Simon Phipps, one of the most vocal supporters of open source inside Sun Microsystems announced in his blog yesterday that he is stepping down from his post as Chief Open Source Officer. He didn’t write his farewell as a haiku, ala former Sun CEO Jonathan Schwartz, but it was poetic nonetheless.
Phipps, a 10-year veteran of the company, didn’t spell out whether he was being forced out or was leaving on his own. But given Oracle is now in a position to be the most powerful open source company in the industry, and could crystallize some of the goals Sun didn’t get a chance to complete, one has to wonder.
In his farewell blog Phipps cites several of those unrealized goals. Phipps said he is “sad” the company didn’t get the code for those projects that remained permanently outside the Sun firewall, that Apache failed to get the TCK license Sun requested, that the company never got to a place where co-developers became more of a priority for a number of Sun’s product groups.
Lastly, perhaps in a jab thrown at Oracle, Phipps writes he is disappointed that despite the overall success of the company’s open source business, “it still wasn’t enough to rescue Sun in the end.”
Still he looks back with pride on what he believes are Sun’s major contributions to the open source community. Under his reign he says Sun got “some of the most important software in the computer industry,” released under Free licenses thereby guaranteeing “software freedom” for people regardless of own the copyrights including Unix, Java, key elements of Linux and the Sparc chip.
Another major accomplishment was creating the Open Document Format, which he said was instrumental in guiding “the quiet revolution” that has helped restore competition to the productivity software market.
Announcing his departure though his blog is appropriate given he believes one of his major accomplishments was starting the first blogs at Sun.blog.com, which kicked off the “corporate blogging revolution.”
Lastly, he takes satisfaction in changing Sun’s attitude about open source turning colleagues who were bitter critics of the technology into defenders of it, and even convincing people to join Sun because of its fervor in supporting open source.
Besides continuing to blog, Phipps says he has not decided what he will do next. Given his resume Phipps shouldn’t be unemployed long, unless he chooses to be. I think IBM might be interested in talking to Phipps and getting their hands on his little black book containing Sun’s open source plans.
Oracle has yet to fully spell out its plans for how it intends to leverage Sun’s rich portfolio of open source products and technologies – aside from committing lots of cash to the care and feeding of the MySQL database.
But open source figures to be an important if not strategic asset to Oracle for both its proprietary and newly acquired open source portfolios. It will be particularly important as it engages Microsoft in hand-to-hand combat at the lower end of the enterprise market. It should be careful about the open source talent it lets walk out the door.]]>
But the EU’s latest objection over the Oracle-Sun deal has little to do with being a watch dog ensuring fair competition and more to do with being an attack dog. We have to assume that EU competition commissioner Neelie Kroes has nothing in particular against Sun Microsystems, but through this upcoming investigation she could do irreparable harm to the company as well as to the fortunes of the newly combined company.
Even before the proposed deal, Sun was a distant fourth behind IBM, Hewlett-Packard and Dell for server hardware market share. Since the April 20 announcement, Sun’s market share is in freefall as Sun customers freeze their buying decisions waiting to see what Oracle will do with its hardware business. Adding insult to injury IBM and HP have had a field day picking off Sun users through aggressive new pricing programs.
In late August, Sun suffered its fourth consecutive loss this time reporting it lost $147 million in its fiscal fourth quarter compared to a net gain of $88 million in the same quarter a year ago. The company’s revenues plummeted 31% in the fourth quarter to $2.36 billion. If the EU investigation goes on for even another quarter or all the way to its January 19 deadline – and some pundits believe it will – there will be very little left for Oracle to work with going into 2010.
With Sun’s server business shriveled to that of a second tier supplier, what does this do to the soup-to-nuts, integrated stack computing strategy Oracle executives have been yammering on about since last spring? More than a few users I have talked to the last few months are now cozying up to the notion of an integrated hardware-software stack from Oracle. They point to the company’s Exadata Database Machine — Oracle databases optimized to run hand-in-glove with HP’s server hardware — saying the system far out distances the performance they get with databases they purchased separately. Can Oracle just walk away from that promise and still look good to users? I don’t see how.
