Kurian discussed the roadmap for JDK 7 and JDK 8, which will be based on OpenJDK

“In addition, Oracle remains committed to OpenJDK as the the best open source Java implementation and we will continue to improve OpenJDK and welcome external contributors.”
http://www.oracle.com/us/corporate/press/173712

“Oracle will work with the OpenJDK code base and the OpenJDK community
like Sun did. We will continue to develop the JDK in the open under a
GPL license. We welcome the cooperation and contribution of any member
of the community – individuals as well as organizations – who would
like to be part of moving the most widely used software platform
forward.”

http://blogs.oracle.com/henrik/

Well, it seems that another piece of the Sun OSS portfolio has floated adrift from the Oracle mother-ship. Some of the key developers have left / are leaving Oracle and have joined Rackspace. With this Rackspace essentially becomes the guardian of one of the major development branches of MySQL. It would be interesting to know what Oracle would price their MySQL assets at today.

The things I care about are the people I know who still work for Sun, the Java ecosystem, and many of Sun’s Open Source projects that I directly or indirectly benefit from – specifically MySQL (this post will live in MySQL), OpenJDK and OpenOffice.

It still seems like a strange merger – sure, Sun and Oracle have a huge shared installed based – Sun Servers + Solaris + BEA + Oracle DBMS was the killer enterprise stack for a decade – that alone gives Oracle a vice-like grip on existing customers; but that’s about the past – not the future. I think someone did the analysis and realized it’s a marginally positive move – so I don’t think this is the big technology vision realized that Oracle are trying to promote.

I’m sure a lot of people at Sun and customers of Sun are glad the uncertainty has come to an end but unfortunately it hasn’t. I’m guessing that it will take until the end of the year before Oracle tells the world what they’re keeping and what they’re dropping. FWIW – here’s my informed guess :

Storage / Servers – the press-briefing made the merger sound like it was all about the hardware (servers and storage). I just don’t see Oracle as a hardware company and they have more to loose than gain by pissing HP off. I wouldn’t be at all surprised if some or all the hardware got sold off to HP, Fujitsu, etc. And I think Oracle could probably recover $3.5-4 bn of their $5bn outlay by doing so.

NetBeans – I’m not changing my mind – it’s done for. Oracle has two Java IDEs already – they don’t need a third. Oracle will pick up some great tools developers to write a migration tool but that’s about it.

OpenOffice – Oracle doesn’t like M$ but I doubt they can find a fiscally rational reason to carry the torch for Open Source and at the end of the day Oracle are doing this because it makes financial sense. I think Oracle will expect “the community” to pick up the slack.

Solaris – I think they’ll milk the legacy installed base (which is huge) but the innovation won’t continue. I think there’s an interesting opportunity for Oracle to manage the migration of the last of the Solaris holdouts to Linux. They could do this by GPLing Solaris and moving some of the Solaris features to Linux; or more likely to an Oracle proprietary OS built on Linux.

Middleware – Oracle have everything Sun has – and Oracle are by and large market leaders with large market and revenue share. Sun have one or two products that might survive and certainly some components. Oracle inherit a commitment to continue to produce the Java EE RI (and others) so Glassfish *might* survive – but Oracle have demonstrated that they don’t have appetite to maintain many products in the same market (see how quickly OC4J got killed after the BEA acquisition ?)

MySQL – I think Oracle will continue to do what Sun did somewhat accidentally – namely slowly kill it off.

Java – I’m confident that Oracle won’t fix the JCP and won’t sacrifice control for the good of the ecosystem. I also think the JCP will become a battlefield for IBM and Oracle and we’ll all be the worse for it.

OpenJDK – no-body needs two JDK’s to maintain – I think Oracle would have to move some of the monitoring / diagnostics from JRockit into a proprietary OpenJDK-based platform. JRockit has a small market share but some nice features for enterprise customers.

Cloud – not that there’s much there there but I don’t expect to see Larry eating his words.

Virtualization – I don’t really understand Oracle’s virtualization strategy other than the feeling that Larry Ellison is unlikely to entertain anything that looks like it might actually save customers money on licenses.

I promised myself I wouldn’t bite; yet here I am. I’m contributing to the “Here’s what Sun should do meme”. And why not -  I’m as good a keyboard-CEO as the next person and I did work for Sun for almost 9 years.

But first I should note – I think Tim Bray was pretty bold starting this (as he’s still a Sun Employee); and Sun is a pretty unique company in that it’s OK to do what Tim did (assuming Tim didn’t get booted).

