Wednesday, March 11, 2009

The Argument for Client Applications & Specialized Hardware

Miguel came up with some interesting news:

Wireless Generation announced today (03/05/2009) that their new handheld of choice--although they would continue to support the Palm and encourage purchasing of the Palm from two resellers--for their products would include the Nokia N810.

Miguel then muses over various facets of the decision, particularly whether or not in 2009 people would prefer the option of a web app running on a netbook.

Everyone is getting so web centric it is worth pointing out why a client app on a pen-based handheld makes sense. Teachers implementing a literacy program based on frequent formative assessment need to do a lot of formative assessment, in class, in the least disruptive and most efficient way possible. This requires a specialized, highly responsive interface. When a kid stumbles over a word, you have to note that while he or she keeps going. You can't stop to tab to the right place, wait for a page reload, wait until the network is back up, etc. You want to tap the word with your little stylus and move on. And you want to be carrying something you can carry around all morning, sit with at tiny tables and chairs, stick in your pocket, etc.

So no, a netbook running a web app is not as good, and Wireless Generation is trying to provide a high quality, vertically integrated service, not the cheapest possible crap. And the fact that they stuck with Palm OS this long, which has been obviously doomed for what, the entire decade, demonstrates that they're sensitive to not pushing schools onto a new platform just for the heck of it.

If there isn't sufficient funding after this giant stimulus, a nice chunk of which is going into educational data systems, to give teachers the best tool for gathering data about student progress in elementary literacy... I don't even know what to say. It will truly demonstrate how much we're living in a Potemkin village wrapped in a fraud, surrounded by a cloud of lies. UPS doesn't ask its delivery men to juggle Eee PC's with USB barcode scanners dangling off them. Walmart doesn't try and save money by putting a shoebox under a laptop and calling it a cash register.

Also, Miguel overplays the connection to the Nokia 810 hardware. The software runs on the open source Maemo platform, which is made up of most of the same components as the GNOME desktop (making it a cousin to Sugar). Making a port that will run on a Linux netbook shouldn't be hard (although Ubuntu Netbook Remix has drifted in a somewhat different direction because Maemo is optimized for pen input). Also, if appropriate Android devices come on line, they'll probably come up with a port. Maemo is not a slam-dunk choice right now, but all the current options have serious drawbacks.


Anonymous said...

Tom, I think you make some very valid points about the use of an application based assessment over a web based one. However, it still remains that Wireless Gen is tying itself to a specific platform that is not even that widely distributed. They could have been a leader in education and developed an app for the iPhone/iPod touch that would work in the same way as their palm app, and be even less obtrusive because the teacher would not have to hold a stylus...they would have one hand free at all times. While the Nokia has much better features than the iPod/iPhone, Apple has a much larger marketing department, and I think they will win out as the handheld device of choice in the very near future. I'm still questioning why they went with the platform they did, and we, as a district will be taking a hard look at whether we want to continue using their product.

Tom Hoffman said...

First, I don't really believe a teacher could do something as complex a score a literacy assessment with one hand -- using their thumb?

Second, I don't think the answer is picking the winning proprietary hardware/software combo. That's what made Palm a complete dead end eventually. Maemo is an open platform, made out of common and widely used open source components. It is highly adaptable and independent of any one vendor.

Miguel Guhlin (@mGuhlin) said...

Tom, you make some fascinating points, in particular, the UPS device. However, a UPS guy doesn't also have to do grades, track attendance, respond to email, keep notes, access a curriculum management system, etc....a teacher does. Why should districts buy a device for every purpose when they can buy one that simplifies things?

It's up to the vendor to make an assessment interface that meets District needs rather than have the District go bankrupt trying to buy and manage and support the device the vendor has decided will work best for their product.

Thanks for offering a reasoned critique!

Wishing you well,

Bill Kerr said...

I didn't quite get this at first due to being an old timer and not realising how far 3G has progressed

Doing some catch up now, the exchange b/w miguel and you has turned out to be very useful, at least for understanding (but implementation in school environments is still a long way away)

Vendors in USA may be further down the track than in aus about the potential in schools - judging from teacher aware comment from Wireless Generation at Miguel's blog

Bill Kerr said...

I ran this past joel and he pointed out:

"New browsers have local storage and database API's built in (HTML5, and Google Gears is a "backport" of these features to older browsers), which can be used to write web apps that are available 100% offline. They can sync up to the network when connectivity is available.
- more here "

The comparison with UPS delivery men and Walmart cash registers needs more context or fleshing out. I'll argue this from the POV of a maths / science teacher. There is a positive role for calculators and graphics calculators but there is also a place for general purpose software which becomes very powerful bit by bit as the student learns more and more about it. Andy DiSessa (Boxer) and Alan Kay (Squeak) have argued this from the margins for decades.

In contrast to the Wireless Generation approach on Miguel's blog , how about general purpose software with Group view as implemented on the xo?