LXer Feature: Strategic Alliance of MySQL AB and Oracle for Linux

Competing against Microsoft as a Free or Open Source based entity is a difficult task with many unknowns, however, I have two candidates in mind that could work in concert to deprive Microsoft of its coveted status in one significant area of the business software market. Both are Linux friendly.

The two candidates include MySQL AB and Oracle, where each for their own reasons have good business logic to strike an alliance on either formal or informal terms in their battle against domination by SQL Server. Despite the supposed hope calling for retribution in bogus reporting by one Forbes writer upon MySQL AB for its misdeeds by the instrument of Oracle in Free Software's name, more consistently friendly voices towards Free and Open Software bear no such grudge.

Just for a moment consider those "unknowns" of Free and Open Software business competition. First. the magic formula to compete is still undiscovered (while all perceived missteps do not escape withering scorn). Second the very rules of rational competition are in flux - too much is still unknown. Third even those entities that have conditionally found a method that seems, for the moment, to have given some business success are perceived to be a threat beyond even that of known, larger more dangerous entities.

The latter is true despite other efforts within those "successful" (FOSS) companies that are still attuned to the ideals but go unrecognized. While a certain fraction of this noise originates with the "true" believers, much more is likely from sources preferring dissension in the ranks (see last two references above) and the discouragement of such open competition.

People will make unavoidable mistakes, however, the greater importance is how these companies correct their course. We are in the midst of an exciting time where still no unanimity of opinion exists that such an endeavor even has a long term chance of survival. So expect experiments and expect mistakes - it's the final result that counts.

Simple Logic

The reasoning here is based upon simple principles, i.e. each MySQL AB and Oracle have complimentary strengths, hence, in combination acting as a single unit would prevail in any conflict with Microsoft. The MySQL AB needs Oracle for reasons beyond the latter's new ownership of Innodb, which the former has invested a great deal of its future prospects.

Oracle could benefit in many ways. The most important point: Together they can harass Microsoft in the business database arena from below and above. By such concerted action they have the prospect of denying Microsoft's SQL Server market dominance.

Microsoft's Early Strategy - a Blueprint?

It would be instructive to see some of the techniques MS applied to become dominant in significant areas of personal computer software. The one I observed most directly was MS's push into word processing with Word, which left me cold both when I was using WordPerfect on DOS and more so when on assignment I was forced to use Word. The competitive means employed was to simply be a product that was "good enough" and lower priced.

While it gained adherents it was not until it foisted off the Windows environment onto most pc's that withholding critical details destroyed WordPerfect when it attempted a Windows version. Of course, MS employed other methods: corrupting the technical press and other sundry dirty tricks. I saw a similar pattern with spread sheets, but Lotus aided its demise putting a gun to its own head in a pitched battle with clones.

The important long-term trend to observe was MS's willingness to lose financially while it slowly gained market and position. To a certain extent it also "innovated" by adding user requested features and began to work on commonality of appearance that helped it in the Suite wars.

With the competitive clones removed by legal edict, Microsoft had the low-end to itself as the cheaper, innovator battling the entrenched, leviathan players in several software segments, e.g. WordPerfect, Lotus 1-2-3 and dBase 3.

The Importance of SQL Server

Let's acknowledge that Microsoft uses SQL Server as a strategic product in its arsenal. We should also take notice of MS's efforts to increase its distribution by many different means including lower than expected pricing. For example, nearly full development database engines are available for download that can then be combined with the full database tool set to yield a fairly function test bed.

Moreover, sample versions proliferate on discs that are easily obtained, but have a limited lifetime. Purchase of developer's tools can have a full version included. I strongly suspect that pricing is flexible for important potential customers.

Beyond simple sales regime Microsoft was attempting to integrate SQL Server as a critical part of the Longhorn file system. In essence, this is one of MS's future cash cows to supplement the aging OS and Suite franchises. I am not alone in believing that SQL Server is the proper target for MySQL cited previously.

F/OSS Strategy

The major goal is to deprive Microsoft of easy revenue gains leading to another monopolistic cash cow. While the steps outlined may not destroy MS as a company, they are designed to make MS compete for the majority of their sales. Moreover, its leverage will be reduced regarding pricing even for those predisposed to do everything the Microsoft way.

Loss of flexibility will be a heavy price MS will pay to compete head-to-head with MySQL. and Linux. In most respects SQL Server will matched at a lower price point with matching features, unless MS is prepared to sacrifice its Enterprise version in a low-end battle (see caveats below).

At this time, even a Window's version 5.0 of MySQL is available for free download, hence, the first stage seems to be in place. However, that does not suffice in confronting Microsoft. Marketing and support efforts must be added to draw new users and to understand what might be attractive to some segments. Stress should be put upon the enhanced performance and stability when running MySQL on other OS's.

Impediments

This is the proper point to recognize SQL Server's advantages that must be at least partially countered. These include:

Moreover, it has the further appeal of helping create tables, indexes, and code with color coded hints. The second item as characterized by the Enterprise Manager allows those less knowledgeable about databases to automate various processes such as partition enlargements for data and index checking.

Finally item 3, while I consider Transact-SQL to be inherently superior to Oracle's PL/SQL that was not my first impression nor might it be to any new user. Moreover, Microsoft's insistence on putting its own stamp on this Sybase originated tool has thrown some of the best features gained from Sybase away. However, more to the point many due to the GUI (item 1) may not care.