#1 Java. I’m part of the Java ecosystem so it’s important to me. Basically, do what James says:

…, at worst go back to SUNW as a stock ticker.”

then do what Tim says :

As for Sun’s role as Steward of Java, and in particular the Java Community Process, let it go already. Java has mostly won and is mostly the establishment, and the community is smart and conservative enough to keep anyone from doing what Microsoft tried last millennium, or in any other way to subvert Java’s interoperability. In 2008, the JCP is costing Sun opportunities and friends and gaining us very little that I can see.

So I’d like Sun to set the JCP free, turn it over to the community, and when we develop some cool Java-based technology in-house, take it to market, try to make some money with it, and after it’s caught on and the bugs are shaken out, consider whether or not it ought to be taken to the JCP.”

But the community is going to have to continue to be smart and I also think that a consortium of interested parties would have to step up and invest some resources in making it work really well. Getting the balance of commercial interest and community is the key – I think Eclipse has it about right. Money makes the world go round – sorry if that offends anyone. Oh and while we’re at it – OpenJDK should be treated like the Linux Kernel – we only need one code base (note IBM and Oracle).

#2 NetBeans

I’ve used both Net Beans and Eclipse (not for a year or two mind) – Net Beans is better, more polished, etc. But it doesn’t matter – it’s not a sustainable product. When Sun stops it’s investment – it will fade away. Eclipse won in mind share years ago; Sun should have recognized this in 2005 – I did. Java needed something like NetBeans in 1999 to attract developers but not anymore.

#3 Solaris

See above. Linux has won. Whatever technical merit Solaris has today will be commoditized next year. The innovation around Linux is relentless.

#4 Middleware

Sun has some great products and technology; some that are popular and fit well with Sun’s demand driven model (eg. Glassfish) and some that make money (Identity Management). Figure out what you need to do #5 – choose the best available OSS solution and run with it.

#5 MySQL

Invest, Innovate. Hire sales people who can sell data-based solutions – steal $1bn from Oracle this year, $2bn next year, repeat.

Demand for storing, archiving, cleaning, replicating and accessing data is going to grow. Forever.

#6 Storage, SPARC, Volume Servers

Keep only what is required for #5.

#7 Open Office

I’m a beneficiary of Open Office – I use it exclusively but have never paid a penny. That’s the problem – I use it because it’s free and it’s the best Office Suite available for Linux. There is no business case here. Thanks for a great product though.

#8 Java FX

Drop it. Nobody will notice.

If this sounds depressing – I’m sorry but it should be clear that Sun isn’t going to be the company it was in 1999. It needs to adjust its cost-model inline with its Open Source strategy; it needs to be much smaller; and needs to do much, much less, much, much more sucessfuly. Even if Sun goes private.

Seeking Alpha has some thoughts on the vulnerability of software companies to reduction in maintenance payments. Read the comments as well. I agree that many companies will scrutinize the value they get from maintenance contracts.

We’ve just launched a new ISV program around our certified platforms. JBoss is everywhere – now we can do a better job of working with the whole JBoss ecosystem. We’re nothing without a healthy, growing and diversified ecosystem of partners.

This graph really puts things into perspective and I really hope I’m not still looking at it in a year’s time wondering where the bottom is.

Finally (finally!!) – Judgment in SCO v. Novell: SCO Loses Again. Who – in their right mind, in the current economic conditions, would be looking to invest in SCO so they can carry on with this nonesense.

But the reality is – no amount of progress with Open Source Java and GlassFish, sales of MySQL, Solaris or Web2oh coolness is going to move the needle enough to make a material difference to Sun’s fortunes as long as it lags the competition in selling servers. Because Sun is still dominantly a hardware company – and that’s where the revenue comes from.

And the issue is, and the reason I’m ranting is that this is fundamentally the same problem that Sun faced 8 years ago as Wintel and Linux started to creep onto Sun’s turf. From the recent IDC WW Server Tracker – that issue still hasn’t been addressed and I’m no longer confident it will be addressed adequately or quickly enough to save Sun. And that is a real shame – but if you’ve been staring at the oncoming steam train for 8 years and still get flattened – no-one is going to sympathize.

I hope I’m wrong.

But first – I’d like to make a point about Sun’s reaction. Opening Java was a big bold move – likely one of the biggest, boldest moves in our industry’s history. Sun has given Java the chance to truly flourish but in doing so has given up a little bit of control and a little bit of the ownership. From time to time others will play a role in advancing Java as the Fedora team and Red Hat did over the last year; and they’ll rightfully take the credit for doing so. That’s a consequence of opening Java. On the whole I was pleased with Sun’s reaction – many Sun people involved in Java actively promoted the IcedTea milestone, recognizing it as progress for the Java platform as well. It could have been different – after all Sun and Red Hat are also competitors. I think if Sun continues to encourage others to promote and advance Java – all will be good.

OK, let me address some of the questions and misconceptions :

Wow – a second Java implementation – this one is free and open source – way to go.

Actually there have always been many Java implementations – eg. Sun, BEA, IBM, Azul, HP even Microsoft had one at one point (but that’s another story). And they were always free ($0) and as of a year ago Sun’s was also Open Source. Java is a pretty open specification – you have always been able to become a licensee and implement your own Java. But that is a huge amount of work.

Who cares – Java is slow.

Too funny. But much funnier back in 1999 when it was also partially true.

OK, but this one’s going to be slow

The OpenJDK in Fedora 9 is largely (ie. 96%-99%) based on Sun’s JDK 1.6 – so it’s likely to perform about the same.

I’m worried about fragmentation – now anyone can have their own Java-like distro.

Firstly, Given the momentum behind Java – producing something that is partially Java isn’t going to be particularly interesting. Sun still protects the Java brand – if it’s called Java, it must behave like Java (ie. minimally pass the Java TCK). I still can’t imagine why anyone (ie. other Linux distros) wouldn’t just take openJDK and IceTea as their imlementation of Java.

Sun Open Sourced Java a year ago – there’s no news here

The initial OpenJDK had encumberences which meant it could not be distrubted fully in some of the more important Linux distributions like Fedora. Now it can. That is significant – we have not seen the true potetnial of Java on Linux. Now we have an implementation everyone can feel good about -  we will.

OpenJDK has been distributed with <your favourite linux> since April.

Sure – but it likely had the ‘encumberances’ – while that may be an OK situation for many – it wasn’t an OK situation for everyone.

But my XYZ app doesn’t run on this jdk

Sun’s TCK can’t guarantee that every possible current and future app. will work – as I mentioned – passing the TCK is a milestone; not the end goal. If your app doesn’t run or behave as it should then help by filling a bug.

At JavaOne in May, 2006, Sun Microsystems announced they were going to release Java as free software under the terms of the GPL. The size of the task (6.5 million lines of code) was only eclipsed by the size of the opportunity for Java as a free and open technology.

At JavaOne in May 2007, Sun announced that the work was largely completed and so OpenJDK was launched. What was less newsworthy was the fact that on release – OpenJDK still relied on code that was encumbered – between 4 and 5 percent of the code was closed, non free source that Sun didn’t own.

Richard Stallman described the encumbered code as :

“The one last obstacle [which] remains in liberating JDK and disarming the Java Trap completely”

and rallied the FOSS and Java communities to

“… work together to replace that code with free software”

So, who would step up to the challenge of making Java truly free and open ?

In June, 2007 – Red Hat launched the IcedTea project with the goal of making OpenJDK usable without requiring any other software that is not free. That in turn would allow OpenJDK to be included in Fedora and other Linux distributions without restrictions. The IcedTea Project made use of previous work developed under the GNU Classpath Project which had been independently driving towards a free and open implementation of the Java class libraries.

This week the IcedTea Project reached an important milestone – The latest OpenJDK binary included in Fedora 9 (x86 and x86_64) passes the rigorous Java Test Compatibility Kit (TCK). This means that it provides all the required Java APIs and behaves like any other Java SE 6 implementation – in keeping with the portability goal of the Java platform. As of writing, Fedora 9 is the only operating system to include a free and open Java SE 6 implementation that has passed the Java TCK. All of the code that makes this possible has been made available to the IcedTea project so everyone can benefit from the work.

The Java TCK is a complex suite of tools and documentation that verifies that Java implementations conform to the Java specification. It consists of more than 80,000 tests and over 1 million lines of code.

From here the initial plans are to make OpenJDK part of Red Hat Enterprise Linux distributions starting with Red Hat Enterprise Linux 5.3 and to expand the platform support. Beyond that our plans are still evolving, but clearly this creates some great opportunities for both Red Hat and Java. For example :

• Improving Java for virtualized, hosted environments – an area where Red Hat Linux has excelled but Java has struggled.
• Optimizing the performance and scalability of the full stack of Java-based JBoss Enterprise Middleware for Red Hat Enterprise Linux and Linux in general
• Being able to better manage the lifecycle of JBoss Enterprise Middleware platforms and the Java Virtual machine on which it depends
• A more fundamental opportunity is for Red Hat to be able to increase the depth of support for the JBoss Enterprise Middleware platforms running on Red Hat Enterprise Linux.

