Oracle and Sun: Match Made in Hades?
Apr 30, 2009 4:00 AM PT
Such was the excitement over Jaunty Jackalope's arrival last week, that the tremors are still reverberating throughout the blogosphere.
It's been a virtual parade of reviews and first impressions, in fact, with the gradual addition of hands-on tips and advice. Lifehacker, for example, posted a "Top 10 Ubuntu Downloads" list over the weekend, garnering no fewer than 1,730 or so Diggs and more than 130 comments by Wednesday.
Jaunty, you've made a lot of geeks very, very happy!
'This Sounds Like Trouble'
There were a few other topics for which bloggers were willing to rip their attention away from the new release, however, and probably the most notable among them was Oracle's purchase of Sun.
As the Linux Loop's Thomas Teisberg put it, "Oracle, maker of one of the largest proprietary database solutions, recently purchased Sun, who makes MySQL, a piece of open-source database software. Naturally, this sounds like trouble."
Indeed, bloggers on LXer, Digg and Slashdot all got into the conversation, pondering what the acquisition might mean not just for MySQL but also for Java, Solaris, OpenOffice.org and even Linux itself.
'Forked to Hell and Very Unhealthy'
"I would have preferred if Sun ended up in IBM's bedroom instead, but time will tell how Oracle will handle the open source projects," wrote hongkongjapie on Digg, for example. "My biggest fear is adding layers of bureaucracy will result in departure of key developers (MySQL anyone?)."
On the other hand: "I'm more worried about OpenOffice; the project is forked to hell and very unhealthy," countered matthekc. "The end product doesn't suck but if Oracle ignores it for long I worry that it may lose precious ground to Microsoft Office."
Concern and speculation reached such a fever pitch in the days following the announcement that we here at LinuxInsider felt compelled to take a small poll of our own.
"On one hand, there are plenty of reasons for Oracle to keep MySQL alive," Chris Travers, a Slashdot blogger who works on the LedgerSMB project, told LinuxInsider. "There is no indication that MySQL users will switch to Oracle, and MySQL could represent a revenue stream for Oracle."
However, there are "a few reasons to see this as particularly troubling," Travers added. "Oracle has pursued a strategy which has placed pressure on MySQL by purchasing important partner companies, including Sleepycat Software (makers of the Berkeley Database engine, used for MySQL BDB table) and Innobase (makers of InnoDB engine for MySQL)," he explained. "This strategy has seemed to target MySQL specifically and predated Sun's acquisition of that MySQL AB."
Opportunity for PostgreSQL?
The result, Travers added, is that "there are reasonable fears about MySQL's future. Oracle thus far has continued to support InnoDB and BDB, so this concern may be overblown, but nonetheless, the future of MySQL seems unsettled at the moment."
One consequence of that uncertainty, in turn, is an increasing interest in PostgreSQL, Travers added. "PostgreSQL is an excellent database, and would be a likely choice of folks moving to a different Free/Open Source RDBMS," he predicted.
Not everyone saw reason for concern, however.
"I don't know why people are so worried about Oracle," Montreal consultant and Slashdot blogger Gerhard Mack told LinuxInsider. "I mean, sure, Sun's software might have been nicer in IBM's hands, but it's not as if Sun was doing all that well with them to begin with."
OpenOffice "suffers from developer discontent, has multiple forks, and seems to suffer from a NIH syndrome where Sun seems to want to reimplement other people's patches rather than accepting them as is," Mack noted. "It may be open source, but it's being run like a standard proprietary project."
'An Outright Disaster'
Java, for its part, "is only recently open sourced and has the same NIH issues as OpenOffice," added Mack. "On top of that, it has issues with Sun not bothering to keep backwards compatibility across point releases, leading to version hell when trying to manage Java projects."
MySQL, finally, "has been an outright disaster since much of their best talent left the company after Sun bought it," Mack asserted. "Not that it really matters anyway: On the low end, you're better off with SQlite; on the high end, you're better off with PostgreSQL -- unless you need replication, which is the one thing that PostgreSQL does poorly."
