One of the largest clients for my Human Interface consultancy (www.humancomputing.com)  this past year has been TiVo, the folks who brought us both the DVR and the ridiculously cute corporate trademark.

I’ve just completed a long contract there, which–among other challenges–saw me helping design their next generation of high-end DVRs and streaming video boxes. This was a terrific project which I was lucky enough to see through end-to-end. On it, I got to do everything from sawing apart old circuit boards and visiting Radio Shack in the middle of the night to assemble testing prototypes… to doing screen design… setting up user studies… and even working with an exceptionally clever engineer to figure out how to get the box installed (and the cable guy back on his way) in a third of the time it used to require.

Best of all, I got to meet and work with some amazingly smart folks both on the user experience team and throughout the company. I may even sign on more adventures there next year, but for now I’m very happy to be back in my robot-strewn office at Human Computing, catching up on correspondence, and looking forward to a terrific Christmas with my family.

In the TV hardware field, manufacturers are beginning to notice that the same 17″-wide components that nestle in your living room’s entertainment center have a harder time fitting in to the back bedroom or study, where many auxiliary screens are kept.

Today, Engadget linked to a new streaming TV product from Sigma Designs which manages to bundle the whole product into a small plug-style box which is commanded via RF Z-Wave networking (so that it can be controlled without the need to hit an IR sensor with your remote control; the radio wave from its remote will find it if you’re anywhere in the general vicinity of the box).

The box itself effectively has almost no interface at all–indeed, the product simply hangs from the bottom of a wall socket with a single HDMI and optical port to connect it to your TV. Networking and communication are handled over the HomePlug powerline networking, so there are no additional networking or video input cables to run. It simply plugs into your wall socket–also neatly handling the problem of where to put/mount the device.

(It would have been interesting to see Sigma design the box to occupy the top outlet of a standard wall socket instead of the bottom one, and configured the video/audio cables to run from the top of the device instead of looping around the bottom. As the TV is almost certain to be located above the device, this might have allowed for a slightly cleaner installation.)

Minimal Styling for a Minimalist Device

It’s clear that Sigma didn’t feel that a device that amounts to an oversized wall wart requires much in the way of consumer design or styling. I’m actually not sure I agree on this point, as the large, flush face of the installed product seems to offer a perfect chance for branding, if nothing else. It’s hard to imagine, for instance, Apple passing up the chance to have yet another glossy, smooth-edged device sporting a tasteful Apple logo as permanent fixture in the home. All that would be required is an change in the plastics and tooling at a minimal difference in cost.

Similarly, I’m not sure sufficient care was taken with the design or placement of the many, minuscule status indicators seen above at right. These are placed so that they are almost certain to be both illegible and unseen when the device is plugged in behind any bureau, desk, or entertainment center. Here, the designers might have been well-served by having fewer, possibly multi-color lights to indicate status. Edge-mounting them so that they are visible from both top and side to account for various types of view obstruction would also help.

Finally, the LED design might do well to stay as low-key as possible, particularly if it’s meant to live in the bedroom or study. A single, low-key light that indicates that the unit is powered up, and which possibly switches color momentarily to indicate remote activity (or red for an error) is probably close to the ideal for such lighting-sensitive environments.

Taking it Further

No price is given in the press release, and it should be noted that this is always a crucial factor in the home consumer market. Having spent much of my past year in the TV space, I also can’t help speculating on what Sigma’s cost basis might be, if the power circuitry (which seems to occupy much of the right side of the board in the picture below) could be handled by a host in an integrated device.

Once you downplay the physical attributes of a device sufficiently: writing off both function (no controls or displays) and style, it seems that the next logical step is to see if you can make the device disappear entirely. Once the physical face of a product has disappeared, the product must then survive by the virtues of its virtual interface. It’ll be interesting to see this area develop.

(HT: Engadget)

At the Mix 11 conference in Las Vegas today, Microsoft took the wraps off an SDK to allow Windows developers access to the the array of sensors and skeleton-tracking technology that make up their popular Kinect sensor. Due out in spring as a free download, it gives Microsoft’s official stamp of approval to some of the homebrew Kinect hacking efforts that have already yielded impressive results in terms of 3-D imaging, robotics control, and other great tricks.

Early demos at the show  included a gesture-controlled version of Worldwide Telescope:

…as well as some Minority Report-style screen navigation.

