Interface design is hard and many of the iterative changes in the 2007 Office betas have been improvements, but I am finding this change a real pain in the (Paste random content here) Send button
From the release notes, and indeed the email with the download link...
Things you should know before installing
Pseudo Localized Text Present
Parts of the OS contain Pseudo-localized text, or non-english characters placed together to approximate English words. (Text that looks similar to the following: "?i?d??§ ?èðì? Þläÿ??") This is used internally, is known and being addressed in future builds.
But what I noticed on MapQuest this morning (checking out Leigh on Sea where my mum will probably move to) was what I think of as people-centered data. While the label that comes up when you hover the mouse is Zoom Level 3 the labels at the size of the zoom control show me that's actually the most detailed view I can get of this location as a place within a country, before I go down into region level. For the most detail at street level the icon is a person, for the least level at country view it's mountains (topographic data here I come). The icons get wider from top to bottom - a handy visual cue if I haven't spotted the plus and minus buttons - but it's the labels of Street, City, Region and Country that let me get information the way people think about it, not the way computers do. Like Today/Tomorrow/This Week/Next Week in Outlook 2007 or tags on a blog, it's data aggregated into a fuzzy structure rather than a strictly normalized data slice.
It also reminds me of a set of tech support war stories published by, I think, Compaq, where someone phoned up because the 'foot pedal' on their notebook wasn't very responsive. The foot pedal on my sewing machine gives me acceleration and deceleration as well as on and off. I've been playing Tux Racing on a THinkPad X41 using the accelerometer in the hard drive to detect how I wave the notebook around in mid-air. I love controlling the PlayStation through the EyeToy camera. One the one hand there's the sense of wonder you used to get from controlling a computer at all; on the other, it's a more intimate connection because you don't need to only use your fingers and your eyes. The MS researchers behind this are in the VIBE team (Visualization and Interaction for Business and Entertainment) who do a lot of cool things. I interviewed the Senior Researcher, Mary Czerwinski, a couple of years ago for a piece on how our brains adjust to using two screens side by side (you very quickly tune out the bezel of the screen in the middle and perceive the split screen as one information source).
The StepMail application uses an off-the-shelf "dance pad"to let a user carry out commands in e-mail - such as scroll, open, close, delete, flag and place messages in folders - by tapping a set of six buttons on the floor. Another prototype application, StepPhoto, allows foot-controlled scrolling and sorting through digital photographs.
“Many information workers spend a majority of their time trapped at their desk dealing with e-mail. We wanted to provide them with an alternative,” said Brian Meyers, a member of the Step User Interface Project Group involved in the prototype. “By allowing information workers to stand and continue to read, delete and flag e-mail messages, StepMail gives them a break from the keyboard and mouse, which reduces the risk of repetitive stress injury in their hands and wrists and engages more of their bodies’ muscles.”
'Traditional tablet PC' can't mean passive digitiser (dumb screen that you can touch with a stylus or a fingernail), because those are limited to industrial tablets and PDAs, and very few of them are PCs rather than embedded OS devices (Win CE and embedded Linux for the most part). The OQO and the Nokia 770 are the main exceptions (Windows XP and a reasonably standard Linux) and they're PDA size. You do need a virtual keyboard on most passive digitisers because of the poor smoothness; even when the OS lets you write anywhere on the screen, the curves of your writing aren't what they should be.
But the modern Tablet PC running the Tablet Edition of Windows XP is a bit of a different beast. The active digitiser samples more often than a graphics tablet (though it's the same technology, just a higher sampling rate), so it's very like writing with an ink pen. The only time you'll hunt and peck on a virtual keyboard is for passwords and URLs where 'usually right' isn't good enough. The rest of the time it's up to the application developer whether you write into an input strip - or anywhere on the page.