And if Oracle is hoping to pawn Sun’s failing server business off to an IBM or HP after this EU mess gets straightened out in three, four or six months, they may face a fierce level of indifference.
In her statement last week Kroes made it clear one of the major reasons for the investigation is her concern over the world’s largest proprietary database maker taking over the world’s largest open source database. With one company owning all that database software, bad things most certainly are going to happen to database competitors in Europe and user choice will be nill, right?
Come on now, Neelie, this is open source we are talking about. No one can truly own and control a popular open source product like MySQL – it belongs to the community. The best Oracle can do is coordinate some of the development going on around the product. Most open source licenses demand that whatever you create must be thrown back in to the open source ocean for others to use and/or build on.
There should be no fear about MySQL being “Oraclized” taking the heart, soul and brain out of the product so it works better with Oracle’s proprietary products. If Larry’s boys do something like that non-Oracle shops don’t have to buy it. There will always be another version out in the community.
And one other thing: does Oracle’s proprietary database comes close to competing head-to-head with MySQL? Hardly, they are at opposite ends of the competitive spectrum.
So Neelie, for the competitive good of the industry, make this a quick investigation. Competitors are quite able to destroy each other and sometimes themselves – they don’t need any outside help.]]>
What was encouraging about the announcement was first, the commitment to using open source technologies, and second using it in a way that allows IT shops and ISVs to address a couple of fundamental challenges and to do so inexpensively.
The VM Templates are essentially virtual machines that contain pre-installed and pre-configured enterprise-class software. IT shops can use them to create, package up and deploy applications significantly faster. The VM Template Builder takes advantage of the Oracle Enterprise Linux “Just enough OS” (JeOS)-based scripts.
Given IT shops growing interest in cloud computing and other SaaS strategies, being able to simply and quickly launch virtual applications figures to have a lot of appeal to Oracle shops.
“We are picking up a lot of interest from our users about cloud computing, with a particular interest in deploying private clouds. They are trying to figure out what it means for them,” said Monica Kumar, Oracle‘s senior director of Linux and open source product marketing. “Virtualization is one of the enabling technologies for cloud computing and a lot of other things. We are looking at virtualization from solutions perspective such as deploying a CRM stack virtually,” she said.
Besides the VM Template Builder Oracle also announced the VM template for its Siebel CRM product, which allows users to set up a complete Siebel environment quickly, as well as a test kit for testing a stack of applications before it gets deployed as part of the Oracle Validated Configurations Program.
The VM Template Builder, the test kit, and Siebel template are free, but users must have a license for Siebel CRM.
“If a user has a private cloud infrastructure, with these templates they can bring up applications and solutions much faster. And not just that, with the Template Builder they can build these templates themselves,” Kumar said.
Offering an example Kumar said if an IT shop had a financial application, it could use the VM Builder product to create a virtual machine image of the software that could then be downloaded by customers for deployment. Otherwise, customers would have to deploy applications themselves from scratch.
“The template has a GUI and the underlying technology is Linux and so together they basically give ISVs the ability to create a software-in-a-box environment,” Kumar said.
Kumar, of course, would not comment on how the new templates might work and play with Sun’s xVM Server hypervisor or Solaris virtual containers or even the technologies it has picked up from its acquisition of Virtual Iron.
Hopefully, it will apply most of these newly acquitted open source technologies as practically as it has its own VM templates.
There hasn’t been a lot of hollering about the price hikes, of course, because many corporate accounts are locked into long term software licensing agreements they can do little about until the agreements expire. And the same for maintenance agreements which go hand in hand with the software agreements.
Still, it is perplexing why in the midst of The Great Recession Oracle would raise prices by thousands of dollars on some products when some of its best customers are counting pennies as well as the minutes until this crisis eases.
It is true that the elevated prices represent merely a starting point in negotiations between Oracle and it customers. There is nothing stopping Oracle from dropping those prices in behind closed doors talks, but why poke the rattlesnake with a stick?
But it is moments like this where some frustrated users leave open that door to exploring opportunities to lower cost options, even if it is only for one-off departmental solutions.
