Though there are lots of other reasons why I switched to a Mac (the ease of web development being chief among them), but the biggest difference in my mind between Windows and Macs is Cupertino’s love of gesture-based interfaces (which I share). My last Windows machine was a small 8″ tablet running tablet XP–the form factor was perfect for me, but the touch as an interface was way behind the mouse. I build quite a few little interface prototypes in Processing at the time, but there really wasn’t a way to take a small little self-contained java demo and push its conventions onto the OS as a whole (I’m certainly not an OS programmer).

Between the built-in gestures in OS X, the iPhone, multitouch trackpads, and the new multitouch mouse, Apple is kicking Microsoft’s ass on the gesture front. Surface and the Courier are promising, but neither of them are exactly nearing the consumer market at this point.  Why not build a touch controller for the XBox? Something small (maybe, uh, Zune sized?) which could act as a secondary display for inventory or controls. There’s nothing stopping them from doing this, and assuming it was available to XNA developers this would instantly get me interested in building more games for the platform (the Zune requirement would mean even less people would buy them, but heck–no one’s buying indie games anyway).

There are rumors swirling that the new Apple tablet will have somewhat of a learning curve, so I’m hoping it’s some kind of new gestural interface. Since it’s not being unveiled for another couple of days, I thought I might as well fantasize a little about my ultimate tablet device:

• An 8″ convertible multi-touch screen with a physical keyboard (I know the Apple won’t have one, but I think an 8″ keyboard is about the smallest still-functional keyboard–and it blows away any virtual keyboard I’ve ever used)
• In lieu of a physical keyboard, a way to dock the thing to a physical keyboard for extended typing.
• An IR emitter with a rich interface for controlling the TV (and a cloud-based Tivo would be nice, too)
• I’d love to be able to just “fling” content from a tablet PC onto a desktop when in blue tooth proximity. Just grab the file, do a little fling gesture, and the file magically lands on the other computer’s desktop. No cords needed.
• iPhone tethering for internet access on the go–or just toss in 3G to the device itself
• The same compass/accelerometer technology currently used in the iPhone.
• A system for slaving the device to a full computer for use as a tablet-based input device. I can’t count how many times I wished I could plug my tablet XP machine into my full desktop running Photoshop to do a quick sketch. Actually, this is dreaming small–I want any piece of hardware to be able to take control of the thing and use it however it wants. Alarm clock dock? Sure. X-Ray machine? Sure. Car dashboard? Sure.

Fingers crossed for Wednesday.

(Image lost in the Great Update of 2009)

I put in my email address and clicked continue. Nothing happend–a message to “Please fill out the entire form.” After trying a few times (all unsuccessfully), I finally said “screw Genius” and went back to my normal playlist.

Fast forward 10 hours or so. Now I’m home, and again listening to some music. Thinking it may have just been some weird glitch on my iMac at work, I tried to activate Genius again. Same problem. I think, “Fine, I’ll try my AOL login.” I switch the toggle to the second option and fill in the second field with my AOL username. Another error. Only at this point do I look at it more closely… the second field is the password field!

There’s a good chance I’m just retarded, but I feel like I’m pretty tech-savvy. If I made this mistake, I’m sure others have as well. The issue is that the two radio buttons line up perfectly with the two text fields. If I could boil everything I’ve learned about usability into one single mantra (even if I don’t always follow it myself), it would be: “PEOPLE DON’T READ.” I’m no exception. Because there was a 1:1 correspondence between the two radio buttons (and more importantly, the two logos), my brain made the (incorrect) leap that the first field was for Apple ID and the second field was for your AOL username. I’m no stranger to multi-screen login/verification forms (especially for things with higher security, like bank accounts or apple accounts with purchase power built in), so it didn’t seem weird that it would ask for my username on one screen and the password on another.

Once I actually read the form and understood what it wanted me to do, it all worked great. What I’m wondering is, though, am I just dumb? Would others actually make this same mistake?

I was playing pool last night for a few hours in a pretty large pool hall over in Santa Clara.  There was a Juke Box there, one of those modern numbers that’s all-digital instead of shuffling discs around (as a side-note $0.50 a song on the Jukebox makes iTunes look a lot more reasonable…).  There were two payment options on the thing–either insert 1/5/10/20 dollar bills, or log in to your “whatever the name of this jukebox is” Account (I’m assuming it was hooked up to the internet).  This seems like such a wasted opportunity to me.  While I’m sure the jukebox makers would just LOVE to have some demographic data and email addresses to try to sell things to, who in their right mind is going to sit there and sign up for a profile on a jukebox interface?  More importantly, who even wants a jukebox account?