'A Lot of Destruction in the Process'
Blogger Robert Pogson had also been intrigued by the possible combination of Sun and IBM. "It looked like it could be a happy marriage in that each did certain things well and together they could do everything well, but egos got in the way," he told LinuxInsider.
Then came Oracle -- "not my favorite Mom and Pop corner store example of Free Software, but one of the icons of proprietary software," Pogson noted. "Ellison has shown insight in advocating thin-client computing, but Oracle is such a niche product, I cannot see how Sun could survive the process with any level of dignity; eventually, the whole thing may make sense, but there will be a lot of destruction in the process."
Pogson uses OpenOffice.org, MySQL and VirtualBox "heavily," he said. "These are anchors for my world on desktop and server."
'It Is All Good'
Ellison "may see these things as useful tools to make friends and to influence people and support them," Pogson noted. Alternatively, "he may see them as burdens or competition and not support them. He spoke of being able to provide a 'complete stack' to customers, so I am hopeful."
There's room for everyone in the world of Free Software, he added. "This acquisition may be the perfect proof of that. I expect Oracle and Sun will use the combined resources of both to cater to a wider base of customers, small, medium and large. It is all good. I expect Oracle and IBM to continue to be major contributors to FLOSS, killing the FUD that one cannot make money from it."
'Oracle Makes Sense'
Oracle actually makes more sense as a parent of Sun than IBM or Cisco would have, Slashdot blogger hairyfeet asserted.
"With IBM there was simply too much of a culture clash, and with Cisco there simply wouldn't have been enough value," he told LinuxInsider by email. "Oracle makes sense, and here's why: integration."
Oracle needs a solid OS and hardware that it can control; it also has built many of its apps on Java, hairyfeet noted. "By having control of Java, Solaris and the Sun server hardware arch, they can VERY tightly integrate the Oracle DB and maximize throughput while at the same time having a rock-solid stable solution to offer their customers."
'Oracle Needs to Fund OpenOffice'
OpenOffice.org is at the heart of Slashdot blogger yagu's primary concerns.
"From a purely selfish view, I'm worried about OpenOffice.org and Oracle's plans to continue support or not," yagu told LinuxInsider. "From articles I've seen over the week, I get the feeling Oracle's not hot for OpenOffice, and that's a shame. If it's true, it's a bit of a shirk."
OpenOffice "is the first thing I install on all friends' and family's new computers," yagu explained. "It's a rock-solid, free, drop-in replacement for Microsoft. It may be one of the few things keeping Microsoft from pushing their prices even higher.
"Oracle has benefited as much or more than many from open source," he added. "If Oracle does anything to damage OpenOffice, it begins to seem possible a behind-the-scenes discussion with Microsoft may have occurred."
Bottom line: "Oracle needs to fund OpenOffice," yagu concluded.
'The Greatest Danger Is Sparc'
One piece of the puzzle hasn't yet been mentioned, however, and that's Sparc. For Slashdot blogger drinkypoo, it's the biggest point of concern.
"Arguably, the piece of Sun in the greatest danger is Sparc," drinkypoo told LinuxInsider.
"For the immediately foreseeable future, it makes sense for Oracle to keep it going --people are paying for Sun kit, after all," drinkypoo noted. "But this change does come at an interesting time -- immediately after MySQL has gained clustering functionality."
'Clustering Is Taking Over'
Clustered databases "are the major threat to monolithic instances like Oracle, which depend on big iron single-system-image servers like the high-end Sun kit. Oddly enough, they are also the major threat to Sparc, which has basically only one reason left to live: operating a large RDBMS," drinkypoo said. "IBM has all but completely destroyed Sun in scientific computing, and clustering is taking over the other remaining reasons for its existence.
"MySQL may stumble, but due to its licensing it can simply be forked," he concluded. "But if Sun couldn't iterate Sparc and keep it competitive, what hope does a Sun owned by Oracle have?"