PRISM — contextual gesture shortcuts

Research demonstrator from an HCI course project: contextual direction-gesture shortcuts with live feedforward guidance.

PRISM is a research demonstrator from a Human–Computer Interaction course project at Stony Brook University. It puts one interaction technique on real web pages so that it can be tried, shown and discussed. It is a prototype, not a polished or supported product, and it is published in that spirit. WHAT IT DOES Hold the right mouse button and drag. Instead of the usual context menu, PRISM reads what is under your cursor — a page, a link, an image, or text you have selected — and offers only the commands that make sense there. Flick in a direction and the command runs. If you hold still for a quarter of a second before moving, a guide blooms around where you pressed and shows the directions available and what each one does. Start moving and the guide prunes itself down to the options still reachable from the direction you have taken. The guide is there to be read while you are learning and ignored once you are not; you never have to wait for it. Right-click still works. A right-click that moves 10 pixels or less always opens the website's own context menu, exactly as that site intended. The gesture only takes over once you have clearly started dragging. THE COMMAND SET 16 commands across four contexts: Page scroll to bottom, scroll to top, back, forward, new tab, close tab Link open in a background tab, open in a new window, copy link, bookmark link Image save image, open image in a new tab, copy image Selection define, search this, copy You can also record your own gesture on the options page and bind it to any command in that context's list. Custom gestures are checked against the existing ones for collisions, are stored locally, and survive a browser restart. HONEST LIMITS - This is a course research prototype. Expect rough edges, and expect it to stop being updated when the project ends. - The technique needs a mouse or trackpad with a right button and a drag. It is not usable with a keyboard alone, and it is not designed for touch. - No study results are claimed. The evaluation this prototype was built for has not been run. PRIVACY No analytics, no telemetry, no server, no account, no tracking, and no data collection of any kind. The only thing it stores is the custom gestures you record yourself, in local browser storage on your machine. Three permissions are requested: "bookmarks", spent by the single call that the bookmark-link gesture makes; "storage", for those custom gestures; and "downloads", so the save-image gesture saves the picture to your Downloads folder. Host access to http and https sites is used for one thing only: when you draw copy image, the extension re-fetches that one picture (without cookies) so the picture itself, not just its address, can go on your clipboard. The full policy, including what the copy commands put on your clipboard and which two commands open a third-party website at your request, is linked from this listing. The privacy policy is public, and the installed code is readable: Chrome extensions ship their JavaScript to your computer, so you can inspect exactly what PRISM runs.