The idea is solid–the fact that not all people carry cash with them, and being able to pick a few songs even if you don’t have any ones handy is nice.  But the hassle of registering for anything is just too great to make it a worthwhile proposition.  Enter the cell phone.

I guarantee you that 99.9% of the people at that pool hall had cell phones on them.  When I was at Georgia Tech, the mobile guys had this one interactive display setup that featured fish floating around in an aquarium.  Each fish had its own phone number, and you could dial in on a cell phone to take control of it–moving around with touch-tone presses as a joystick.  THAT would actually take some programming chops to set up.  You also see things on TV all the time saying “text in your vote for $0.99 to ######.”  If the jukebox operators really want to make some money, they need to get in bed with the cell phone carriers and figure out a way to text in how many song credits you want to buy.

Let’s see… scrounge around for a few one-dollar bills or text “5″ to the phone number written on top of the jukebox?  I know which one I’d rather do.  The sad thing is that people are spending millions and millions of dollars trying to figure out how to make cell phones do more, and no one seems to be focusing on what they already do well–call things.

The sign-up process is a major chore.  I’ve never done it at another broker, so I don’t know if it’s this bad everywhere or this is a “you get what you pay for” scenario.  I actually filled out all the paperwork back in August, but I never got around to mailing it off and faxing them my info (yes, MAIL and FAX–I didn’t realize anyone still used either).  I got a little more motivated last week, so I checked back in with them.  Because I hadn’t mailed my paperwork in by their 30 day deadline, I had to fill everything out again from scratch–there was no record anywhere of the application I’d previously filled out.

Once I got everything filled out, I headed over to Fedex/Kinkos, figuring I could make copies of my driver’s license, fax them the sheets they required, and mail everything in one stop.  No such luck.  Their mailing address for paperwork is a P.O. Box, so after standing in line for half an hour (December, go figure), they told me I’d have to take it to the post office.  After standing in line again at the post office, I finally had everything sent off properly.

I got the email saying the account had been a few days later–and that I could go to the site and log on using a combination of random personal info and the last 4 digits of my account number.  I tried to do so, but I didn’t actually have an account number yet.  When logging on to their site, it just said “Roth IRA–Account set up in minutes.”  It said that for a few days.  After it finally went through, I was able to log in and take a look at their trading platform.  In a word: ugly.  It looks like it was designed by programmers, with no real care paid to user experience.  I’ll hold further comment on it until I’ve actually used it some (though if the ACH-setup interface is any indication, it’s going to be a major chore).

Then my next issue: management fees.  Somewhere in the paperwork I filled out, I remember reading  that they charge you $30 a year to manage your retirement account (though, as far as I can tell, it’s still up to me to make all my own trades and pay trading fees of $4.50 until my balance hits $2500).  The yearly fee hits in February, and if you fund your account after February they charge you immediately.  It’s now December.  I called their customer support (and, to their credit, actually got a real person within 5 minutes) and asked if this fee was pro-rated or I was really going to be charged $30 so they could manage the couple of hundred or so dollars I’m likely to put in before February.  Their response: full $30.  I asked them if it was a specific date in February (I’d hate to put money in Feb 1st only to get charged again Feb 10th) or if I could safely put money into the account any time in the month.  He didn’t know.  He said it was usually a certain week in February, so if I wanted to be absolutely sure it was best to wait until March 1st.
So I’ve got a Roth IRA now.  I’m just not going to put any money into it until March. That’s not so bad, I guess.  I’ll just have to do a good job of mentally earmarking money in my checking account as “IRA money” and not “shiny new toy” money.  The game I mentioned in my Thanksgiving post got a really great sponsorship offer, so once I polish it up to their specs and finalize the deal I should have a pretty nice chunk of change to act as seed money (though I still don’t see myself maxing it out for 2007).

