Musings on software product development.

As everyone who reads technology news knows, on June 9^th, Apple is rumored to unveil the second generation iPhone. It’s predicted to have all the features of the first plus twice as much memory (16/32 GB), 3G network connectivity, and a GPS. In addition, AT&T or Apple is likely to subsidize the phone by 50% to $200. What the analysts aren’t talking about is the fundamental shift in mobile computing that this release will bring. Apple may have lost the desktop operating system wars, but they are about to become the Microsoft of the mobile phone.

A subsidized iPhone is going to penetrate the smart phone market fast and deep. In 2004, the Motorola RAZR launched at a $600 price point. It was the sexy, sleek, must-have phone of its day. Sure it was expensive, but within 2 years it was carried by virtually every GSM network operator and could easily be purchased for $100 (subsidized with a two-year contract). The RAZR has now sold more than 65 million units worldwide. Apple has shipped approximately 6 million iPhones thus far. By subsidizing, I expect exponential increases over the next 12 months. As if that weren’t enough, Apples latest back to school promotion is giving every student or teacher purchasing a new computer a free iPod Touch. This is adding up to a whole lot of iPhone OS installations.

So what does it all mean? For one, if you are a web application company, you had better be seriously thinking about what the iPhone means to your business. Every iPhone user has constant access to your application via mobile Safari. This is a dramatic paradigm shift that will result in new business opportunities. Desktop and enterprise line of business software companies also have great new opportunities. A mobile device capable of rendering desktop-grade user interfaces with high-speed connectivity is what mobile computing is all about. Implications to industries requiring high levels of mobility (i.e. healthcare, network management) are substantial.

A lot of hubbub has been made over the restrictive Apple App Store business model, but serious software companies will make it work. Apple’s services and fees really only pertain to distribution. Once a customer has your application installed, alternative revenue models can be pursued outside of Apple’s domain. Think monthly service contracts, online feature unlocking, etc.

Who are the big potential losers here? First off, Dash needs to seriously rethink their business model. Their hardware (GPS + 3G connectivity) all of a sudden costs too much ($299) and does too little (Internet connected GPS). Compared to iPhone 2.0 their only remaining competitive advantage is the software. An iPhone compatible version better hit the iPhone App Store before a more nimble startup beats them to it.

Second, Microsoft is going to suffer some major market share losses. Windows Mobile has been widely criticized elsewhere so there isn’t much I can add to the conversation. Despite the tiny form factor, limited screen size and puny hardware, Microsoft attempted to cram in the desktop Windows experience. Like every mobile OS predecessor it’s a marginally usable platform that users only put up with due to lack of alternatives. On Monday, Apple’s (relatively) cheap and exceptionally well designed alternative is going to push Windows Mobile in to obsolescence. (Yes, I’m aware of the Windows Mobile 7 and 8 leaked specifications and screenshots. Sure, it looks improved, but right now it’s vaporware. They are playing catch up.)

What about Google? The videos from Google IO were impressive. They’ve got a more open platform and more manufacturers slated to build devices with Android. You’d think, surely Apple has major competition here. Perhaps. However, it’s unlikely Google will achieve the same level of UI polish that Apple has risen to. Open source third party software is notorious for ugly, cumbersome user interfaces. Android apps will suffer the same fate. Apple was smart to spend significant time establishing user interface tools and specifications from the get-go. Third party iPhone apps will feel as polished as the ones coming directly from Cupertino. With Android devices sporting nearly identical features to the iPhone, the battle will be waged on software. Apple has a distinct advantage here.

I’m excited to see how these predictions pan out. Mark your calendars because on Monday everything changes.

Account verification workflows have always perplexed me. What I’m referring to are the common steps you usually have to go through to register a new account on a site:

I understand the logic behind this (spam prevention, guaranteed means of contact, etc), but the task is arduous and repetitive. It’s also problematic if you sign up for an account on a computer without easy access to your email account. Not everyone uses webmail. In the process of brainstorming solutions I started pondering whether it might be possible to use SMTP email as a mechanism for performing stateless web services. The fundamentals are the same as services over Http. The only difference is that, email servers do not expose hooks for acting on incoming messages. You’d have to use a specialized SMTP server in order to get this functionality.

Or would you?

Every halfway decent email client I’ve ever encountered has a rule system for preprocessing incoming mail.

So, in your client you could setup a rule to forward certain messages on to a third-party mail server with Xml message handling capabilities. If the email met the criteria of an web service request, it would be forwarded on to a third-party mail server for processing. Your personal email address now becomes an Api for any number of exposed services. Getting back to the start of this post, one of the Api calls could be a check to verify accuracy and authenticity of a user’s email address. I could see this easily tying in with centralized authentication providers such as Windows CardSpace or OpenID.* Taking things even further, I could imagine using something like Yahoo! Pipes to redirect incoming email Api calls to fourth-party web services translating responses back to plain text or html via XSLT.

Doing a bit of searching, I see I’m not the first
  to come up with this idea. I even found a ready-to-use Python implementation of the whole process. However, no one has really embraced this approach. It could be I’m crafting up technology for technology’s sake here (admittedly not the first time). However, there is something appealing about the idea emailing a friend with a WSDL request and getting a list of services their email can provide me.

Of course, this does leave me with another unanswered question. How do you avoid the email address verification for the third-party provider?

* Yes, I realize this is a naive brain-dump that doesn’t consider security or privacy.