Zooming

Martin Stein Hi Nathan,

Thanks for the opportunity.

Canvas is fantastic stuff. I want to thank you for making my life easier. I was about to give up on OneNote because of the limitations in manipulating the "sections" as a information tree. (There was no easy View in the GUI to deal with big chunks of the information tree at once.) This is a brilliant fix.

With this one, you are poised to leapfrog ahead and make ON a more desirable app than you could by sticking with a traditional tabbed interface.

Am I correct that some of this object manipulation was inspired by StickySorter? And BumpTop (presented at TED) was the project that came up with expanding and shrinking icons according to user importance. Correct?

Now, some suggestions. There are some immediate functions of interest, and then there is where Canvas is pointing more broadly.

1) MrGroove (commenting in the OfficeLabs blog) is correct. It's important that Canvas be more thoroughly synced with ON's tabs. Any groupings or renamings done in Canvas must seamlessly update existing notebooks. Currently, it seems it doesn't.

2) It would be helpful if Canvas could JOIN different notes together -- "smoosh" them together, so to speak. At a bare minimum, if you can edit a page, it should be possible to cut from one and paste to another.

More broadly, and more boldly, here is where the rubber hits the road with your Canvas project:

3) Canvas is a prototype for overhauling the traditional Windows Explorer icon-parking lot approach to file management.

I gather that's what Surface Computing is about. I haven't seen the OfficeLabs roadmap video yet, but plan to. The following metaphor might clear it up: Rather than doing file management through unweidly layers of "sliding window frames" the Canvas is a step toward surface computing and the GUI equivalent of an "infinity pool."

More specifically, whether it was designed with this consciously in mind or not, it's exactly the right step to take to move beyond where we are today, with static icons and opening/closing documents as discrete app tasks.

Essentially, with a Canvas, you're bypassing a desktop and rendering the app -- and the OS -- totally invisible. OneNote is a great platform for this, as you probably knew going into this, not just because it's good at OCR and handling images, but because each file is a pastiche of smaller documents / files. With Canvas clearing the OS out of the way in the visual field, a user is automatically interacting directly with file *contents*, rather than with file "formats" and"icons." Certainly, the "transparency" features in Win 7 (when windows are called up and made to vaporize as previews) hints at bigger things to come. Canvas is a peek at a very different OS. At first, I assumed that Vista and Win 7's Aero Glass was just a gimmick, but now I can see more clearly where MS' research into transparencies is headed....

What's best is being released from manhandling Explorer frames. In other words, you can extend this beyond OneNote Notebook groups to think of this as an Explorer interface. It's the perfect tool for calling up multiple folder content groups onto a single canvas. (Laying all the contents out alongside each other.)

Indeed, all we need is one Canvas. Just give it the ability to right-click to open another folder's contents somewhere else on the Canvas in order to have a much less confining replacement for windowing. All the file management and dialogue should be done not as matte boxes, but through the equivalent of a fighter-pilot's heads-up display. Where the user clicks on the surface, a colored font can be superimposed on the Canvas / image surface with instructions. With a click on an image surface, a user can swiftly call up files by tag searches, rather than by directories. The search form just opens up as an unobtrusive gap in the image surface, The search results in that superimposed colored font will also be clickable.

4) A further element of this GUI, which could be implemented even more easily right now, is a "mind mapping" web surface for letting the user organize data the way he/she prefers. One app to look at quickly is Personal Brain. This is the first app I've seen that does something wonderful with the file structure -- rather than require that the user navigate directories, the user decides how he sees files or even just ideas as related, and tethers his own files to nodes. Those file tethersmove in any direction, like shortcuts and hyperlinks. (In a glass GUI, the tethers needn't always be visible clutter, but can be invisible for clarity, until you click on a piece of data.)

The other advantage of a mind mapping approach is that contexts needn't be limited by a file hierarchy, but can be multi-modal -- the display of nodes will vaporize and reform, and call up new data links, in any number of different ways, depending on which of many "current project" the user selects from the starting node.

The desktop, then, is a mutable space. It is no more than an "overlay" for manipulating all the relevant semantic objects needed to continue work on a particular projects -- and it calls up at any time only those objects that are needed or most relevant. Each project would be stored as metadata, so that it would take no more than a macro keyword to unfurl one of any number of overlays for less cluttered desktop manipulation. (In other words, overlays would be embeddable as hyperlinks within kindred overlays. A mind map is only a fluid, hi-def wiki of sorts after all.)

5) It's also worth broaching a way that drag-n-drop for folders could be handled on such an Explorer overlay: Start with a top-level home folder on one edge of the canvas, and when an object is hovered over it, it expands -- not as a window or a linear folder tree, but rather as a set of radial spokes to the subfolders. Then when the user moves the object onto that folder, BOTH the folder and the user's pointer (with the object it's dragging) can reposition again to the Canvas edge, as the new hub for the subsequent sublayer of spokes and nodes. By constantly repositioning the center of the spokes to the edge of the screen, this hover interface would completely avoid the problem that plagues complex menus, with deep submenus capable of blotting out the screen's entire real estate.

Those were my immediate thoughts on seeing the surface you've designed with Canvas. It's brilliant. I wish you luck with this, especially with becoming plugged into a larger project to remake the OS' GUI, not just the GUI for the individual office apps, I found myself suddenly a huge fan of the new "vista" you've suddenly opened up on desktops and windowing.

It's late, and that's all I have for you at the moment, Nathan. Take care, and best wishes.

Cheers,

Martin

Martin Stein Nathan (who knows if you're checking this),

Writing to congratulate you on Canvas. Very fluid, and more important, brilliant at fixing OneNote. I hope this GUI is a preview of even bigger things to come. Great work.

Can I send you a PM with some ideas on where you could take this interface? I felt too nerdy replying with a "concept" on the OfficeLabs blog. The forum for OL is hard to find as well.

Cheers,

Martin Stein (alsoknownas at live dot youknowwhat)

Nathan Fish replied to a message RE: Canvas for OneNote.

Nathan Fish posted a new message Canvas for OneNote.

Nathan Fish replied to a message RE: Introduce yourself.

Nathan Fish joined our group!