I like mobile phones. When traveling internationally a good phone is pretty important. More than just a way to talk to people, my phone has increasingly become entertainment and guidebook. I tend to buy phones based on what I think I’m going to want to do, not what I do today. I want a new phone, but I’m trying to convince myself I don’t need one.

Somebody who travels a lot recently asked me what sort of phone they should get. Initially I couldn’t really give them an answer, but I wrote most of the following as an email that works through some of the background noise in my head.

According to Market Share data synopsis from Canalys report “Worldwide smart mobile device market” which I pulled off of wikipedia, the top Smartphone OSes are,
1. Symbian (has UI Variants of S60 and UIQ) 57%
2. RIM (Blackberry) 17.4%
3. Windows Mobile 12%
4. Linux and variants 7.3%
5. iPhone 2.8%
6. Palm 2.3%

In the last few year or so we’ve seen several things happen to change the mobile os playing field. I’ll talk about that a little later, but in the existing market Symbian is king by the numbers. We don’t hear much about it in the States, but it’s hugh overseas. Symbian is the core OS and there are two graphical overlays that can run on top as UI. S60 is used by Nokia and has the lion share of the market. Sony Ericsson uses UIQ which is adds touch and stylus support. I have a M600i which I really like, but the software is flakey and UIQ doesn’t have much the market penetration, so is probably going to go by the wayside unless Sony-Ericsson does something amazing. Nokia’s uses the S60 overlay and their N Series for lifestyle and E Series for business have always been really interesting to me. I would have probably gotten the E70 instead of the M600i if they weren’t so expensive. Symbian is a good OS, but I think that it has too much momentum in it’s current paradigm to be able to react to recent changes successfully.

RIM has been killing it with their Blackberries. I had a Blackberry and I liked it, but it made me just too connected. The way that they do email makes the data plan expensive and for an international traveler like myself it’s probably tricky to use with multiple sim cards. The storm adds a touch screen but it’s still focused primarily on business so it has some perceptions and limitations to over come for consumers.

There has been speculation that Microsoft will buy RIM. It’s debatable how likely this is. If it does, it would be an acceptance that Windows Mobile is a failure. Microsoft hasn’t been doing anything interesting with their Mobile OS for a while. Windows Mobile 7 starts to be interesting but that’s not due to 2009 and it won’t be till Version 8 that they’ll completely catch up to what the consumer thinks the iPhone does today. So if Microsoft doesn’t acquire RIM it will take them some time to get Windows Mobile to be competitive and even if they do it will take them time to rebrand and integrate it into the rest of their line up. End result is they’re going to be a backseat driver until 2010.

As for Palm well they haven’t updated their OS significantly in a long time and while it was a great product, the company has been mismanaged, lacking vision and unable to figure out what they’re doing. Again unless they do something amazing they’re not going to be a player.

Like I said before, there are things that have mixed up the status quo which I think are game changers.

1. iPhone – Techies have been complaining about how bad Mobile Phone OSes are. They were all bad and most of them still aren’t great. Along comes the iPhone and shows them how bad they really are and in a little over a year it has gone from a vapor to nothing to an icon. What this taught mobile manufacturers is that it doesn’t matter how good your hardware is, it’s only as good as the software it runs. They’ve shipped something like 6.9 million of them this quarter which was more than RIM did. It’s an amazing phone. They’re expensive though and only sold unlocked in Hong Kong.

2. Android – Google’s entry into the Mobile space. It’s not a phone but an OS. Aside from the fact that it has the Google brand and presumably their special sauce, the interesting thing about it is that it’s been open sourced. There is only one released phone with this OS, the HTC G1. There’s some controversy on how many they’ve sold, but it seems to be selling pretty well.

3. Convergence – The One Laptop per Child project by Nicolas Negroponte was a great idea. It failed because a certain level it was sabotaged by Intel and Microsoft. Even in it’s failure it catalyzed a whole new market of computers known as Netbooks. In the race to the bottom they’re getting smaller and cheaper. As phones get larger and more powerful the line between Netbook and smart phone starts to blur. ie is better to get a expensive phone or tether a cheap $200-$400 netbook to a cheap phone via bluetooth? Add devices like Peek (http://www.getpeek.com) into the mix which is an email only device for the states that’s contractless and only $20/month, MIDs (Mobile Internet Devices) and wifi based skype phones, and things get even more gray. For example, a device like the Nokia N810 WiMax edition which I would classify as a MID is only slightly larger than a phone and runs linux. WiMax is a next generation high speed data protocol which promises high speeds, large coverage areas and cheap access. With that scenario, I’d only have a WiMax data plan and get phone service through Skype.

