To print: Click here or Select File and then Print from your browser's menu

This story was printed from silicon.com, located at http://www.silicon.com/

Story URL: http://comment.silicon.com/0,39024711,10004214,00.htm

The Bloor Perspective: Ballmer on Linux, IBM's dinosaur and SQL's future
This week Robin Bloor and his team of analysts consider Microsoft's approach to the open source threat, IBM's faith in mainframes and the pros and cons of SQL databases...

By Bloor Research

Published: Monday 19 May 2003

Recently Microsoft held a meeting in Berlin with leading European industry analysts to discuss Linux and Open Source Software (OSS). The day concluded with Steve Ballmer, CEO of the Redmond headquartered giant, discussing Microsoft's position in the world.

The discussions commenced with Microsoft acknowledging that Linux/OSS is now appearing in a number of different segments. To kick things off the company acknowledged that in the past it has lost credibility by criticising the Linux core (kernel and web server stack). It accepts that this core now possesses credibility in the commercial world.

Broadly speaking, Microsoft sees the areas where an impact has been felt as 'community' software (simple file and print services, email and firewall/cache), 'commercial quality Community'-based software (Linux kernel, networking, web serving) and true 'commercial' software (ISV applications and databases such as Oracle and DB2).

The company believes that Linux solutions appeal to different organisations for different reasons. For some, especially price-sensitive SMB organisations, the low cost of purchase and its ability to function on comparatively low cost hardware platforms appear to be major attractions. For others it may be the ability to configure OSS in a bespoke fashion that is the prime attraction.

When discussion turned to which platforms Linux is currently gaining share from, opinions were divided. Microsoft believes that the major movers behind Linux (in its opinion IBM and Intel) are using the OSS movement as a means to disrupt the Unix market and that Unix systems comprise the major source of movement. Others present proposed that organisations are migrating from both Unix and Windows to Linux.

The Microsoft CEO was quite candid when he was asked about the current attention that the company is now paying to Linux when compared to its former stance of apparent unconcern. Ballmer simply smiled and responded: "We were neither unconcerned nor unaware of Linux. We just tried not to show it!"

Ballmer went on to state that it was obvious that the advent of Linux and OSS means that Microsoft can no longer rely on the "we are cheaper to acquire" message to win and retain customers. In recognising this the company is now focusing much attention on its belief that its solutions provide a much better TCO case than those currently available using Linux-based tools.

He also stressed that 'innovation' and the creation of complete solutions rather than simple software functionality would hold the key to future success.

Ballmer once again made it clear that Microsoft had absolutely no plans to port any of its substantial software library to Linux. Apart from concerns centred on the possible impact that any accidental inclusion of GPL code could have on the integrity of Microsoft's own intellectual property, he felt that it could be counterproductive to the development of its products.

The day indicated that Microsoft is now happy to recognise that the influence of Linux is growing. It is clear that we can now expect Microsoft to attempt to build its case for Windows as an operating system based on rational arguments rather than a simple dismissal.

It will be interesting to see how the company sets about promoting its qualities versus those of open source. It will be even more fascinating to monitor the response of customers, OSS advocates and the world at large.

*Mainframes far from extinct*

T-Rex is IBM's tongue-in-cheek code name for its latest mainframe model, the z990. It was once suggested that mainframes were dinosaurs and due for extinction. They were rescued by two factors.

First, for the very large organisations that deploy them, IBM mainframes represent very economic computing. They always did because the cost per user was lower than any of the alternatives. Nevertheless the sales began to collapse in the early 1990s mainly because few new applications were being written for the mainframe as it was not a good vehicle for the then-hot client/server world.

The second factor and the one that brought the mainframe back to life was Linux. Implemented in a virtual machine environment on the mainframe, Linux proves to be very economical 'per instance' and cheaper to configure and run than on any other platform.

If you add the manageability factor and the small footprint factor, the mainframe is actually a natural Linux environment - the equivalent of a large Linux server farm. IBM claims that about 20 per cent of mainframe sales are now Linux-based and a good number of them are consolidation projects where Unix and Linux servers are replaced by a single more economical box.

