Oracle picked up a trail of SAP in hunting for Open Source

Consecutive acquisition of Innobase and Sleepycat companies by Oracle corporation, attempt to purchase MySQL and also plans to purchase JBoss and Zend Technologies drew attention, both a business community, and community Open Source. On the one hand, it is curious as the corporation is going to use the resources attracted with acquisition, on the other hand potential impact of similar acquisitions on development of projects with the open code and communities Open Source in general attracts interest (and even concern).

Concerning InnoDB use (development of Innobase) and BerkleyDB (development of Sleepycat) the answer was given by Oracle – acquisition of the specified two companies is in the strategic plan of strengthening of positions of corporation in the market of the built-in DBMS. Obviously, the planned acquisition of the developer of solutions for J2EE (JBoss) and the developer of a core of the interpreter and corporate assets of development for PHP (Zend Technologies) should complete functionally creation of a skeleton of future platform for development of the built-in solutions of Oracle. Confirmation of the read plans of corporation is also opening of specialized research centers in India and Southeast Asia.

At the same time, there is an opinion (stated by representatives of community Open Source) that acquisition of Innobase and Sleepycat pursues other aim, namely – blow to MySQL, as on the competitor in the market of DBMS. Despite explicit incompatibility of "weight categories" of Oracle and MySQL, such assumption has under itself the reasons. The brightest to that confirmation – unfortunate attempt of acquisition of MySQL. Therefore, Oracle tries to take the developer of the most popular open DBMS under the control. According to us, to that there can be three reasons: �  elimination of the competitor in the market of SMB and the market of the built-in DBMS, �  cost reduction at entry into these markets due to use of already ready-made and popular solution, �  establishment of control over development and promotion of MaxDB.

Not only DBMS

Large players in the markets of DBMS (and CIS) are actively positioned in the market of solutions for medium and small business now, mastering it after relative saturation of the sector of large business, and Oracle – not an exception. A main competitor is Microsoft offering the SQL Server of 2005 + .NET platform which is highly integrated and at cost much more available than "heavy" solutions of Oracle here. And here the main competition to products of Microsoft is made by the open LAMP platform (Linux + Apache + MySQL + PHP) which though has no such high integration scale, is cheaper than solutions of Microsoft, is very popular and also is much more democratic in respect of license policy. Attempts of Oracle to take control of two last LAMP components are also directed to transformation of this fact into the trump.

The second reason of interest of corporation in ready-made solutions with the open code – a possibility of their use for creation of own decisions without need to incur a number of expenses, occasionally considerable. First of all, InnoDB and BerkleyDB, as well as MySQL, are the mature, well worked products. On writing of their analogs by own efforts it should incur considerably big monetary (and, above all – temporary) costs, than on acquisition of the companies together with their practices. Secondly, at an output of a new product to the market costs for its advertizing, promotion and promotion, costs of time for acquisition of experience of implementations and gaining trust of users are inevitable. All this is not required for the purchased products as they for a long time are known and widespread, there is a whole community ready to share the knowledge and experience in the field of their application and implementation there is a wide experience of implementation of these solutions. The corporation got all this "non-paid, i.e. for nothing".

At last, not such obvious, as two previous, the reason – local interception of an initiative at SAP in the development plan for MaxDB which was SAP DB (it is Adabas D). The matter is that Oracle and SAP have difficult relations not only in the market of ERP – since 2003 SAP through MySQL extends MaxDB under the license GPL and positions it as the certified DBMS for ERP-SAP solutions. Certainly, total ownership cost the ERP system with MaxDB as a core is much lower than the cost of a similar system on Oracle DB kernel. And again the tempting market of SMB for which such difference in cost is basic steps on the stage. Besides, the license GPL does not impose any restrictions for the nature of use of DBMS, the number of installations and number of the used processors that is undoubted advantage from the point of view of the customer. It is represented to Oracle which is quite proved attempt to put it an end and, at least, to attribute the name to the MaxDB brand.

Achilles' heel

Having dealt with the purposes of corporation, we will try to understand how means of their achievement will be reflected in the involved projects and community Open Source in general.

The "loudest" reaction to the last actions of Oracle, likely, is Bruce Perens's opinion expressed about it. Its essence is that the Oracle corporation did not understand that the Open Source project cannot be purchased and wasted the millions. The author of determination of Open Source also showed joy that projects with the open code gained such strength now that even corporations were frightened of their competition.

At the same time, in comments on Perens's opinion it was right there specified that both MySQL, and InnoDB and BerkleyDB are delivered by their developers according to the scheme of dual licensing: on commercial terms and on the terms of Open Source (MySQL – GPL, InnoDB and BerkleyDB – BSDL); and that it is not necessary to confuse "Open Source Release" and "Open Source Project". Commercial distribution of MySQL with InnoDB and BerkleyDB just also is MySQL business model "Achilles' heel".

Actually, hardly Oracle can reproach with the naive relation to Open Source, the corporation did not show "monetary interest" in Linux or Apache (the first two LAMP components) which are really open projects impregnable for any attempts of external control. On the contrary, the corporation concentrated attention on "weak links" – the developers gaining income from commercial implementation and support of the products, in parallel extended and as Open Source. Such practice is quite often applied "not quite free" by developers of the open source as a business model. And if the free project cannot really be purchased, then "quasifree" – quite.

Purchases of such project will implicate not necessarily catastrophic effects for its open part, but, undoubtedly, will lead to considerable deceleration of its development. How Eric Raymond eulogized "market model" of software development, collectives most effectively work with centralized operation by development (in case of development of a kernel of Linux all process is controlled by Torvalds almost solely, development of Apache, Gnome, KDE is controlled by rather closed core of developers). Besides, in case of InnoDB, BerkleyDB and MySQL the most qualified developers of these products are a staff of the companies supporting these projects. Respectively, acquisition of such company at once deprives community Open Source of a core of group of developers of the project.

"Insubmersibility" of the Open Source project directly follows from the name – its source code is open. Therefore until there are enthusiasts and persons who are financially interested in development of the project, the project will live. In this sense it is necessary to agree with Bruce Perens. At the same time, it is worth to remember that the serious program project contains not one thousand code lines (and not always – documentation) which to understand quite not easy. Therefore if, let us assume, Oracle after all purchases MySQL, on creation full "forka", even in the opportunity, – inclusion in a game of the new company which will undertake further support of the project, – will leave not one month. During this time, until the community recovers the project course, the corporation with the resources will manage to make much more. Most likely, in this situation of Oracle will be able to produce integrated solution and to bring it to the market. At the same time both own reputation, and reputation of acquired company will work for her that should affect the decisions made by potential users significantly. Ideology – the last that concerns them.

Temporarily we are not on sale

Fortunately for community Open Source, MySQL is not going "change freedom for money yet" (to be exact, to lose an active position in the perspective market) and already found a quite good way out of the situation – not only to replace the DBMS components depending now on Oracle, but also most to purchase the developer of a component replacement which became FireBird DBMS. On the one hand, such approach will allow to resolve tactically developed dangerous situation, with another – does not eliminate strategic threat of acquisition of the project, moreover, strengthens possible negative effects as facilitates to potential "aggressor" a problem of merger of all developers entering the project.

However, most likely, end users anyway have nothing to be afraid – even if corporations will manage "win against freedom", it will not be profitable simple to them to set on the updated products the high price as because of it they and had to refuse market promotion of SMB of own developments and to look for cheaper solutions.

Sergey Sereda