“Some people, when confronted with a problem, think: ‘I know, I’ll use regular expressions.’ Now they have two problems.”— Jamie Zawinski

As attributed by Jeffrey Friedl.

Broadly speaking, mobile frameworks can be divided into two categories: those that focus on Javascript utilities, and those that also provide cross-platform interface elements.

Frameworks differ vastly in scope: Sencha Touch is huge (over 300 KB), but it also provides the most robust development platform across devices and operating systems.

One big difference to note is that unlike most other frameworks, Sencha Touch is written entirely in Javascript (rather than building on HTML markup).

On the opposite end of the spectrum is Zepto, aiming to be under 5 KB in size, and MicroJS, which allows you to pick and choose small components as you need them.

jQTouch is an interesting beast, but it’s definitely more keyed to creating specific kinds of interfaces for mobile Web apps.

I think jQuery Mobile is somewhat poorly named; it really should be “jQuery UI Mobile”.

Two that seem to fall somewhere between jQTouch and Sencha Touch is appML (heavy) and Jo (very light).

One important thing is to remember that smartphones aren’t limited to the iOS and Android platforms; there’s also webOS, Blackberry, Windows Phone 7 and Symbian out there, to name the ones off the top of my mind. This is the whole reason interface-oriented frameworks exist.

Finally, you should always keep in mind that perhaps you don’t need a framework at all.

Nokia’s former head of design Adam Greenfield touches on its troubles innovating.

Nokia’s problem is not, and has never been, that it lacks for creative, thoughtful, talented people, or the resources to turn their ideas into shipping product. It’s that the company is fundamentally, and has always been, organized to trade in commodities. […]

Nokia’s engineers were and are brilliant at this. I am so far from an expert on the topic it’s not even funny, but I’d feel comfortable wagering that there is still no organization on the planet more capable at designing the guts of a phone, the various antennae and radios-on-a-chip that allow a handset to communicate with a network. Nor are there many who can compete with Nokia on the ability to optimize a supply chain and bring in a given bill of materials at a given (and generally astonishingly low) cost.

Greenfield illustrates this with an example on NFC:

I was given an NFC phone, and told to tap it against the item I wanted from the vending machine. This is what happened next: the vending machine teeped, and the phone teeped, and six or seven seconds later a notification popped up on its screen. It was an incoming text message, which had been sent by the vending machine at the moment I tapped my phone against it. I had to respond “Y” to this text to complete the transaction. […]

It’s not that the NFC-based, phone-to-object interaction didn’t work. Of course it did: it had been engineered perfectly. But what it hadn’t been was designed. Those responsible for imagining the interaction apparently wanted to protect users against the (edge case!) contingency of someone making off with their phones and running up a huge vending-machine tab. They failed to understand that, for low-value transactions like this, at least, the touch gesture is a useful proxy for consent — and that if someone’s got physical possession of my phone, I’m likely to have bigger problems than whether or not they order a few cans of Coke with it.