Amie and I were talking about Wes Molebash. Wes draws cartoons. In fact, Wes is the guy that made this:
Wes has recently decided to shift from making a web comic to making a children’s book. He’s documenting the process and had one post where he drew the same character over and over again. Amie commented that she never thought about if you were making a book, that being able to draw the character consistently from all the various angles would be necessary.
When people don’t understand about how something is made, they have a tendency to oversimplify it to basic components. And for some reason, mobile apps seem to have it really bad. Here’s why:
Usually when someone comes up with the idea for an app, they can state it in one or two sentences. But those sentences usually infer stuff that, unlike literature or a song, they must be done. If they say “download from the Internet”, that means something has to serve up the data. If it’s not something that already exists, then it requires that the server part be written.
And then there’s all the stuff that has to occur when stuff doesn’t work. On occasion, I’ve whipped up some applications over the weekend as a proof of concept. What I do with these is I’m usually the one demoing them, and I show how the app functions when it behaves properly. For the purposes of the demo, some things are faked. Other things I know won’t work, so I avoid doing them. But there’s so much to consider on the error path for a real application. What happens if I can’t send something? Should I try to do it automatically later? Or will doing it automatically later not make sense? What if it’s optional to do it later? Then I need to provide an interface so that the user can choose to do it later or forget about it.
And interface design is tricky. In PTO Tracker, people email me often for things that are specific to their company’s policies. I never knew how vast PTO policies could be. But with these requests, I need to find a balance somewhere. Can I make the app handle everything? Yes, but that leads to a complex setup where 100% of the users have to answer a ton of questions just so that it functions perfectly for 0.1% of users. The balance here is to forget that 0.1% so that it’s better for the 99.9%. Of course, I don’t always take the easy path. It depends on the impact for that 0.1% user. Is it something that won’t occur often anyways? Or if it does, is it just annoying? And if you’ve ever had to do 10 clicks to do something, you know how an overly complex design can be awful.
Do you do something that seems to always get oversimplified?