OneNote, Journal, Art Rage, Grafigo: applications that are designed to work on a touchscreen let you use the screen without needing to go back to a keyboard (virtual or not). Utilities like Sensiva Symbol Commander and ActiveWords let you trigger actions with gestures or individual letters instead of keyboard shortcuts. ritePen is a great handwriting recogniser for desktop and Tablet PCs that lets you write anywhere, even if the application isn't designed that way. A video editing or musical composition app that understands pen input shouldn't need the keyboard input that's designed to be faster than mousing through menus when your input with a pen is both fast and accurate. The next generation of touchscreens won't have the alignment issues that have made it hard to recogise input close to the very edges of the screen (which is why Word's write anywhere option doesn't cover quite all of the screen).
'careful integration'. Definitely important.
'responsive software'. Absolutely.
'eating up screen real estate'. Not when you design and integrate it well. Touchscreens aren't the problem; it's understanding how to use them to replace the keyboard rather than replicating the mouse.
This is another place where I'm impatient for Vista; it will have a cross-hair cursor to make it obvious you’re using the pen, on-screen ripples that let you know you’ve clicked on the screen and eight gestures called flicks that mean you can copy, paste, undo and delete just by flicking the pen in a particular direction. Vista can also learn what your handwriting looks like to make it easier to recognise what you write. If you want to use your finger to touch the screen to select something and it's a combined active and passive digitiser so you can, there will be a little magnifying ring to show you more clearly where you're clicking.
As to the form factors, the first is very like the Fujitsu Siemens Lifebook with touchscreen sitting on my desk at the moment. Folding the screen down to use it without the keyboard is the principle used by any convertible Tablet PC, although I can't see room on the mockup for any hinges ;-) And yes, narrow widescreen is a little off when you rotate it to portrait. This falls in what Ken Delaney at Gartner told me he calls “the 1kg wasteland” because so many products of this size and weight have failed, compared to standard notebooks or smartphones and PDAs. “You don’t have the benefits of the larger devices or the portability of the smaller devices”.
The notebook using a second screen as a configurable keyboard looks huge fun. OLED would do that nicely. It would, as the blog remarks, be pricey. It's price that's held people back from buying devices like Wacom's Cintiq (a touscreen monitor that's a desktop Tablet PC) because everyone one I know who's seen one wants one but not enough to pay that much. And I don't think a screen would be comfy to use for typing - with no key action your fingers get tired very quickly on projection keyboards. We've been using pen and paper for centuries and we're good at it. Making software as good at it shouldn't take that long.
And it's Tools > Customize > Options> Always show full menus to get rid of the irritation.
The site has four competing navigation tools: the buttons across the middle, the buttons up the side, the buttons across the top of the page and the scrolling images of products above them. Except that when you click an item in the scrolling images, while that item does change from wireframe to photo and a pointing finger cursor does appear, you can't actually click to make anything happen. It's Flash to make things pretty, not Flash to make pretty things an interface. And I still don't know what the yellow strips look like.
- Current Mood:ranting
I don't often ask for review copies of O'Reilly books on paper. I write about them and refer to them frequently but I usually read them through Safari, the online library where I can search, browse or read page by page like a normal book. I did ask for a copy of Designing Interfaces: patterns for effective design (Jenifer Tidwell) because I thought it would be a book to pore over. It is.
First thing I noticed; the cover is the usual O'Reilly animal - but in attention grabbing colour. There's a whole section of CSS Zen Garden styles. It's packed with clips of interfaces from applications and the Web. I'm going to sit down and read it properly, but I'm going to recommend it straight away anyway ;-)
Getting the interface right is half the battle (functionality matters too, hence the rant that will be in my next post about the rumoured RIM workaround) and I've been thinking about design styles for supporting navigation habits a lot lately because of the gender design preferences piece I've been researching (now to find a home in .net magazine). Press the user's joy button in the interface, or at the very least don't whack them on the funny bone. At AOL I had to spend a significant proportion of my daily life in a CMS that has what I would nominate as the world's worst interface: eleven tabs with 20+ checkboxes and fields on each, of which a minimum of two needed changing on each tab. Add in a garbage collection mechanism that was so aggressive that it collected database record locks and you have a user who develops strong views on user interface. So I like that here's a book you can give to programmers along with Understanding Comics and say 'read this and then we can argue'.
- Current Mood:busy