The mobile device is evolving into a standalone platform and what we’re seeing is the equivalent of the OS wars in the early 90′s when you had Microsoft, IBM, Apple and a bunch of smaller players, Amiga, BeOS, etc. etc. duking it out. But by the time 95 comes out we have a fairly clear winner. Over the next year or two this space is going to dramatically evolve until there are only a few players left. My bet is that unless something dramatic happens to change the status quo in 2 years that will be Android, the iPhone OS and what ever RIM evolves into.

Addendum: I would have posted this earlier, but Robert Cringely wrote made pretty much the same observations more eloquently about the time I finished writing up my ideas. http://www.pbs.org/cringely/pulpit/2008/pulpit_20081023_005500.html

The reason I actually write now is that I saw Gartner’s numbers for 2008 and it’s not pretty for Nokia and Microsoft.

The new list is
1. Symbian 49.8%
2. RIM 15.9%
3. Mac OS X 12.9%
4. Windows Mobile 11.1%
5. Linux 7.2%
6. Palm 2.1%

This is pretty much confirms the trends I’ve been noticing. When it really comes down to it, we are the cross roads of the next generation of computing. For better or worse this is the point where the functionality of mobile devices starts to compare to that of full blown computers. There is one device out there that can be purchased today that is part of that future and has a roadmap. The iPhone. If I was in the States I’d walk down to the Apple store and buy one, end of story. But I’m in Bangladesh and that isn’t an option. Everything else may have promise but isn’t quite there yet. I’m most tempted by a Nokia N series device, like a N79. It’s a very cool phone and while it would work very well today with a few more hardware features, it wouldn’t be significantly more capable than the 2 year old phone I have. The other thing is it doesn’t have a migration path to have a place in that mobile future. Operating systems like Android and mobile Linux have the potential but don’t have have a handset that offers the hardware I want. I could come from the other direction and get a Nokia N810 but we don’t have WiMax in Bangladesh quite yet. Net net is that the mobile devices are going through a lot of growing pains and I might have talked myself out of a new phone.

It’s Thursday which means that it’s the end of the week. Because Muslims worship on Friday, the weekend becomes Friday and Saturday. Come Sunday it’s back to work. I’m looking forward to this weekend because it’s a much needed break. Last weekend was lost to a MCC all staff training that was mostly in Bangla, and getting sick. Bangla learning is coming along well but not well enough to follow a training like that. So far I can stumble through sounding out the worlds in the script and I’m up to 435 Bangla flash cards. The goal is 2000.

As I sit here in front of my computer, I hear my roommate in the other room talking to his Xbox 360.  He’s actually playing in a multi-player online game and communicating with people via the in game voice chat.  I think it’s Halo 3 but the actual game he’s playing is irrelevant.  They all seem to have the same technology these days–high resolution, hi fidelity landscapes that one interacts with with friends and strangers alike.  So what relevance does this have to IT?  It’s _simple_.  Simplicity is something we in IT seem to run away from as fast as possible.  His video game presents a cohesive interface that allows him to interact in a virtual world more vibrant than the sci-fi portrayed by major motion pictures 15 years ago.  It is high technology made accessible and put in the hands of the masses.  Comparatively the interfaces of our IT computing systems are in the stone age.   At best the collaboration tools I use are overlays in my environment.  Band-aids slapped on a series of deep wounds.  So what am I complaining about?  My problem is that software development and management should think deeply about the medium in which it is presented.  I spend my day working in 4 windows.  My email, my web browser, my instant messenger and my shell session.  I happen to like my shell session most of all.  It’s the most powerful interface I have because it can encompass the other 3.    I can control just about anything from it and with one command I can travel from Durham to Boston, Florida even China.  I’m talking about bash and ssh by extension.    There are days when I want to be entirely in a shell session.  I run screen, I have finch for IM on one window, mutt for email on another and listen to music or control video playback on a third.  I used to work on mainframes and there it was TSO over TN3270.  Terminals are very powerful interfaces to those who know how to drive them.  Individual pieces of software can be wonderful.  However, whenever we have to change context, change from one user interface to another, it takes time.  It’s the integration that is the problem.  The glue.  We need to thing about changing the paradigm of the way we get to interfaces and creating a landscape that fosters innovation.  My roommate has a persistent Voice over IP (VOIP) connection with everyone in his game.  He doesn’t have to start to initiate a voice call, it’s already going.  Everyone is in the same landscape with different views of the same information.  In a day and age of working mobile, virtual team, and reduced travel budgets, IT solutions need to make screen sharing and collaboration as intuitive as Halo 3.

