You cannot fail to be impressed with Microsoft's all-encompassing IT offering and seductive marketing. I'd like to share some thoughts on how Microsoft has got to where it is now, and to prepare you for the next level onslaught from Redmond, because believe me, what's brewing around COM, XML and SQL Server will affect us all.

Microsoft's strong starting position is not so much the OS as the application-end of the business. It has been by providing massively functional applications and selling them cheap that Microsoft has achieved dominance. Once everyone in the organization is on Office, the choice of NT follows. But not yet that of SQL Server. So how is Microsoft going to rein-in that segment of the market? By offering massively functional database applications just like they did with the word processor before.

Don't believe me? How about being able to perform 'region growing' and other sophisticated statistics on data in your production database from your desktop? But I'm getting ahead of myself, I'd like to walk you through three high-impact solutions as seen from the key end-user standpoint.

Do you remember how the database vendors sold us SQL a while back by saying that we all needed ad-hoc queries? We did and we probably still do, but how many ad-hoc queries have you performed today? My guess is none because using SQL to query a moderately complex database is like "kicking a whale across the beach with your bare feet". Rather than spending three weeks (a real-world figure) writing one SQL query, the trend is to have a separate 'data cube' database, with data extracted from the production database and pre-arranged so that it can be attacked by On-Line Analytical Processing (OLAP). Now Microsoft has not invented OLAP, but BG didn't invent the OS, the word processor or the spreadsheet. They just made all this technology accessible and cheap. This is what is about to happen with OLAP. Instead of spending $50,000 on some obscure software, you will be soon have 'in-your-face' region-growing and data mining, right out of Office.

Another potential killer application is data loading and synchronicity. You know the problem, multiple databases, multiple data-entry points and ensuing chaos! Enter Microsoft Transaction Server – providing a cheap, configurable link to all your organizations databases. Pop a Visual Basic ‘business logic’ layer on top of that, and you can ensure that one validated x-y coordinate is transformed and replicated across the organization.

But my main take-home concept from the show is Microsoft’s enthusiasm for XML as the preferred route to interoperability. Of course XML is intended to be a CORBA and Java Beans killer, but the spin-off, unless the technology is hijacked, is a credible route to interoperability that avoids the old problem of standards that only work if everyone implements them in exactly the same way. Next month’s PDM will look at XML in depth, in fact if you would like to contribute something, feel free. Meanwhile, you’ll fine lots more on TechEd on pages 6 & 7 in this issue and also an exclusive interview with Microsoft’s industry architect James Utzschneider.