Ever since Oracle announced the acquisition of Sun Microsystems along with MySQL, all hell broke loose in the open source community. With EU questioning the deal, there is a war (of words) erupting inside the community with one side asking EU to block the deal or, at the very least, change the license to another open source license from GPL and the other side urging EU to allow the transaction to go through. Even though I have no love for Oracle, I think it is time to let the deal go through at least for the sake of Sun employees who are sitting there with their future unknown. At the same time, I am not unduly worried about the future of MySQL because I have complete confidence in the open source license of MySQL. Let me try to explain my position here in this post.
For the sake of argument, let us consider the hypothetical scenario of Oracle killing off MySQL. This leaves us with the only option of going with one or more of the MySQL forks. Such a scenario is perfectly fine for most of us except a handful of people who are either packaging MySQL into a proprietary software or keen on building a business making money from a dual licensing scheme like the one that MySQL uses now. If you look at how MySQL is being used in the world, it is easy to see that only a small percentage of users are affected by Oracle killing MySQL. Now, if you factor in the odds of Oracle resorting to such an action and the associated PR impact, you are left with only 2-3 people getting affected. They are the ones who are planning to run a business using a dual licensing scheme on a MySQL fork. I can assure you that these 2-3 people are quite capable of taking care of themselves and we need not waste EU’s taxpayer money and our valuable time fighting for them. I want to emphasize once again that the only business model that will get affected in the case of Oracle killing MySQL is the dual licensing scheme. Everything else, including the hugely successful support services model, will continue to thrive.
Now I am going to address the “Cloud Computing” part of the title. Richard Stallman and some of the free software evangelists dismiss Cloud Computing as an attempt to push vendor lock-in through the backdoor. There are others like Tim O’ Reilly who advocate the line that we should not worry about licensing and, instead, focus on ensuring open architectures and standards. On the other hand, I have argued many times in this blog about the importance of open source from a moral and strategic point of view. However, as noted by the industry observers including Matt Asay and Index Ventures general partner Bernard Dallé, cloud computing may turn out to be the best way to monetize the open source software in the coming decade.
If you are wondering how this argument fits into the MySQL show, I would like to argue that cloud computing will ensure that (forks of) MySQL will continue to live and serve the needs of the users as before. Here is my line of argument but feel free to poke holes into it (it will help me tweak my understanding of the open source marketplace). Already big cloud infrastructure players are offering cloud based services relying on MySQL. Amazon has started offering MySQL like capabilities in the name of Amazon RDS (note the absence of the term MySQL in the name), Joyent is offering MySQL accelerator and Rackspace is partnering with FathomDB to offer their own MySQL offering. Of the above three examples, Amazon and Rackspace jumped into the game after the announcement of Oracle-Sun deal. These players are not there because they are trying to squeeze out as much revenue as possible before MySQL goes under the water. They are there for a long haul which implies that they will be forking MySQL or supporting one of the forks. Let us not forget that Rackspace has committed resources to Cassandra project because they have plans to offer an open cloud alternative to Amazon’s SimpleDB and Google’s datastore. I am pretty sure Rackspace will spend some of their resources on a MySQL fork to keep the project going. There is a good chance that Amazon might also support the project even though they have the ability to absorb many of the current MySQL employees and keep the development for their in-house usage (Thanks to GPL’s still existing SaaS loophole). In fact, these cloud vendors may even end up supporting a neutral MySQL fork and, then, use their in-house expertise to differentiate themselves from the competitors. With more and more cloud vendors offering some form of service based on MySQL, the longevity of MySQL increases. These vendors, in their own self interest, will ensure that a fork of MySQL live as long as people use relational databases.
I think it is time for some sanity to prevail in the community and let Oracle absorb Sun and MySQL. The very nature of open source will ensure that users are never left in the lurch. MySQL and any other open source software absorbed by proprietary vendors in the future will survive irrespective of what the new owner does to the OS software they buy. Along with other factors, cloud computing will also help them survive.