The launch by Microsoft Corp with enthusiastic support from Ashton-Tate Corp of an SQL database product (CI No 847) is seen as the first shot in a battle which aims to pre-empt the forthcoming challenge from IBM’s database product for its OS/2 offering. But the alliance between Ashton-Tate and Microsoft has raised the spectre of a deepening divide between MS-OS/2 and IBM OS/2 and future developments threaten to push the two systems further apart rather than closer together. Ashton-Tate UK managing director Paul Sloane plays down the fact that the SQL Server venture is a direct challenge to IBM. What this alliance and the SQL represent is a product which will make micro-based products more competitive than databases on minicomputers, he declares. Crucial victory But he concedes that the SQL Server venture will also widen the schism with IBM. This means two proprietary standards will be unavoidable in the OS/2 arena. That’s a pretty good guess. I think Ashton Tate and Microsoft have recognised the need to get together to reach a critical mass to ensure our product becomes the new standard, he says. The approach these two companies have taken seems simple enough. By unveiling their product first they hope to establish it as the standard in the database arena thereby scoring a crucial victory for Ashton-Tate which wants to safeguard its lead in the database market against the coming IBM database product. And if users accept the SQL Server then they will have to take MS-OS/2 rather than IBM OS/2 and this is good news for Microsoft. Hewlett-Packard UK marketing director John Golding agrees with that: We are all trying to work toward an industry standard but we want to influence what that standard is going to be. The trick is to get in first and create the standard, he says. But as the battle lines are drawn the question arises: can the software makers really expect to withstand an IBM onslaught? The answer is probably yes. If one listens carefully to what IBM has been saying, the most important role for the Personal System/2 is to populate every corporate desk-top currently occupied by a 3270 display. That IBM’s OS/2 Extended seems like an attempt by the company to evolve its own proprietary operating system into which users will gradually be locked is therefore no surprise. The big question is whether there is good reason for any user who does not spend much of his time talking to an IBM mainframe buying PS/2 rather than any of the myriad cheaper clones that will run OS/2 whether or not they are Micro Channel compatible. The mutually beneficial alliance between IBM and Ashton-Tate – which provided the former with good database products for the IBM Personal range and turned Ashton Tate into a major applications house – is also under threat. Ashton-Tate will sell the SQL Server under licence from Microsoft thus risking incurring the wrath of IBM by ensuring that existing dBase users follow the SQL server route. But then the whole drift of the IBM approach to database on the PS/2 is not to provide a super stand-alone server solution for clusters of networked PS/2s, it is rather to provide an ideal environment for drawing down data from DB2 relational databases on the mainframe. And Sloane plays down the threat to Ashton-Tate from IBM. Tough time He argues that the hardware giant has not proved adept at software sales in the past. So IBM will have a tough time trying to combat the combined expertise of its Ashton-Tate and Microsoft in that field, he reckons. Jan Feaster, managing director of rival database software developer Migent Inc takes a contra view. She thinks that Sloane has underestimated IBM. Who will win the showdown? I would put my money on IBM. It is very good at everything it does, including system sales – which is what we are talking about. Whether the software vendors can fend off a challenge from IBM will also depend on how much backing they can get from other software houses. The SQL Server has been made with the help of SQL interface technology from Sybase Inc. One industry watcher reckons software make
rs will hedge their bets and write software which runs under both the Microsoft and IBM version of OS/2. When you load the software, it will ask you whether you are using MS OS/2 or IBM OS/2. Software companies can’t really ignore either market because both have too many users. According to Ms Feaster, Migent will definitely be making a product that will run under IBM’s version of OS/2. We have got to play ball on its patch, she says. She is not sure how Migent will respond to the Ashton Tate/Microsoft move. Other software makers are likely to take the same view. The SQL Server deal also has wider implications for the market. According to Sloane, it will continue the transfer of power from minis to micros in database management technology. On a broader scale this has been a feature of the computing scene since the arrival of the IBM Personal in the early 1980s. But more importantly perhaps, the SQL Server venture also heralds the beginning of closer co-operation between software giants. This isn’t the only collaboration between Ashton-Tate and Microsoft. We are working on products around the SQL Server such as tool kits and additional utilities, Sloane notes. An immediate problem for software developers is that the schism between the software giants and IBM means they have to spend more effort and resources to develop products that can straddle the gap between the two systems. Sloane says that Ashton-Tate has already got some software makers working on products for its version of OS/2. No doubt IBM has for its forthcoming offering too. Divorce courts Some also see the Microsoft challenge as heralding a session in the divorce courts with IBM – and without the guidance of recent history, that would look like a dead certainty. But Microsoft has tweaked the lion’s tail and got away with it so often in the past that the case for saying that this time it will be different is difficult to prove – especially as, two years ago, IBM put its relationship with Microsoft on a more formal footing that it had been previously. Another reason why IBM may not offer anything approaching the attractions of the SQL Server is that if it came out with a comparable product, it would threaten sales not only of the 9370 but even of Silverlake: System 36 users may well find that a network of 80286 micros with an 80386 server are a much cheaper and more attractive proposition that a move to Silverlake, and from IBM’s point of view that 80386 chip is just too damn powerful. Don’t say it too loudly in IBM’s hearing, but the 9370 addresses 16Mb real and virtual memory, the 80386 addresses 4Gb. Which, on that basis, looks like the future? But which does IBM want to encourage users to buy?