In some sense, for tablets the browser is a killer app. Maps is a killer app to some extent. Being able to share the screen with other people -- that it's a social device -- also might fit the bill. I think that for tablets, there isn't and won't be one killer app for everyone. It's more that there are apps that are killers for individual people. It's the sum of all those that is the killer app. This has been true since the original Palm Pilot.
Advice Line: In your note-taking app, you don't offer handwriting recognition. Is this because it's just too hard for a small software company to develop the technology, or do you think it isn't all that useful a feature now that just about everyone grows up using a keyboard?
Dan Bricklin: First of all, it's a complex thing to do. You work at it for years to get it right. Architecturally, either the company that provides the OS builds handwriting recognition into it, or we'd have to license someone's engine. If we license it, we have to charge for it. And then there are patent issues.
Another thing: I've been in the ink world since the '90s, and I've found that calling the problem "handwriting recognition" oversimplifies the challenge. Here's why: Not all of the information is in the text of the writing itself. Some of the information is two-dimensional -- how it's placed on the page -- so just transcribing the text doesn't do the job. Where it is, and what you circle, underline, and so on have a lot of meaning. Searching is probably more important, and that takes a different kind of recognition. Then there's the challenge of correction, which is more complicated than you might think, because with handwriting, corrected text is in the right position, but entered way out of sequence.
Anyway, I'm not sure it's all that important. Until someone is able to figure out how to recognize handwriting without losing the spatial context, it will probably be more of a niche need than something broad enough to justify the investment it will take to develop. I haven't gotten around to it yet. The way I look at it is, when a user needs the text to be perfect, they can use a keyboard. If they're standing or a keyboard isn't practical for some reason, they can dictate and use a speech recognition engine.
Advice Line: More about note-taking: From the small sample I've looked at it appears that except for OneNote, everything in this space builds on a page metaphor rather than an infinite scroll. Why did you make this design choice?
Dan Bricklin: One reason is memory. A scroll requires a lot more of it, or more sophisticated memory management. The other is the need for PDF output. That puts you in a page metaphor anyway. And you have to have pages when sketching is part of the feature set.