This week, a relatively small open source database make, Enterprise DB may have given some frustrated users reason to keep that door to alternatives open.
With some clever marketing the company announced its Oracle Migration Assessment Program — which the company with a wry smile calls the Oracle Bailout Program – is designed as an enticement to attract users away from Oracle’s database lineup and sky high licensing fees, and over to the decidedly less expensive Postgres Plus Advanced Server.
“Oracle’s price hikes might be good news for those on Wall Street, but they’re terrible news for IT departments trying to function in the worst economy since the Great Depression,” said Ed Boyajian, the president and CEO of EnterpriseDB. True enough.
Admittedly however the program sounds a bit gimmicky, EnterpriseDB dreamed it up in direct reaction to the July 1 price increases. And the company is not releasing names of those who are interested in migrating away from Oracle. But the company’s timing could be just about right. So it is safe to say there isn’t a land rush toward the program.
The “bailout” essentially helps enterprise customers migrate any and all applications running on top of Oracle databases to the Postgres product, which is an open source based database with a built in Oracle compatibility layer. EnterpriseDB officials said they are guaranteeing no disruptions to an IT shop’s operation during a migration, truly a brave statement.
Enterprise officials may be going too far in saying the price hikes to the performance and tuning packs and other database options are indicative of what they will do with their newly acquired open-source bauble, MySQL. Only the good Lord and Larry Ellison know what they will do with MySQL, and I am betting the former is even a bit in the dark.
But if EnterpriseDB and other open source comrades can come up with some creative approaches that make financial sense, they could apply pressure on Oracle to lower pricing on its proprietary products and even force them to make a more obvious commitment to open source.]]>
Oracle’s continued lack of public enthusiasm, let’s call it, about gaining possession over MySQL Server continues to puzzle me.
The folks at Redwood Shores certainly haven’t indicated they will abandon the popular open-source data base, but neither have they acknowledged what the strategic importance to its overall data base fortunes might be.
I know Oracle can’t offer up too many specifics about its plans for MySQL until its acquisition of Sun is complete (the latest speculation is that approval may not come now until at least September over some concerns expressed by the European Union). But it could do a better job of making a general statement or two about its potential value, which might go a long way towards making its customers using Oracle and MySQL databases side by side feel more assured.
The company shouldn’t wait too much longer to do that.
A couple of weeks ago, Monty Widenius, MySQL’s founder who left Sun before the Oracle deal, has formed an independent vendor-neutral consortium that will serve as a hub to create and maintain code and binaries, as well as offer training and technical support for MySQL.
The fledgling organization, called the Open Database Alliance, will supply a range of software and services for Widenius’ fork of the MySQL MariaDB version of the product. The group will not wait for Oracle’s endorsement or formal participation.
At the announcement of the Alliance Widenius expressed more than a little concern that MySQL’s development efforts could be set back years if Oracle either lets the product languish without regular updates, or lays off many of the product’s programmers at Sun once the deal is completed.
Widenius pledged to work closely with those MySQL developers at Sun, to prevent a significant forking of the code which would fracture an otherwise united development community. This, of course, could result in diluting the product’s competitive powers against Microsoft, or make it less attractive to Oracle shops as a departmental-level compliment to its higher-end databases.
Not just that. Some of Oracle’s database competitors could join the new consortium and make technical contributions to MySQL. It is hard to imagine that Oracle would be comfortable with that scenario having just paid over $7 billion for Sun.
Widenius also made it clear his new organization is quite open to any company or individuals joining in the group. Given there are only the two founding members who have joined – namely Monty Program Ab and Percona – new members are likely to have more than a little influence in the product’s direction.
I have already written about MySQL’s more obvious advantages to Oracle’s data base business: a strong lower-end compliment to Oracle’s proprietary line of data bases that could effectively compete against Microsoft; another source of maintenance revenues; and a way for Oracle to take a leadership position in the open source world and improve its credibility there.
No matter what Oracle’s longer range plans are for MySQL, the company would be wise to throw The Open Database Alliance a bone soon, letting it know it intends to work cooperatively. It would be good for not only MySQL users and developers, but for its own strategic good.]]>