Have you seen any of these articles saying how Vista is terrible and Vista has failed?  The truth of the matter is that Vista tries to be everything to all and failed.  This is no fault of Vista.  The problem is that the paradigm of individual operating systems is changing.  A slashdot meme that is slowly fading is “imagine a beowolf cluster.”  Now we talk of virtualization.  What changed?  The hardware got more powerful than it wasn’t as important to make look two like one as to make one look like two.  I read an article recently that touted the IBM mainframe because it was able to have 85% utilization vs. 15% utilization of commodity hardware.  In the same article a VMware spokesperson made the claim that their software also allows hardware to be 85% efficient. I used to work on mainframes and  everything was done on a tn3270 session.  Mainframes were carved up with z/VM to allow lost of z/OS images to run in parallel, but it was always done with a tn3270 session.  One interface.  I never saw most of the machines I worked on because I didn’t need to.  The only time I had any sort of physical interaction was when I had to load tapes for maintenance and even that went away at the end. 

Command line interfaces aren’t going away any time soon.  Microsoft’s offering a headless version of Windows 2008 is an acquiescence of that fact.  However, a text mode console isn’t always the best interface.  They are ill suited to deal with images, or non-linear editing or casual users.  At the other extreme of a command line interface is immersive 3D.  There’s been a lot of talk about Second Life recently.  To anyone who’s read Snowcrash by Neal Stephenson, you recognize that Second Life is a poor interpretation of The Steet.  Second life is a video game.  It’s a user interface that attempts to be a representation of data.  It is a top down approach and as a result is semantically entirely unlike the street.  The street was an abstraction of data, an aggregate of interfaces to systems, but it’s a vision of what might be. 

These days I work with the previously mentioned “commodity hardware” and while I’m working with decentralized hardware,  I still can do remote administration via KVM over IP.  KVM over IP is wonderful although it can have a few limitations, like the need for an addition module for remote hard power cycling.  The rest of my time I use VMware.  When I’m sitting at my computer doing remote administration.  KVM over IP and VMware present a very similar interface.  The catch is that I can’t access them from the same interface. The latest VMware server beta has a Web interface.  The console doesn’t work properly in Firefox on Ubuntu, so I go to a VMware image running windows and access it that way.  The abstraction of processing data and power is no longer at a individual OS level but data center level.  A result of this shift is that I don’t care if my computer crashes nearly as much as I care if my connection is down.  Whether testers/developer/it professionals/users are presented with VNC, Remote Desktop, a RMA or VMware console or KVM over IP doesn’t matter, to the end user it’s a window. The paradigm of “my machine” has changed.   So the way we give them the box should be changed.  We need to envision our entire IT infrastructure as a cloud which we carve out bits which are given to people as needed and returned to the cloud when finished.  Amazon’s Elastic Compute Cloud (EC2) and Simple Storage Service (S3) are early examples of this mentality.   S3 provides a simple API for data storage of any size and similarly EC2 provides a API that allows creation and control of an Amazon Machine Image (AMI).   These services are too homogeneous for an test organization because they currently only supports x86 Linux images.  However it points toward a comprehensive test framework with network sandboxing, storage and different architectures all controlled from a common API.

What should IT look like in the future?  The paradigm of “my computer” are going to look a lot more gray and we must answer the question of how we bring together communication and work sessions.  How do we pull together VOIP, chat, shell sessions, email and other collaboration tools?  Lets start by looking at video games.