I still think this is the best way to do Web or GUI apps, but I just don’t do enough work in this space to justify keeping this as a goal. There are too many other things that are closer to my domain.
So, I’ve moved to the next stage: HTML mockups. The way things progress using this technique is fascinating—I’m already understanding the requirements and interface much better than I had before (in any of my other attempts to implement this beast).
I’m going to drop back to paper and pencil now, though, and start sketching out the pages for some of the more “core” functionality of the application. Great fun!
As I’m nearing the end of the sketches, I wonder what I should do about the back-end process I need to deal with. I could sketch the data flow, then build stories to implement from that. I’d love to hear what other people do though.
I’ve just about sketched out all the screens for the first version of my new app. I’ve noticed three things:
1) Sketching out the screens has made the app’s design and (future) implementation very clear. I’m really looking forward to starting to write some tests and get coding.
2) The sketches have also made me realize where I’d skipped important functionality in my mental doodling. Making the changes now while things are ‘just pictures’ has lowered the cost of change dramatically.
3) It’s really hard to keep from sitting down laying out the DB tables and writing code—everything just looks so easy at this point.
Now I just need to talk to someone with some real design skills to walk through the design sketches with so that I don’t end up with angry fruit salad.
After listening to DHH talk about his preferred development model at about a Seattle.rb meeting and reading about Jamis Buck’s experiences trying to follow DHH’s lead, I’m ready to start.
On the bus ride in to work this morning, I sketched out several screens for an app I’ll be writing (with Rails). I’m pretty psyched.
