The Bloor Perspective: Java's big hope and Microsoft's major headache

In their final analysis of the year, the Bloor team look at Java's potential on mobile platforms and the recent wobbles at Microsoft - both of which will undoubtedly continue to dominate the headlines next year...

Java and mobile have much in common. Both technologies have a high theoretical and conceptual promise: Java promised platform independence and applications delivered remotely on demand, mobile promised a world of roaming data away from the ethernet desk chain. But the first Java products were grindingly slow to execute, and their claims for uniform inter-operability were quickly exposed as absurd. The backlash was tempered by anti-Microsoft sentiment, but was still merciless. Saturation point had been reached, and media experts and pundits wrote wry columns on the Java hype phenomena for a while, after which the technology was relegated to the headline B-list. Engineers and developers were free to press on in their labs without being distracted by commitments to the publicity machine. Mobile is currently weathering its own backlash. Circuit-switched WAP, popularly misinterpreted as simply 'WAP', was rightly panned for its poor usability. Packet based GPRS services are poised to deliver packet billed applications with better bandwidth than circuit-switched WAP, but the interim nature of the technology and hype-bred anticipation promise heavy weather. And right out there is 3G, which has already been written off by those with pocket calculators and near horizons as a total waste of money. When a technology has been over-marketed to the point where it is simply too naff to criticise in the server room or at the free publicity bar, it usually follows one of two paths. The truly dreadful will sink completely off the radar. More likely a decent product suffering from over-saturation will go from strength to strength until it becomes too boring to criticise. It is because it is simply everywhere. The word 'Microsoft' springs to mind here. Java may have failed to flood the PC LAN with applets on its initial launch. However, mobile and Java have so much in common, and it is beginning to look like they are A-list celebrities made for each other. Whether you are an enterprise that wants to remotely manage a fleet of business devices, or a carrier looking to score a major dollar from a volatile and unpredictable consumer market, Java's promise of re-usable, downloadable applications on demand is compelling. The Java developed for wireless, Java Platform Micro Edition J2ME, is not the all-singing, all-dancing everyman promised in the original hype. There are other microJava JVMs. However, in September, the J2ME Mobile Information Device profile was released for application developers. The membership list was comprehensive. All the major handset manufacturers support J2ME. Epoc and Symbian support J2ME. DoCoMo supports J2ME. The only major player that doesn't is Microsoft. At the moment a mobile device needs at least 256k of memory to run Java. The current bottleneck is bandwidth for the downloads: an applet will be about 10 to100k, a sophisticated application between 100 and1000k. In contrast, the maximum bandwidth of GPRS is estimated to be about 40 kbps. Applications look set to be downloaded once and stored in persistent memory until erased to make way for the next one. There are two different versions of Java developed applications, as outlined by the white paper on the Sun website. The first high-level APIs have a high degree of abstract content and concentrate on delivering an accurate service from a central site. The developer has little control over the final look and feel of the high level API based application. Low level APIs, on the other hand, are mapped much more firmly to the hardware of the device, and allow the application to exploit the physical hardware such as the screen. Games are the most typical application of this type. Last week Bloor saw its first wireless application server written entirely in Java, 'elata senses'. Demonstrated with a Psion 7, elata senses showed strong potential for remote management. Profiles were generated on the server and mapped to applications, and then dragged and dropped to a specific fleet of devices. If a new application, such as a car rental expenses form, was needed, this could be wirelessly downloaded onto the mobile system on demand. All applications could synchronise on connection, and everything could be recorded on the server. These applications and platforms are here now. The major weakness in the plan, acknowledged by both elata and Sun, is the current lack of decent Java programmers on the market. Microsoft, with its .Net initiative, is very hot on this issue. When .Net applications are finally available, Microsoft claims they will be cheap to build, as they use XML-based data, SOAP, Visual Development 10 suite and the full range of the programming community. Microsoft versus Oracle: Let battle commence Microsoft has just issued its first ever profits warning. It stated that sales were weak in all areas, especially with US PC sales dropping and corporate spending on IT declining. The company believes that this effect is simply highlighting a global slowdown in the world's economies. At the same time it reported that its online service has experienced a downturn in advertising revenue. So should Microsoft be worried? It is interesting to contrast Bill Gate's business situation with that of Larry Ellison over at Oracle. Oracle has reported booming results, on both the database software side, with sales up 19 per cent, and the applications software business, which registered sales up by a staggering 66 per cent. Total software license revenue rose to $1.1bn and services were up to $1.5bn. If Oracle can deliver a functionally rich application suite to complement its strong database products, Ellison has every reason to smile. After all it has long been known that Ellison wants to displace Microsoft from the number one software company spot. With the rapidly changing desktop market place and the continuing battles on server platforms, Microsoft faces a hard battle to maintain its margins. Oracle on the other hand may be well placed to exploit its traditional strengths and be able to expand the growing application revenue streams. While Microsoft has few immediate reasons to panic, it is clear that the maturing worldwide PC marketplace leaves several challenges to overcome. There is also the issue of the long running legal disputes on both sides of the Atlantic. During the next couple of years it will be fascinating to watch how the fortunes of these two software heavyweights vary. Will it be taciturn Bill or extrovert Larry who ends up with the biggest grin? For more analysis, see http://www.it-director.com