Posted on

Mobile computing: Whither?

Ten years back one would not have imagined that the cell phones would become much more than a cordless phone with a wide signal range. From a communication device that can store a few contact numbers it has, over the years, become ‘smarter’ by providing integrated features such as digital camera, FM radio, music/video player, games, multiple connectivity modes for the common user. For the advanced or business user it provides a Personal Information Manager (PIM), mailing features (push or pull), other office products and most importantly the ability to run business applications. There are, however, several other parameters that will dictate the way in which future mobile applications would run on modern mobile devices.

Hardware – PDA v/s Smart Phone

The gap between a Personal Digital Assistant (PDA) and a smart phone is getting shorter day by day. Most of the advanced phones these days have QWERTY keyboards, higher form factor, excellent processing power and almost the same memory capacity of comparable PDAs. As smart phones get more and more data-centric, the possibilities of running powerful LoB applications get better. Hence if the current trend continues, the long-term future of mobile computing will be revolving around smart phones.

Online v/s Offline

The bandwidth worries are to become the story of the past! As technologies are being developed to transfer megabytes of data at lower costs and in a matter of seconds, ‘almost-offline’ kind of applications will make way for ‘almost-always-connected’ applications. This would mean that the mobile client technology will run applications that are online with the enterprise backend systems but would not mind even if the connection is disturbed for a while. On restoration of the connection it would continue synchronizing the data that was collected or changed during the disconnected time period. This behaviour is comparable to that of the Microsoft Outlook mail client. Standard client side protocols will be soon available to manage session, state and queue persistence to facilitate the above concept.

Enterprise mobile apps v/s custom build

Built-in mobility would be the theme going forward. This would mean that you build your enterprise or mid-scale applications once and mobilize it using ‘a mobile infrastructure’ without significant effort. The organizations who build such mobility frameworks and build their enterprise applications keeping this flexibility (possible usage modes and channels) in mind would be the leaders in future business computing. These naturally mobilized applications would surpass custom build mobile applications that are difficult to maintain in the long run.

Public v/s Private network

A future mobile application would be able to run within a private corporate network as well as a public wide area network facilitated by a service provider without any difference in behavior at all. However, the usage pattern and the type of mobile applications in reality would vary within a private network and a public network. For example, a mobile service application may use a public network more often than an assembly line monitoring applet that is running within a manufacturing plant. The point here is that the usage pattern and location should not determine the architecture and technology behind the application.

Accessibility

In the future, voice activated NLP commands and touch screen entries will be used more often than keyboard entries. Mail usage will be around templatized content and smarter word fillers would make whatever entries to be made, a lot faster. Probably there is a lot of limitations when it comes to input mechanisms but something really innovative needs to be implemented here. Seamless integration and accessibility between office applications, PIM etc and the mobile enterprise applications will be what power users will be looking for. In order to achieve this the standards for exchanging data between PIM and mobile applications need to be defined.

Which technology – Java v/s .NET v/s ToBeInvented

As long as a standardized mobile platform protocol is available, it will be insignificant to talk about the technology used to write the applications and/or frameworks. Even the operating system may not play a bigger role here. The present problem is that the cool proprietary features that various mobile devices provide are often mistaken as built in O/S features rather than added application features. With conscious effort the integration points (PIM, Enterprise backend connectivity, Media players, Synchronization, Device management etc) can be really standardized. At the moment, open standards such as OSGi is purely talking about a raditional open application architecture that is far away from the current needs and capabilities of mobile computing

Future mobile computing device (MCD)

If the technology grows at the current pace, we would soon get to see devices that have computing power and memory capacity of the current desktops. They will also have foldable LCD/Plasma screens (something like the airport multi-display-monitor hoarding) that can change form factors easily and could substitute a laptop or tablet PC via retractable keyboard mat. Seamless synchronization with other devices (not just computers) will be a reality. Video streaming and recording features will soon be utilized for mobile TVs and video conferencing services while on the move (using 3G or ‘moreG‘). Also, don’t be surprised if all your TV channels are accessible via your smart phone in less than two years from now!

Posted on

Enterprise SOA for Agile Enterprises

To improve is to change; to be perfect is to change often – Winston Churchill

Sir Churchill seems to be completely right in a general sense, and it is very much relevant in this modern era of information technology. In a competitive, complex and dynamic business environment, organizations are fast realizing that the ability to transform their IT for their future business needs will determine their success. An agile enterprise or an organization that continuously reinvents itself, would be successful only if they have a flexible IT infrastructure that depicts their current organizational structure and business domain. The need for this kind of flexibility resulted in what is called an enterprise architecture that is nothing but a bundle of loosely coupled functions and processes from the organizations’ point of view.