Once we got there, though, we found that we could’ve started working on it weeks ahead of time. Whoops. Undeterred, we stayed up all night programming Bike vs Car. This was an idea that I had a couple of weeks ago (after riding my bike for a couple of weeks). I’d like to ride even more than I have been, but motivation has been a little lacking (some days). When I do ride, though, I tend to hop on Facebook to update my status to something like: “I just biked to work!”

I figured I might not be the only one to do that sort of thing (and since FB is still mostly colege students, biking to class is certainly an option), so it might make a great application. I went ahead and registered the bikevscar domain and then just put it on the list of things I wanted to do. The BarCamp seemed like the perfect excuse.

After we got back from the speakers, we spent a few hours reading up on the Facebook API. Eventually we decided to create our app as a standalone, get it working decently, and then integrate with facebook afterwards. I talked my friend into trying out Ruby on Rails, so I spent a little while configuring the server while she did a couple of tutorials. At around 1 a.m, we actually started programming. I did most of the database and backend stuff while she did most of the HTML and CSS (thank goodness… I hate CSS). We got a reasonable facsimile of a social network put together by presentation time (2:00 p.m.), but didn’t actually get any of the facebook integration done (other than loading a splash screen).

The judging was completely crowd-based, and I think we did a pretty good job of selling not just we’d completed–but where the app could go. We won 1st place. The prize was basically a lot of Java merchandise, a bunch of free programming books, and a few other cool things. The fact that we were able to sell a bunch of people on the app tells me it might have some potential if we ever finish it.

Basically, Bike vs Car is a personal tracking application. Rather than focus on other users, our target user is someone who wants to motivate themselves to ride their bike more often. This isn’t necessarily targeted to the spandex-clad guys on $2000 bikes you see riding to work (motivation doesn’t seem to be a problem with them). This will be a tool for the rest of us. There are several major components to the application:

USERS. We allow users to upload a picture, but we don’t really get any info other than their zip code (which we can then use to calculate average gas prices). Because we’re targeting the site for Facebook, we can just graft our functionality into their already robust inter-user network… and we get to focus on just the functionality we want.

VEHICLES. Though there are many types of vehicles, we’re only supporting the two namesakes for now: bikes and cars. Users can upload a picture of their vehicle, list the make and model, give it a name, and (if applicable) enter what kind of gas mileage they get.

ROUTES. Google maps recently added a great feature. You can get directions from one place to another, alter the route as you wish, and then hit “Save this Route.” Rather than just giving you a link (which has been around forever), they now offer you several different sizes of pre-configured embed code to place the map on your own site. So, rather than scrape Google maps ourselves, we just allow the users to post in the embed code for their route, then give the route a name/description and list the mileage.

Users, Vehicles, and Routes. These are the three essential “nouns” for Bike vs Car. The majority of interaction, though, will come from keeping track of which vehicles you take on which routes. To be fair, not every trip you make is a good candidate for a bike ride. We’re encouraging people to only keep track of trips they make that are ~3-5 miles (or less), with only the driver in the car, and where it’s not going to make much of a difference if they show up a little sweaty.

Once you make one of these trips (on a bike or in a car), the users would then log in to our site (or Facebook, which most of them do already). They would be presented with a list of their most common routes (Work, School, Grocery Store) with bike and car icons next to each. Clicking on one will add a tally-mark for that week. The goal (which will likely be rewarded with some sort of token) will be to choose the bike more often than the car (>50%) for a given week. Based on the MPG of your car, the length of your routes, and the area where you live, we’ll be able to generate an estimate of both how much money you’ve saved by biking and how much carbon you’ve spared the environment.

This is something I would find valuable as a standalone application, but making it a Facebook application is even better. Not only would a user be able to track their own bike usage, they’d be able to track how their group of friends are doing. Or how their network is doing. Or how the entire facebook community is doing. Furthermore, instead of logging in to Facebook to update my status, Bike vs Car could automagically create feed events whenever I bike to work (or let me enter an excuse for why I didn’t).

After spending almost 12 hours straight on it, we’ve got a pretty decent prototype in place. The 1st finish place sort of vets the fact people might go for this, so now it’s just a matter of finding some more time to work on it.

Using some more basic YUI components (I actually like their simple pieces a lot better than the complex components, as skinning them is a bit of an undertaking if you don’t want them to look YUI-like), I whipped up the two easiest examples of this I could think of: a 2×2 pane and a 4×1 pane: