have used 20, then I'm assuming this one I am working on could take many more.