An enterprise architecture gets even better if it is a Service Oriented Architecture (SOA). An SOA is a collection of services that can interact with each other to carry out the business processes. Each service in this case, would be self-sufficient, in terms of data and state maintenance, and would form a node in a distributed computing environment. Each business process, thus becomes a series of request-response cycles involving one or more services. Please note that we are not necessarily talking about web services here.

SOA really puts some of the beautiful technology concepts into practice – Reusability, Encapsulation and common protocol for usage and data exchange to name a few. Each independent service as such could be implemented in any language or technology platform. It also provides tremendous possibility for organizations to integrate their existing solutions with other external systems by keeping similar protocols for publishing and accessing the services. This is achieved via the standardization of protocols by leading independent research & standards organizations like World Wide Web Consortium (W3C). We may recall that first generation service oriented architectures like DCOM or CORBA didn’t help everyone mainly because access protocols across heterogenic systems were not standardized.

Enterprise SOA seems to be the new buzzword now, that promises a scaleable, adaptable and open service oriented architecture for developing and/or adapting enterprise solutions. Leading business software vendors like SAP is investing heavily to port their traditional ERP and domain specific enterprise applications and suites towards service orientation. While newer applications can be fully service compliant and will work as per the current vision, it is yet to be seen how organizations can adapt their existing applications to the SOA platforms that these vendors offer. One-time adaptation may not be a feasible option for many huge organizations. So they can buy the new enterprise SOA applications for fresh requirements and integrate them to their existing infrastructure. Over a period of time, probably they can adapt the systems towards enterprise SOA, and until it is fully achieved they have to be satisfied with a side-by-side model in their IT infrastructure.

A clean enterprise SOA solution is not there in the market yet. However the next two or three years could witness a surge of service oriented enterprise offerings for both medium and large scale organizations. The success of these solutions at the end may depend on enterprise-wide protocol standardization (how long the current standards last?) and lower cost to get started (for IT optimization). After all, these technological shifts should help the enterprises to concentrate more on their core business rather than spending time and resources on IT infrastructure alone.

Posted on

Microsoft and the handheld market

There are not too many desktop or personal computing related opportunities that Microsoft missed out to capitalize in the past twenty five years. In most cases they were early enough to react to the changing market needs by either innovation or adaptation of other cool technologies. However, Microsoft seems to have missed a trick or two when it comes the voice-centric devices and the related platform requirements.

According to Gartner the handheld devices can be broadly categorized into data-centric devices and voice-centric devices. The PDA devices fall into the first category where as the smart phones represent the second category. There could be PDAs that offer cellular connectivity but they are still ‘data-first and voice-next’ devices. PDAs usually have a higher memory, processing power, bigger form-factor as well as keypad compared to the smart-phones. The other differentiator is the usability – PDAs are supposed to be used by both hands where as smart phones are targeted mostly for one hand keyboard/wheel operation.

Microsoft has been the undisputed leader in the PDA operating system market for a while now. According to the sales statistics for the year 2006, they command 56% of the market share where as their nearest competitor Research In Motion (RIM) have got less than 20% of the PDA O/S market. Palm may be a leader in the Americas but overall they get only 12% of the market share. The story is totally different when it comes to the smartphone O/S market.

Microsoft devised a plan five years back to enter the smartphone market but the ground reality is that they were not successful at all in this war field. The Nokia led Symbian O/S commands 67% of the market share in the smartphone category as far as 2006 sales data goes. Blackberry has been a disruptive innovation (Read: Disruptive Innovation Model by Clayton M. Christensen) that thwarted Microsoft and several other market players and going forward probably even Symbian. Probably the Windows Mobile story didn’t click as much in the smartphone market.

Though a little late, Microsoft seems to be realizing their mistake now. Since 2006 they have been a bit more aggressive on stabilizing and positioning Windows Mobile. Year 2006 saw Microsoft market share growing by 200% but still that figure contributes to less than 10% of the market share. Gartner projects that the PDA market is growing at a much slower pace (CAGR of 5%) compared to the smartphone market that is growing at a CAGR of 47% for the next 4-5 years. If Microsoft is smart enough to do a good job with respect to Windows Mobile for smartphone, it is highly likely that they can rule the market by 2010. After all, the success of smartphones is generally governed by the quality of its O/S, usability and accessibility – and Microsoft is good at all these aspects. It is also interesting to read another research report published by the Diffusion Group. They project that Microsoft will have 29% market share, followed by 26% and 22% respectively for Linux and Symbian in the year 2010. This could be true if they continue to grow at the way they did in 2006!