Over the coming months, we’ll continue working with our communities of users, customers and partners to better understand the opportunities that OpenJDK and IcedTea present to us.

Working with Sun Microsystems and the broader Open Source Java community; Red Hat’s OpenJDK team included Tom Fitzsimmons, Lillian Angel, Gary Benson, Keith Seitz, Mark Wielaard and Andrew Haley.

Tom Fitzsimmons will be at the Red Hat Summit in Boston between June 18th and 20th, so if you want to chat about the project – swing by the Exhibit Hall, grab a beer, and ask him how much fun the Java TCK testing was.

Java, Java SE, OpenJDK and Java TCK are all Trademarks of Sun Microsystems Inc.

Time for a quick update. In many cases I’m comparing with Sun which is a) a little unfair because they’re companies in very different situations; and b) unfortunately my only reference point. So, I apologise beforehand if this post reads like I’m laying into Sun. That’s not my intention.

Business

Firstly, Red Hat and JBoss in particular, seem to be on a roll right now – every week seems to have bought news of Red Hat’s continued success. First the Fiscal Q4 results, a really successful Java One, then really encouraging reports from  Nikkei Market Access, Aberdeen Group, Gartner and Forrester. News like this assuage any ‘job switchers remorse’

Technology

One thing that still impresses me about Red Hat is the pace of innovation. Red Hat isn’t a big company and JBoss is still a small piece of the whole but we seem to be punching way above our weight. This is fuelled by constant innovation in the JBoss.org projects. During my short time here; and specifically for JBoss, I’ve seen the following releases :

• JON 2.0 GA
• JBoss DNA 0.1
• JBoss Tools 2.1
• JBoss ESB 4.3
• JBoss Messaging 2.0
• JBoss RESTeasy 1.0 beta
• EAP 4.3 CP02
• EAP 4.2 CP04
• Seam 2.0.2

In addition there are some pretty interesting things going on behind the scenes – eg. EAP 4.3 is starting evaluation for EAL 2+;  AS 5.0 and EAP 5.0 are taking shape and OpenJDK in Fedora making progress. One thing I know for sure is that the cadence isn’t going to slow – this summer is going to be busy.

While I’m talking about technology – I should also mention I’m back to being a full-time Linux user (Fedora 9 to be specific). I keep having OS/X withdrawals but so far Fedora has kept me out of the Apple store.

Culture and People

Superficially,  the Red Hat culture isn’t radically different from Sun. They are both technology companies – full of smart, very technical people where technology rules. Questioning corporate mandates seems to expected at Red Hat – as it was at Sun – though I’m sure changing them is a lot easier at Red Hat due to its relative size.

Many of the JBoss people I work with share one thing in common – they used to be a customer. Whether, VP of engineering, support engineer or product manager – you’ll find people who joined JBoss because being a customer wasn’t enough. What this means is that there is a huge amount of customer insight in the organization. OK – I’m going to slap Sun around a little – sorry. In the last few years at Sun I saw a worrying trend – hiring people with absolutely no knowledge of the products Sun sells; people who have no interest in understanding how the products worked or what problems they solved for customers. For a product company – that’s wrong. Wrong.

Work Environment

I’m spending about 50% of my time in the office – that’s 50% up on the last 3 years – travel takes most of the rest. The change is not as bad as I thought it would be – my commute to Red Hat’s HQ in Raleigh is about as easy as it gets (typically less than 20 mins, often just over 10). Actually getting fully dressed in the morning before work hasn’t been a problem either. One thing I miss is  being able to hit the Gym outside peak hours – my Gym hours are way down currently.

JBoss is probably a lot more distributed than other organizations – I doubt there’s an office anywhere in the world with more than a handful lof JBossians – so phone, mail and IRC rule. Twitter – not so much. Actually meeting people is a pretty infrequent opportunity – Red Hat summit will be the next.

I guess Sun see’s JavaFX as a cornerstone of it’s mobile strategy – which I still don’t get. I get that mobile devices are increasingly important – what I don’t get is how Sun thinks it can be come a comsumer electronics company nor why it feels it needs to. I’m not even completely sure about Java on mobile devices – I’ve spend the last 10 years with Java and was an early adopter of the cell phone and have lost count of the devices I’ve owned and the money I’ve spent on services. Yet I still haven’t seen a compelling / unique use for Java on a cell phone; and I simlply don’t buy that Java needs to become a consumer brand to become sucessful. It’s already wildly sucessful.

Java is one of the most sucessful enabling technologies ever but it it’s not a consumer brand.