OneNote and Journal

Those of you who have a TabletPC are probably familiar with the built-in note-taking application called Journal. You're probably also confused as to how Microsoft could release two programs (OneNote and Journal), that seem to behave so differently, and don't even interoperate well. The answer to that is everyone else who does not have a TabletPC. Let me explain.

First, if you do not own a TabletPC, you are in the great majority, and you probably don't have many serious complaints about the way ink is supported in OneNote - that's because you don’t have a good way to generate the ink in the first place (yes, you can use an external digitizing tablet, or a mouse or trackball, but these are clearly inferior to just being able to write on the screen where the ink appears, as you would with paper or a TabletPC). The point is that you see OneNote as a great place to type and collect text and images by drag/drop or copy/paste. You can even record audio synced to your typing, manage to-dos and important items with note flags, etc. In short, almost all the features of OneNote work great on a laptop or desktop.

Now if you do own a TabletPC, and have used Journal, to you it seems the most natural thing in the world that another product from Microsoft that claims to do note-taking and support the TabletPC should be a seamless extension of the built-in Journal application. Clear as day.

So why isn't it? The answer is partly in what I wrote below about The Myth: if you think OneNote is only for the TabletPC, then you're right it is madness to make it different. But it is not only for the TabletPC. Of course we wanted to make the handwriting experience ideal, but as I mentioned in Handwriting and Humility, we didn't get as close as we wanted to. But these don't seem like good explanations - they’re more like excuses, right?

Well, here's what happened (from my perspective and recollection of course). The TabletPC team was formed from one of several groups in Microsoft that I'll call think-tanks. They are collections of people (smart and good-looking of course) who don’t build products. They play with ideas, prototyping them and trying to see what would truly be innovative and useful by playing with actual code and devices, not just thought experiments. As you probably know or can understand, when you actually handle and use a thing, you get many more ideas than you do just by trying to imagine it existing. These are the people who had digital movie servers in their living rooms in 1994, and used interactive television in 1995. Of course pen computing had been around a long time. But the team that became the Tablet team wanted to take a fresh approach and think about the problem end to end from a customer point of view.

BillG is a big fan of Tablets since he believes in the natural pen interface for activities like annotating web pages, reviewing documents, note taking, gesturing, etc. Thanks in part to his interest a product team was formed to turn the Tablet into reality. Some of the principals from the think tank remained attached, and the rest of the team was built up of people from around the company who had experience building and shipping products (several from Office). They got the religion, and worked really hard to build hardware prototypes and software that would make the Tablet into a great product. They used the hardware prototypes to convince OEMs (computer hardware companies) that the hardware was feasible and interesting. Ironically, some of the early designs from the OEMs themselves were just clones of the prototypes that the Tablet team gave them. The dirty not-so-secret of the computer world is that no one makes much money in hardware, so with razor-thin margins they save money by not doing research beyond that they need to integrate new chips and hardware, which is usually only incrementally better than last year's hardware, because their suppliers have the same problem. In fact, most computer companies do no research whatsoever and just assemble parts or pay other companies to assemble the parts and they just sell the completed units. So companies like Microsoft and Intel increasingly do research for them and pass it on to make the whole ecosystem move forward (USB, Tablet, optical mice, etc are all ideas from Microsoft or Intel)

The Tablet software team was made of two parts. The founding group was from the think tank, and they were focused on handwritten notes as the killer app for the tablet. They already had prototypes of what came to be known as Journal. Separately, there was a pre-existing handwriting technology team in Microsoft that had done work for the PocketPC and other products. They were the "platform" team. These were joined together with the hardware guys to make the Tablet group.

Journal came to define the Tablet to many in the Tablet team. They knew they needed to have some killer app on the Tablet from day one, and Journal was going to be it (and it was). But in making Journal, the Tablet team had no interest in building an application that was also good at handling text and keyboards - this was a Tablet app after all, and in order to be the best it could on the Tablet, they had to focus on ink and pen scenarios.

When we started OneNote, the Journal application was up and running in crude form. As I mentioned earlier, OneNote was an Office product, in a different group from the Tablet. Accordingly, we had different priorities - we wanted to be broadly useful to as many people as we could because we wanted to get revenue. That meant supporting text and keyboards, since the Tablet had not even shipped yet and would have very small share when we launched, although we all suspected at least the ability to write on the screen would become a ubiquitous laptop feature eventually. And ink would never replace keyboard - the pen is useful for some things (quiet, unobtrusive, natural), and the keyboard was better for others (fast, compact, accurate input). Of course the Tablet was a key factor for us - we wanted to make sure we made a great application for all PCs, and since the Tablet (with a keyboard) was a superset of other PCs, we would be most powerful on the Tablet, and ideally a “no-brainer” purchase for Tablet owners.

Now here's the rub. How do you design something that is great at both text and ink without compromising either? Optimizing for text was well understood. Optimizing for ink was what Journal was trying to do. But both? That was our struggle. We began with the idea that we didn’t want to have two user interface "modes" - one for text and one for ink. If we did, it might as well be two different applications. All the features of text and ink should work the same way for things that could be made the same, so you didn’t have to learn two different user experiences. With either ink or text you could start entering information anywhere. You could grab things and move them around freeform. You could even do ink outlines and rearrange the parts, collapse hierarchical trees, apply bullets and numbering, etc.

I described below where we got a little off track with ink by focusing too much on structure and outlining and had to scramble to fix it, but we are very happy with how the text support turned out.

So where was Journal in all this? Well, we couldn't simply extend the Journal interface model, since it was quite text-unfriendly. And using Journal produced files that had no structure - just ink on a page. How could we import those meaningfully into OneNote, which had a defined page structure? Well, it turns out we could try, but the result would not be pretty, and like many other useful things it fell by the wayside as we prioritized. After all, there were hundreds of millions of laptop and desktop users, and there would only be a few hundred thousand Journal users at first. And theoretically once you had OneNote you would use it rather than Journal, so note import would be a one-time thing. It pained us, but we couldn't afford to take on an import feature. Before you write to ask, bear in mind that we do want to do this at some point - it is on our list. Just as is making ink as great as it should be, rather than just serviceable.

So, we ended up partly by design and partly by circumstance with two different applications which operate a little differently and don’t interoperate much at all. Welcome to version 1.0.

Update 1/2006: Fortunately, there is now a PowerToy available to import Journal notes into OneNote as ink. Yay!