IBM never lost faith in its premier hardware platform and continued to invest even in the thin years. The T-Rex range itself is the culmination of a four-year $1bn plus project involving 1,200 IBM developers. It has almost three times the capacity of the z900 with double the number of Channel Path Identifiers (CHPIDs), double the number of logical partitions (LPARs) and four times the number of HiperSockets (Internal LANs).

Performance improvements have been added for Linux and z/VM (the virtual machine environment) and IBM claims it is capable of more than 11,000 SSL handshakes per second. It's a big beast. In IBM's words it "roughly equates to an entire data centre on one server the size of a refrigerator" - an American refrigerator of course but let's not be picky.

IBM intends T-Rex to be a major vehicle for delivering its On Demand message. It will serve as an engine in IBM's On Demand Data Centers where IBM customers of all sizes can buy power on a hosted pay-per-use basis. The service offered will include processors, storage and standardised middleware. The fact that IBM is chewing its own gum here suggests that the economics of z990 are attractive.

So the dinosaurs are truly back and now at last they have a dinosaur's name.

*Bye-bye SQL?*

Although it may seem a no-brainer today, SQL took some time to get established with by no means all vendors subscribing to the notion of a common database query language. That, remember, is what SQL was originally designed to do - provide set-based query processing - the fact that it was soon appreciated that it could be used for other purposes, such as writing stored procedures, was a bonus.

However, by the early 1990s SQL was pretty much adopted by everyone (or, at least, the relational vendors - there were still plenty of pre-relational databases going strong) but there were lots of proprietary extensions to SQL. However, by the middle of the decade there was a real sense that new versions of the ANSI standard would subsume most of the differences between the different versions, and we would actually get a genuine SQL standard that was widely deployed and only minimally extended.

And then Informix bought Illustra and declared that it was introducing the Universal Database that would store everything. One consequence of that announcement was that companies like Oracle and IBM introduced extensive object capabilities into their databases. As a result they extended the SQL they used, in order to support the new capabilities of the database.

To cut a long story short, then came the internet and OLAP, followed by XML and, more recently, data mining and ETL, all embedded within the database. Now, leaving aside any questions of whether it is sensible to manage all of these diverse things from one place, these all involved further extensions to SQL.

For example, IBM has introduced more than 100 extensions to SQL just to support XML. And it is not limited to generic sets of extensions. There are also vendors including SQL extensions for specialised functions that SQL can't handle, such as time-lapsed queries (try asking a question such as "Who bought a barbecue within seven days of purchasing patio furniture?").

The question is: If companies are introducing hundreds of proprietary extensions to SQL has it actually gone so far away from the standard that to talk about it as if it were a common language is absurd?

The short answer to this question will depend on how you look at it. For all practical purposes, we should now be thinking about SQL as a language with a standard kernel and potentially unlimited extensions. However, the market has a vested interest in making us believe that SQL still remains a truly portable language. It is not or, at least, only at the kernel level.

There are some signs that suppliers are moving away from SQL. In particular, it is no longer the only language you can use for writing stored procedures. Vendors like Oracle and FirstSQL allow you to use Java as an alternative.

But there is no such movement away from using SQL for query processing. Yet the language is clearly inadequate, not only because of the extent of the dialects that are supported but also in the way that it allows the execution of very poorly formed code (which is why database optimisers have SQL rewrite capability).

Arguably this is a good thing - it allows novice programmers to still create useful queries - but it leads to bad habits and can hardly be considered a sound premise.

The fact is that we won't be saying goodbye to SQL any time soon but it is surely becoming over-bloated and it is not the panacea that the marketing departments of some suppliers would have us believe. Companies making buying decisions need to recognise that large numbers of SQL extensions make SQL a lock-in just as much as proprietary operating systems or hardware.

Bloor Research is a leading independent analyst organisation in Europe. You can find out more at www.bloor-research.com or by emailing mail@bloor-research.com.