The Bloor Perspective: Java skills in demand, Corel's future with Inprise, and how to make the Web more relevant

This week, Robin Bloor and his team of analysts discover that Java skills are more in demand than C++; consider what the Corel/Inprise merger means for Linux; and examine a new tool that cuts out the wild goose in the chase for information on the Web

It's official: according to a survey by Bloor Research of over 40,000 job adverts in January, Java has taken the top slot in development language skills, with 36.8 per cent of all programming job adverts in our sample now specifying the language. This figure is just ahead of demand for C++ skills. Meanwhile demand for VB and C are holding up with perhaps a tendency to decline, both hovering at around the 25 per cent mark. These latter skills are still in widespread use but the slight decline suggests that we will start to see a drop in their use in 2001. To use a well-worn phrase, there are lies, damn lies and statistics. A previous survey showed that the requirement for Java-based projects was holding static relative to Microsoft technologies. It has to be said, however, that a survey of the job market does present a pretty clear picture of exactly what is wanted right now, as opposed to projections and anticipated requirements. According to the ads, Java is what the world is using today. Fifteen years ago, few would have predicted that developers would still program in third-generation languages such as the ones discussed here. The brave new world was one of self-generating code and dynamically customisable applications. Such visions remain pushed ever further into the future in the meantime, it looks like Java is set to rule the roost. (Silicon.com will be publishing the results of its Skills Survey 2000 in the week beginning 6 March. Remember to tune in for in-depth text and video analyses of all the key issues facing IT departments this year.) *Menage a trois: Corel, Inprise and a penguin* Corel has announced that it's to merge with Inprise (previously Borland) with the resulting company retaining the name Corel. While we don't know how long these negotiations have been going on, both companies have been setting the scene by putting out compatible marketing messages. The joint press release positions the combined company as a "Linux Powerhouse", and this can be recognised in recent announcements by both companies. The combined product line-up covers the entire development through to deployment lifecycle on the Linux platform in addition to Corel's Office product set. This capability will certainly give new Corel a very commanding product set for those sites that are moving to widescale Linux deployment (as opposed to the deployment of Linux "appliances" such as Web servers that do not impact the rest of the infrastructure). This merger appears to be in the best interests of all concerned. It transforms Corel into an organisation that has "critical mass" in its ability to develop and sell commercial products to the Linux sector, while at the same time there is no external pressure to discontinue the development of non-Linux products or to reduce the service commitment to the "legacy" customers. The remaining unknown factor is the growth of Linux in this general purpose infrastructure sense. Much of the current Linux growth in the corporate sector is based around single function server appliances and the use of Linux as the de facto infrastructure operating system is still in the early stage. It's highly likely Corel will be successful in its level of penetration in this market, but the rate of growth and ultimate size of the market is still unclear. *Straight to the e-point* Lack of information as an excuse for poor decision-making should have gone away years ago, but the glut of information available on the Web gives us an equally difficult problem. We are all expected to base decisions on the most current information available, but there is so much information and it changes so quickly that the activity of staying current can easily absorb all of our time without actually managing to do anything productive. Better by far to be able to acquire first-rate relevant and current information at the time that it is needed. Instead of trying to "read and remember", this is becoming the "information on demand" society. Connextra has built technology that provides additional relevant information using an architecture which, to the best of our knowledge, is unique. Content providers deliver a list of Web domain names, individual Web pages and specific keywords about which they can deliver added-value information. End users download a small Java application (around 120KB) that tracks the use of the Web and identifies when a relevant page is browsed, delivering back the information to the content providers. The information sent back via Connextra is displayed in a small window alongside the browser and can contain hypertext links so that only a short initial description needs to be displayed. As the user drills down through a Web site, the Java application re-examines the pages for keywords and updates the context of the information displayed in the Connextra window. Absolutely nothing has to be done to Web sites to permit the Java application to analyse the content, and because it is keyword driven there is no need for a special multilingual version. The actual value, of course, will be largely dependent on the quality and quantity of information provided by Connextra's partners. In the beta project, the partners consist of a major broadsheet newspaper as well as established informational Web sites. How this builds will depend on the success of the business model proposed by Connextra. As with many Web services, the business challenge will be to attract a critical mass of both end users and content providers. This will require a substantial marketing effort for which the privately-funded company is currently negotiating second-round funding. Connextra is about to start Beta trials with business partners and will launch a limited pilot service in April. The best ideas are often simple, and this particular idea appears to add value to both users and content providers. But it is an idea that could be copied. Connextra will need to attract a critical mass of users and content partners as quickly as possible. For more analysis, see http://www.it-director.com .