@tomiahonen: Windows Mobile & the lack of a migration path

I used to read Tomi Ahonen‘s “Communities Dominate Brands” blog religiously. Stats, stats and more stats, and analysis that took place outside of bubble of the US blogosphere. Now I mostly just skim what has generally turned into a collection of rants about the stupidity of Stephen Elop (though, strangely, never the Nokia board which has signed off on the strategy), the evil of Microsoft, and the general foolishness of believing that the US are anybody in the mobile space. I also used to be active in the comments there, first often in assent, then trying to argue against the more obvious misgivings and factual mistakes.
Now I don’t bother much anymore. But, very occasionally, I still feel the need to comment on something. In this case it’s a tweet from yesterday: “The collapse of MS happened only after MS moronically announced ‘no migration path’ from Windows Mob to Phone7”. The collapse in question here is Microsoft’s market share in the smartphone sphere.

The opinion expressed in this tweet contains two huge mistakes:

It is first of all taking correlation for causation. There is the direct assumption that Windows Mobile was still a viable OS at the time Microsoft announced the switch, and that what prevented users from buying Windows Mobile handsets in the time between the announcement and the launch of Windows Phone was just the announcement of no migration path.
But when you look back at any review of new Windows Mobile device at the time, the main question software side was how well the manufacturer had skinned the device to avoid bringing the user in contact with the system itself. The ideal Windows Mobile device would have been one that completely hid this fact (were, apparently, the HTC HD2 came pretty close). Windows Mobile wasn’t just criticized by reviewers – it was ridiculed, in far worse a way than either Symbian or Blackberry OS have been since. This was universal, not just by the US tech press/bloggers. HTC, as the main quality provider of Windows Mobile devices, had already started to enter the Android Market, and all the other manufacturing partners that Microsoft had were either starting to explore alternatives, or were already on a shift away. Windows Mobile was about to crash and burn no matter what.

The second mistake is assuming that a migration path from Windows Mobile to Windows Phone was an option at all. Migration paths are not a simple matter of decree by the management. They are hugely difficult technical undertakings. Keeping the old enabled while introducing the new doesn’t just add complexity – it multiplies it in parts. The resulting system is universally less perfomant, less reliable, and presents a worse overall user experience than a complete fresh start would have provided. This was doable up to a point with Windows on desktop systems, where each new generation of hardware has always had computing cycles and memory to spare, and the electrical juice to run the extra burden. Processing power, memory and, most of all, battery life are at a premium on mobile devices, so that the extra overhead presents much bigger problems here. More importantly, on Windows for desktops, the input paradigm for the UI has remained constant across versions up until now: mouse-operated pointer control. Migration paths have been a question of under-the-hood technology. Windows Mobile to Windows Phone represented a transition from resistive screens and stylus operation to capacitive screens and touch. This is not just a change of technology, it’s an entirely different UI paradigm. Integrating the old would have meant accommodating these two different input paradigms – which, on a mobile phone, is a recipe for sure disaster. Offering a migration path that enabled the old software to run on the new devices would have accomplished little for the users (just try hitting that 5 x 5 pixel button in a row of three like ones with your finger instead of a stylus), but dramatically worsened the overall experience, and terribly restricted what Microsoft could do in terms of new architecture. There was no way to salvage anything from Windows Mobile without endangering the  viability of Windows Phone in the marketplace.

So: Windows Mobile had to die. That Microsoft made a clean cut and did not offer a migration path is what enabled them to come out with a system that at least has a shot at #3.

PS: regarding the developers, who really were pissed off at Microsoft at the time? Windows Phone currently is the third ecosystem in one respect only: developer mind share. It’s surpassed the Blackberry AppWorld and Symbian, and the new developer tools are universally lauded as the best in the business. There are now more applications for Windows Phone than there ever were for Windows Mobile, and that with the new system at a fraction of the old one’s market share. This was a gamble, but it didn’t turn out as quite the catastrophe that could have been expected.



Filed under mobile

3 responses to “@tomiahonen: Windows Mobile & the lack of a migration path

  1. mirmit

    I think the migration referred by Tomi Ahoneen path is not about the ability to run a new OS on an existing machine but to have the software developed for a generation of OS/device able to run on newer platform.

    The migration path is hard to put in place. We saw some nice implementation with Nokia, giving some extra life at each ‘technology gap’, except when moving from Symbian 8 to Symbian 9, to the existing applications. And still enabling developers to migrate their applications with some ease.

    In case of Microsoft, it has been decide there will be NO migration. All the existing software asset has to vanish – almost completely.

    The limitations put by Microsoft on the system lead a lot of professional users to look at alternatives, as there was no mean to have some VoIP application on the device, or other background apps.

    Those customers have migrate their solution to Android for those looking at mass market devices with a very low chance to have them return to Microsoft platform.

    • I was also referring to the migration of existing software.
      At the UI level, the differences both regarding the input and the changes that Metro brought were so great that allowing old software to run unchanged would have made little to no sense. Easing migration of existing code to the new platform would have both increased the overhead (versions of old libraries that run under the new system, offering some of the old APIs) and meant that Microsoft would have had to open up the system more than they want to at the moment.
      Regarding the migration of professional users: It seems that Microsoft wanted to target the consumer market first, with the professional market as an afterthought. So yes, for the time being Android is the OS to go to if you want to do custom things that require e.g.background tasks, or certain kinds of access to underlying system layers. As with any radical shift in strategy, this one was and is not perfect.
      Personally, I would have greatly welcomed the ability to run native code on Windows Phone, and I have a lingering suspicion that Microsoft are going to open the system up eventually, just as was the case with Android.

  2. Pingback: Disclaimer

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s