Personally I believe that the success of smartphones will also be driven by the easy to use Personal Information Manager (PIM) features along with accessibility. Since Microsoft is the undisputed leader in office products, they are the ones who should emerge as leaders in the handheld market as well.

Finally, symbian and RIM may be the leaders in the cellphone (smart or otherwise) market purely based on usage. However, both these platforms are
nowhere near Microsoft Windows Mobile based handhelds and smartphones when it comes to running powerful applications for Mobile Sales, Service, Retail or Logistics business scenarios. For that RIM and Nokia smartphones have to be reborn and have to start supporting pure J2ME than limited device profile frameworks that can be used only for small applets, alerts and basic PIM usage.

The beauty of the Microsoft WM platform is that the same .NET application, without any change, will run on a Smart phone as well as PDA device. This is the
area (on top of PIM) where Microsoft will beat their competitors big time as the question of recoding the business applications several times for device types will not arise in the case of Windows Mobile. So my final take on the handheld market is that Microsoft and Linux will be ruling the enterprise mobile business market (including smartphones) in the mid-term future where as Nokia will continue to be the undisputed leader in the cellphone market.

Posted on

Windows Vista is here!

Microsoft has just announced their new fancy operating system namely Windows Vista. Vista in English language means ‘panorama’ or landscape view. Well, the naming could not have been better as the capabilities of the new O/S ends very much around how your application windows looks and are viewed. Beyond that is it really worth? I guess it’s not.

I used Microsoft Windows for application development since Windows 3.X (Windows/386) era. Since then I have seen how the Windows O/S evolved and I can say that Microsoft is pretty good in marketing whatever Windows flavours they came up with. After Windows 98 Second Edition which had a very good user interface for common man and reasonable stability, I thought NT 4.0 was a pretty cool Operating System for the developers. It may not have been a gamer O/S but
it had everything in it for application developers as well as corporate server requirements. If you categorize PC users broadly into three categories – Application Developers, Information workers and Gamers – I would think that the 98SE belonged to the information workers and NT belonged to application developers. Now that leaves us with one category of RIG lovers and probably Vista is for them?

Though I used Windows 2000 professional edition extensively for development as well as normal PC usage of browsing, chatting, editing etc for almost five years I somehow liked Windows XP better – Well, probably except for its stupid Start menu that requires your mouse to be moved vertically and horizontally several times before I could achieve something. I thought Windows XP was almost complete as a UI-rich Operating System until Microsoft announced the arrival of the Vista! Now, let us see if it really a major technology and usability revamp over XP.

If you believe in love at first site, well then Vista is for you. For the first time in Microsoft’s O/S history you get to see all your application windows arranged in an ‘almost 3D’ manner. This is quite an eye-catching feature along with the translucent menu bars. Microsoft calls it an aero desktop – again quite convincing pet name! But beware, the basic home edition doesn’t support flip 3D windows. This is a setback for normal home users who are probably the main target audience for this kind of features.

Beyond the catchy looks the other value add that Vista has is in terms of the number of tools that it is bundled with – Windows Defender and firewall, Windows DVD maker, Instant search, Windows movie maker, a bunch of 3D games to name a few. But will somebody buy an operating system for these fancy tools? God, err… Billy, alone knows. Wait, the worst is still not over – The so called Business Edition is something in which I had great hope on as I thought it is a
professional developer environment. But Microsoft feels that a few administration features like scheduled complete PC backup/restore, remote desktop, drive encryption etc would automatically make your Home edition a Business Edition of the O/S. Should we blame them, probably not. I think they have got their market and anything that is dumped on the users will be gracefully accepted.

The impact that Vista rollout has on corporates would be in terms of additional investment on aero-supported graphics cards. Also, if home edition PCs are equipped with these expensive cards, the prices for the final assembly are surely going to shoot up which will adversely affect the PC market for sometime. One of the main features WDDM (Windows Display Driver Model) that claims to be the PC saviour in case of a display driver failures definitely requires you spending a few additional bucks while upgrading – All this on top of the $300 that you pay for the Vista Business Edition.

All in all, other than for a little bit of enhanced usability for common users and some pyrotechnics that comes along, I would not buy Microsoft Vista. But then, it’s the Microsoft’s market and sooner or later it will get you there. I am counting down days before I am dispossessed off my good old XP 🙁