I’m guessing, however, that the best is yet to come, as thousands of developers crack open the SDK and start mucking around with the astonishing range of control that the Kinect offers. I’ll also be very interested in learning whether it will be possible to integrate the SDK with Microsoft’s own XNA game development system so that a single code-base might work on both PCs and XBox platforms. The possibilities seem immense for making some truly exciting new types of interfaces.

If folks can stand a bit of personal news: The Verizon VCast for Android design project that had me chugging espresso for several months last year finally shipped! (and is now out of non-disclosure).

Much like a movie, the final cut of the product morphed a little from the original design, but the fundamentals seem to have made it to the marketplace intact. Kudos to our visual designer, Mark Castaneda.

Engadget even linked to this rather nice preview (from Droid Life):

http://www.droid-life.com/2011/01/13/verizon-enters-the-text-app-arena-with-v-cast-messages/

All that said, there are some very interesting developments happening as regards Flash, and it is definitely making me rethink, if not our platform development efforts, at least how we might want to prepare for our own efforts in the future.

It started with Apple’s seemingly inexplicable lack of support for Flash on the iPhone (and later the iPad). In practical terms, this meant that our web sites (which use Flash in a minor way) would display blank “Plug-in not available” icons where such Flash content should have gone. It was nothing critical, though, and we simply coded around it for the iPhone and hoped that Apple would resolve whatever issues they had with Adobe in time for the iPad.

Now, however, it seems that the bad blood between Adobe and Apple has turned into a full-on conflict. Starting with some snarky blog posts between the evangelists of both camps, Steve Jobs at Apple kicked the conflict into high gear this week with his “Thoughts on Flash” open letter in which he took several pages to lay out six reasons why Apple and Flash have parted company, seemingly for good.

This in turn came on the heels of Apple’s announcement that apps made with cross-compilers (like the one in Adobe’s new CS5 suite) would not be accepted in the iPhone/iPad App Store. This effectively killed Adobe’s promising effort to to easily allow Flash developers to package their applications into iPhone apps the day it was born.

Adobe fired back in an interview with the Wall Street Journal, and later with a kind of retaliation by immediately ceasing development of their Mac Flash products, making Flash on Mac OS X an also-ran to Windows in the desktop world.

Interestingly, Microsoft also piled on Adobe, albeit subtly, in stating not so much their antipathy for Adobe as their belief that “open standards” (HTML5) are the future of the web, and that Internet Explorer 9 will only have built-in video support for H.264 video. Although it seems unlikely that a release of IE9 will appear without Flash video support in some manner, it can’t have been a great week over at Adobe.

Android, for its part, gleefully embraces both technologies.

Regardless of the technical merits of the discussion, what it really means is that Flash isn’t coming to products like the iPhone or iPad ever unless something very dramatic changes in the business of either Apple or Adobe.

While I don’t really have a dog in this race, I–like most web developers–want be able to make sure the stuff we create can be viewed well on as many devices as possible. Despite Apple’s claims to the contrary, HTML5 isn’t an easy replacement for many of the sort of interfaces we’ve seen in Flash. At a minimum, they’d need to be completely rethought and re-coded. More likely, completely alternative designs–often less…well, “flashy” ones–would have to be developed in their place.

So yes, although we don’t use a lot of Flash in our daily life, it’s looking like we’ll be spending some efforts in the days ahead to see what the alternatives might be for the sort of work we currently use Flash for. It looks like this is one argument the squabbling kids aren’t going to be sorting out any time soon. They’re effectively taking their toys and moving to separate playgrounds. Those of us who used to play with these guys will either have to pick a side, or spend a lot of extra work to make sure that we can operate well in both camps.

Welcome to the Flash Wars.



From designer Gonglue Jiang, an award-winning concept for fitting infinite USB ports in a tight space:

http://jianggonglue.com/DesignUSB.html

Oh, how I want to like this design. It’s elegant. It’s simple. It’s sort of pretty with the color coding…but I’m just not sure it works in the real world.

Here are the major difficulties as I see them:

1. When used with more than a couple of devices, power becomes an issue (although there’s another cable that can be hooked in, apparently, to address that).
2. It also sticks out and become a long torque arm after a couple of devices are attached, risking bending and obstruction (I’ve now wrecked two 3G cards this way).
3. And then there’s the whole, “it requires the world to issue new device cords” part.It’s hard to see a device vendor wanting to pony up the additional cost required to make the change to let other devices play nicely with their device. Your hardwired mouse, graphic tablet, or camera, keyboard, bar code scanner, etc. all would need to fight for the end position, since none come with detachable cables. In those situations, implementing such a solution would require you to do it through external extension cables in order to get any benefit. At that rate, you may as well just pack a USB hub and call it a day…

…or maybe I’m wrong on this. What do you think?

