Recently, I met with members of Microsoft’s Platform Strategy Group, including the group’s director, Robert Duffner, to talk about their company’s activities around—and evolving stance toward—open-source software.
After assuming an initially hostile position toward open source, Microsoft has adopted sort of an experimental approach—the company developed a pair of bona fide open-source software licenses, maintains an open-source project repository at codeplex.com and is working at making popular open-source components work with the Microsoft product family.
For instance, Duffner cited Microsoft’s work toward embracing PHP—the “P” in the popular open-source LAMP stack of Linux, Apache, MySQL and PHP—by tackling the barriers to use of the Microsoft-alternative WISP (Windows, IIS, SQL Server, PHP) stack. To this end, Microsoft has been optimizing PHP to run well on Windows and helping adapt in-demand open-source applications that have been hard-coded to use the popular MySQL database to work with Microsoft’s SQL Server.
These efforts will certainly bear dividends for Microsoft’s PHP-loving stakeholders, but no amount of integration and optimization work will address the primary barrier to a LAMP-like ascent for WISP: open source.
Open-source development and distribution will continue to expand because between functionally similar open-source and proprietary platforms, the open options offer developers and architects a fundamentally better deal: You own the code and are free to spin up a thousand copies of MySQL with no more licensing burden as a single instance.
For now, Microsoft is operating, tentatively, at the margins of open source, with an agenda marked more by toleration than adoption of open source. I contend that if Microsoft approached open source aggressively, the company could solidify its prominence in computing potentially for decades to come.
In many cases, the best option for a Microsoft customer will be MySQL. MySQL runs on Windows, so Microsoft might do well to extend its embrace beyond PHP to include MySQL. After all, if certain customers are destined to choose an open-source alternative to SQL Server, Microsoft might as well make money on it and maintain the customer relationship.
If Microsoft chose to become a MySQL distributor, Microsoft’s deep roots in the market, along with its support, services, learning, events and channel resources, could enable the company to extract more money from MySQL than anyone else, including Sun Microsystems.
Alternatively, Microsoft could address the MySQL challenge not by joining, but by beating MySQL at its open-source game. An open-source-minded Microsoft could release SQL Server as open source. A project like this, which I’ll call OpenSQL, would be no small effort, but Microsoft has the examples set during past open-sourcing efforts, such as those of Netscape’s Navigator to Mozilla and of Sun’s Solaris to OpenSolaris migrations, to help guide the way.
Also challenging would be the business model changes required by an OpenSQL effort, but I believe that a well-executed OpenSQL effort would deliver a healthy market share boost for Microsoft and its channel partners to tap with their service and support offerings.
The move would be a major boon for the company’s developer ecosystem, the members of which would be free to weave Microsoft’s database product into more of their applications, and focus their resources higher up the stack.
Speaking of moving up the stack, we could move another letter to the left and make a similar set of arguments regarding IIS, Apache and what a hypothetical OpenIIS could do to solidify the “I” in WISP.
As for the W, that will have to wait until my next column.