Archive for November, 2004

PC Tablets

Sunday, November 7th, 2004

My company provides software to pediatricians to improve pediatric care. The doctor) use our software during patient visits on PC tablets which have pens to capture text input. T-‘incurrent (y i up.s,.,,….of a tablet that F am testing so I can have experience with take tablets themselves.

I am writing this entire post using the pen input device. As you can see, the first part of the message was quite successful but things went awry when I tried to say “I’m currently in possession…”. I thi nkt hat’ she cause I was trying top u-i too much in the little handwritin, input box; plus, I have awful handwriting, so the tablet should actually be applauded for being able to read my chicken-sc rate h in the first place. (I’ve decided not to correct any of the mistakes made by the recognition system so you can see how well the system works, and where it fails).

So, my analysis is that the handwriting recognition on the tablet is pretty darned good. However, there’s a lot more than hander writing recognition necessary to make a pen-based interface work. Sometimes, the white space needs to be tweaked. When that happens, the pen-based interface breaks down. Also, if you screw-up the strokes of a letter and change your for mind, then there’s no recourse other than resorting to the keyboard or using a small number of buttons to the right (armed to return to the mistake and correct it using “bksp” and “del” buttons that you can tap with the pen–it acts just as a mouse in that context.

I’ve usedadi iferen.itablet in the past (this one is an HP, and I’ve also used a Toshiba) which had a be double -ended, Jen. The other end let you erase your strokes, which helped quite a bit. This one does not seem to have such a feature, which I miss.

Simply entering tee* seems to work relatively well, but when there’s something to be done that’s not handwriting -related,” the process gets bogged-down with exception cases.

Another problem that I see with the handwriting interface is entering passwords. Since are deal with medical data, which is very sensitive, most of our applications have password authentication. Entering a password using the pen can be done in two ways: by writing on the screen and using the handwriting. recognition system, or by switching to an oh-screen keyboard, where goa can enter your password, hunt-and-peck style, into the password field. Both choices make it painfully easy for someone to serrepticiorsly view the password being entered-i either by simply reading it off the screen, or by watching you type-in the password trey-by-key. Both of these options pretty much suck.

All in all, the technology in use here is pretty sexy. I’m hoping th a-ii in-‘.’me, th chard w.: l, any System will get better and be able to understand any chicken-scratch I can throw at it. -l-also hope that the non-handwriting stuff gets better, as well as’ the ways that sensitive information is entered (such as passwords).

Wow. The HW recognition system finally gut a pair of parenthesis right. Things are already getting better!

Super Ouch!

Thursday, November 4th, 2004

My very good friend Dave Edwards lives in Salt Lake City, Utah and continually makes us all jealous that he’s got season passes to Snowbird and I think a couple of other places in the immediate area. He can basically finish up classes for the day (at U of Utah) and run up the hill for a few runs.

Whils planning this year’s week-long ski trip, we received a message from him on the mailing list, which includes this excerpt. I’ve in-lined the image thumbnail to make for a better blog entry.

Two weekends ago, I decided to try something called “sand skiing,” which is exactly what it sounds like: you find a large sand dune, hike to the top, and ski down on some old, beat-up skis that you don’t really care about. Needless to say, it’s a fun way to spend an afternoon.

But if you’re going to try sand skiing, here’s some advice:

  1. As you ski, the sand really tends to come up over the tops of the skis and bury them, which is a problem since sand is so much heavier than snow. If possible, I’d recommend a set of fat, powder-type skis; if you can afford to ruin a pair without straight edges, so much the better.
  2. You will want to set your bindings as low as possible so that they release easily. If your bindings pop out when you look at them wrong, so much the better. Heavy sand on your skis can generate a LOT of torque quickly, and it’s better for your legs not to absorb that if you can avoid it. Also, the sand tends to work its way into the bindings and jam them up, so it’s better to be jammed with lower release settings than higher.
Nice tib-fib fracture!

Unfortunately, this was my first attempt at sand skiing, and I didn’t know
these guidelines. I had my bindings set just slightly lower than I would
for snow skiing, which turned out to be a bad idea.

So, to make a long story only slightly less long, I’m in a hard splint for two more weeks,
followed by a cast for six to eight weeks, followed by rehab. There’s a
chance I might be able to ski again by the end of the season: we’ll just
have to wait and see.

In the meantime, the mountains continue to be buried by snow, as I hobble
around my apartment emitting primal screams :)

I nearly cried when I saw that picture. I mean, as soon as I read the words I decided to try something called “sand skiing”, I knew where the story was going. I wasn’t exactly prepared for that image, though. He pretty much cracked the bone in half, and then right down the middle — just for good measure.

Get well, soon, Dave.