Why are we still putting the menus at the top and leaving users with less space for editing instead of setting those menus at the left?

I have been wondering about this exact topic. With all this widescreen revolution, screens featuring more horizontal real estate than vertical real state. Why are we still putting the menus at the top and leaving users with less space for editing instead of setting those menus at the left? Does this have to do with users used to this specific setting with menus on top or is there really a real scientific logic behind it?

Here’s my response:

Two good reasons come to mind. The first is practical: convention and user expectation demands it. This isn’t the end of all considerations, but it means that anything that defies such convention needs to be not just better, but A LOT better in order to justify making the change.

The second is physical and a bit more subtle. It relates to Fitts’s Law (which boils down to the idea that big, close targets are easier to hit than small or far away targets). Since menus are “bound” to the edge of the screen which they border, they are essentially infinite in that direction–you can slam the pointing device as far in that direction as you like and get the same point.

However, menus on the left and right sides are only as tall as their line height, while menus at the top and bottom as wide as the text width–usually several times the effective target size. This leads them to be far easier to “acquire” (hit) than their horizontal brethren.

-Pete Bickford
(for uxtalk.wordpress.com, and a HT to Bruce Tognazzini)

Although the graphics are certainly more pleasant than the IDEs I’m used to, I find myself constantly lost as I try to establish my context within the overall program, and the sense of watching a program’s flow through this metaphor is unusual to say the least. Ultimately, I find I’m more drawn to the team’s other ideas (such as code factoring via a simple direct manipulation gesture) than I am to their overall bubble metaphor.

For more, see the team’s site at: http://www.cs.brown.edu/people/acb/codebubbles_site.htm

What do you think?

Julia Turner at Slate has an interesting article on how the beautifully designed “Man leaving via exit door” pictogram hasn’t caught on with Americans who are used to our big red Exit signs.

http://www.slate.com/id/2246107/

This is a great issue where the expectations of the installed base (a.k.a. the entire American population) can make unworkable a potentially more attractive or functional interface that doesn’t meet those expectations. C.f. Dvorak vs. Qwerty.

The article also includes a wonderful example of the Slate offices, sporting two different versions of exit signs (the one on the right apparently was placed as part of a governmental edict). Note the Band-Aid approach of labelling the sign “Exit” in addition to the pictogram (icon), as well as the need to place the sign smack dab in the middle of the door. Even then, it’s likely an inferior design for the vast majority of staffers since everything from its color to its icon violates the cultural expectations of exit signs.

It’s interesting to note that “Way Out” signs (The British equivalent) are also frequently labelled in “caution” colors. Despite the article’s implication,  green (“affirmative”) and soothing is often not the color scheme that you should use when you need to direct the user’s attention to something important.

Traditionally, the gating factor on how small mobile phone buttons could be was set by the width of the human fingertip–around 2 cm (3/4 of an inch). Technically, half that width serves as a workable target in most cases (10 mm), since it’s possible to compute the intended target, even if there’s overlap of adjacent areas, although this depends a great deal on the type of switches or touch-screen technology used.

You can also buy yourself a bit more room by adding vertical and lateral “pitch” (curvature) to the keys of a physical device. This effectively gives you a larger surface area while staying within the same overall length and width of the button. Keyboards with strong vertical pitch also give you a tactile targeting point, further reducing errors (and the lack of both on the Droid’s slide-out keyboard explains why it’s such an error-prone design, despite the relatively large width of the keys themselves).

With proper pitch, physical buttons as narrow as 9 mm (0.35 inches)  can prove surprisingly functional, but for all of this, the minimum width of a standard 12 or 15 key phone keyboard still winds up being 4-5 cm wide (1.6 to 2 inches) and 5-6 cm tall (2 to 2.4 inches).

But what if you want to go even smaller?

Enter an interesting October, 2007 patent from DoCoMo, US Patent 7280101 – Fingertip tactile-sense input device and personal digital assistant using it.

Using a grid of tactile dots much as you would a tiny track pad, it claims to offer a workable way of making the entire control surface of a phone about 22 mm high by 20 mm wide. (.78 inches by .86 inches) — in other words, smaller than a US Quarter.

…and you thought losing your cell phone was a problem now…