Caution

I would advise great caution in trying to match the first two particularly on Windows, where MS might pull some dirty tricks that so much effort is totally wasted. Another way MS might try to induce wasted efforts are having bogus users demanding worthless enhancements.

When it seems best to be demure it should be quite explicitly stated that due to MS's past history of dirty tricks against IBM it does not seem wise to expend excess efforts on an operating system that yields lower performance, lower stability and less security readily available on other OS's. Onerous should also be placed upon those most vocal demanding special enhancements to contribute to fulfilling those needs with either cash or usable code.

One other competitive threat from SQL Server exists; that is it has been able to run on Unix and probably could be made to do so under Linux without much effort. Hence, a release onto other platforms without warning is a possibility. However, strong caveats make such a scenario unlikely; that is until a drastic change at the top of Microsoft or competitively, their back has been pushed against a wall - this is an unlikely move. Such a step of: running on Unix or Linux does not mean it functions nearly as well as upon Windows nor could its performance nearly match those designed with Unix/Linux advantages in its base code. Moreover, the timing of such event would be a window into Microsoft's state of inner confidence or lack thereof.

Rapid Time to Market

Oracle could help MySQL immediately, perhaps by some mutual exchange of favors. As part of an agreement between the pair to continue licensing Immodb as a reasonable rate (including continued GPL on the base code) the funding could be channeled into a campaign to educate Window's users to the advantages of MySQL on that platform and others.

Perhaps too, if the serious users of MySQL desired a nearly or completely free of cost version, they might be given copies and updates just by registering. Oracle might be willing to bear some of those costs, because it makes market share of a freely available product more measurable, hence, more accurately gauging the success of various marketing tactics and likely market penetration. The latter would have monetary value to both partners, hence, could justify lowered support charges to none for some.

Playing further up the scale on Windows is a fools' game, but there are advantages to trying to be competitive even on MS's turf. Probably can charge higher rate for the support option buyers and if MS lowers their pricing it is still a win where subscribers could be urged to change OS to gain fuller advantage of the database system at lower pricing. Even where the competition seems hopeless it would not be unwise to plant the idea that any competitor is at a disadvantage on Windows because Microsoft refuses to compete fairly.

Going Forward

Those steps outlined above pretty well outline an attack from below that MS will not be able to counter with the effectiveness they usually expect. Their resources will be drained to fully fight the battle at the low-end and pursuing that approach could negate their long term prospect of making SQL Server as part of the a current triumvirate of cash cows to be milked in pursuit of a new exponential growth spurt they so ardently desire.

On the contrary they may decide as others that end of the market does not deserve their attention, because the larger monetary return resides in beating Oracle at the high-end and running all high-end database users on Windows. The latter is their more probable response, with a rear guard attack in a display of contempt for both the low-end product and those opting for it.

Then the question becomes: must Oracle and other highend database purveyors continue to retreat ever higher and into vertical markets to avoid sure losses against the rising bottom feeders. Well, in the case of SQL Server the answer is a resounding No!

The problem Oracle might have is choosing the best option to confront Microsoft from above in the database market. Perhaps the simplest approach would to use an older and/or trimmed version of Oracle, which should be offered on any OS but not Windows. In my personal opinion, this does not allow Oracle the freedom of action another less costly approach might return.

The Focus

I am going to concentrate on the latter course, because Oracle is better equipped than I on how to decide upon what components would be most suitable. Moreover, making MySQL Premium would allow less resources to be expended and allow greater sense of fun when marketing this new product.

One must understand too that MySQL Premium will not be a static product. It would probably begin with some proprietary addons from Oracle that would enhance performance or processing. Some might be later moved into the Immodb as open source under GPL that would still deprive Microsoft the option of using the technique.

To me one of the marketing ploys Oracle has used to tweak Microsoft's image was to offer a large sum of cash just by having SQL Server perform at a fraction of what the Oracle database could perform. Nonetheless, the terms were such that it was a near impossibility that MS could have ever collected whatever its performance might have been.

Additionally, the taunt had the sought after effect of making MS seem risk adverse. Here the jokers on Oracle's marketing can demand that SQL Server outperform this light weight database that MS would refrain from even considering a worthy competitor.

The challenge should include that the benchmark be verified by an independent panel and performed upon any OS but Windows (listing AIX, HP Unix, Solaris, BSD's, Linux, etc.). For Microsoft this is nothing but a lose - lose proposition and the taunting need not be short lived.

An alignment as described above would be of obvious benefit to both players to Microsoft's disadvantage. It would increase the competitive challenge faced by the latter and probably deny it a win of a new cash cow in the RDBMS arena. MySQL would gain a formidable partner with deep pockets and Oracle could actually gain financially in proportion to Microsoft's losses and in a more subtle fashion gain reputation by its support of a free database tool it owns.

This is not a prescription to destroy Microsoft, since it has too many financial resources as to be easily dispensed with, however, should it not learn to be a more honest competitor such a scenario played out by others could slowly bleed it to irrelavence.

After Word: If Microsoft's complete destruction is the only goal that will satisfy some individual's blood lust it will not be found above nor in this short closing comment. Moreover, this is a business plan designed to fluster, to frustrate and just perhaps at times drive MS into fits of frenzy. However, this is not a plan for a Mafia like entity like Murder Incorporated – that is more Microsoft's style not mine. My hope is two competitors can align their interests to confront a more dangerous enemy of Free and Open Source Software.