Follow me on TwitterUnless you live under a rock you have by now heard that Oracle will be grabbing SUN. And if you care about databases, especially free databases, you know that SUN bought MySQL for a cool billion US$ not too long ago. Oracle is best known for its Oracle database and Oracle database is the largest contributor of revenue for Oracle as a company. MySQL is the most popular database product used by everyone from a pre-teen aspiring to be a hacker to the largest Internet properties like Facebook, Yahoo and Google. MySQL value proposition has never been its technology. As a matter of fact, technology is something that MySQL is quite light on. No, the entire value prop of MySQL is best expressed in the bold claim on the MySQL Enterprise Unlimited website.
Deploy an unlimited number of MySQL Enterprise Servers for the price of a single CPU of Oracle Enterprise Edition
Take that Larry Ellison. There have been a number of industry watchers putting on a brave face and musing about a possibility that MySQL will somehow thrive as part of Oracle. I have no idea what drugs they are taking but they seem to be hallucinogenic. MySQL within Oracle is as good as dead. No, it will not be a quick and painless death of a press release announcing that MySQL is no more. It will be a torture of many years of MySQL of being a “walking wounded”. Publicly Oracle will profess its love for open source and it desire to leave MySQL alone to help it conquer the hearts and minds of the masses. Oracle will use examples of InnoDB (which it acquired to control MySQL oxigen suplly in the first place) and BerkleyDB (Sleepycat). The reality will be very different. One does not have to go very far to see what happens to free products at Oracle. Oracle XE is a free DBMS that Oracle put out there to get tractrion in the community. That was 3 years ago. It has not been updated or patched since despite hundreds of known security vulnerabilities in Oracle 10g R2 code base that Oracle Express Edition is based on. I expect that the future of MySQL in Oracle will not be much different. Progressiv slow down of new development, deteriorating quality of product support etc.
If Oracle does not sell MySQL to someone else, MySQL is as good as dead. It will slowly but surely loose steam with departure of key MySQL developers (many have left already). The pace of work on MySQL within Oracle will slow down to a crawl. At the same time, there will be friction with the true stewards of MySQL in the open source comunity. I just don’t see the open source folks who made it their philosophy to fight Oracle’s exhoirberant software license charges all of a sudden getting cosy with the enemy. Those that leave Oracle (and most of the best will leave because they are true believers of the ideals of the open source) may join one of manyMySQL forking efforts that are already under way. However, I don’t believe that open source MySQL will thrive either. First Sun acquisition of MySQL and now Oracle acqusition, have fragmented the community and set the MySQL progress back years. It is possible that someone like Monty Widenius (founder and original developer of MySQL) will rally and unite the MySQL community and wrestle effective control out of Oracle’s hands. It will be a difficult thing to do, even for someone with Monty’s passion and determination. The exodus of MySQL talent has started long ago with SUN’s acquisition of MySQL and many of the key people have now dispersed and will be hard to bring them back even if Monty is prepared to hire. I think he realizes this; you can almost hear him say it in his blog.
The pundits have pronounced that Oracle acquisition of SUN is an industry changing event. I have no doubt that it is. However, for the world of databases this is not a change for the better. MySQL has always been an important player in this world applying significant pressure on commercial DBMS vendors like Oracle, IBM and Microsoft. I think this era is over. I don’t believe that MySQL will ever play the same role it once did. Maybe PostgreSQL will step in and fill the void. I know IBM will continue to offer DB2 Express-C free of charge and will continue to provide an alternative to the cripled Oracle XE product. I am just sad to see that MySQL will not be joining the fight as its future is getting more and more uncertain.
Dont worry. Companies depending on mysql will not let it fall.
I am certain that MySQL will survive in one form or another. I know that Monty, Percona, Brian Aker and others will do their best to keep it going. The problem is that all of these efforts are very fragmented on their own and they fragment MySQL. Googles, LinkedIns and Facebooks of the world are OK with that but for the rest of the world, I just don’t think MySQL will continue to play the role it ones played.
Unfortunately, I think Leon is right. MySQL doesn’t really fit in with Oracle’s strategy of going after the high end of the market. Oracle wants you to spend alot of money on their products, and doesn’t really seem to be much interested in the lower end. Any effort at this level is pretty much for PR, and that’s about it.
I have never been much of a fan of MySQL, but I did appreciate the pressure they applied to Oracle, to remove them from the low end of the webspace, and this in turn relieved the startup costs for many web entrepreneurs. I am guessing this blog is running on WordPress, and WordPress typically uses a MySQL backend. So many web apps have seamlessly integrated into MySQL, that it has been a breeze to administer.
Oracle has a history of gutting competitors, and not through technical superiority. For example, in the late 1990s, they raided the talent at Informix and snatched away a core group of Informix’s finest. Informix, in my opinion, was the finest database around, more stable, scalable, and way easier to use than Oracle. Informix responded with a lawsuit, which nearly killed the company. Luckily, IBM picked them up and have been moving some of the nicer usability features into DB2.
onstat = best command line db tool ever. Want to upgrade? Just rename the Informix directory and install Informix into the new one. DONE.
I like Postgres alot, because of it’s unbreakability, and the high polish of its engineering, and wonder why it has not picked up as much traction in the commercial space.
[...] 15th, 2009 · No Comments Last week I blogged about uncertain future of MySQL in Oracle’s hands. At that time a number of people pinged me to say that they thought that I [...]
MySQL’s real strength is its brand. If you go back about eight years ago it had a technology & cost edge over its competition – postgresql was too slow and didn’t have a native windows install. The *vast* number of defects and non-standard features were accepted by users as a small price to pay for saving many thousands of dollars.
But that was then. Now Postgresql is far better positioned than MySQL in everything except branding. Postgresql is faster for many non-trivial apps, runs on as many platforms, is more free (no Byzantine licensing), has no worries about its future, and very importantly – does not have a buggy product.
MySQL should be in trouble from postgresql – even if Oracle or Sun didn’t buy them. But branding can trump technology, so who knows?
[...] blogged about my opinions about what the future holds for MySQL customers in light of Oracle’s acquisition of SUN and MySQL. I predicted, contrary to the opinions of some of the industry pundits, that MySQL within [...]
I don’t think brand recognition carries so much weight. I’ve used MySQL for years and affirm its speed, ease of install, and strong cross-platform support, if it occasionally needs a table rebuild. Many queries are easier to write for MySQL but PostgreSQL certainly offers easy to swallow licensing, robustness, data integrity, and many advanced capabilities. db_STRESS benchmarks applied in http://planet.mysql.com/?tag_search=5844 indicate MySQL-5.4 outperforming PostgreSQL-8.3.7 by 13.5 to 7 in read-only and 7.5 to 6.5 in read-write mode. I think the trade-off is MySQL performance, ease of use, and vast community versus more favorable PG licensing, data integrity, advanced features, and an open source future.
In the short term, PG suffers from a Berkeley UNIX world view. Building client PG applications for Win32 is hampered by a need for the source tree that still has missing pieces (e.g. pg_config.exe), that even with source often refuse to compile. A substantially smaller PG community means fewer source examples on forums and blogs that could address externalities like these. If the PG team cleaned up some long ignored Win32 client development issues, narrowed the performance gap, and adopted some of MySQL’s SQL extensions (e.g. IF [NOT] EXISTS…), PG might accelerate its adoption rate, perhaps becoming the heir apparent to MySQL’s free database dominance – at least until Larry Ellison figures out how to take PG out of the picture.