Skip to main content

Dear Community, Get a Clue

Posted by fabriziogiudici on October 24, 2010 at 10:57 AM PDT

Ian Skerrett posted a good article a few days ago, which is recalling how the communication between Oracle and the Community is not working. I strongly agree with 2/3 of it, in fact I voted it. Summing up, the basic points are:

  • Oracle is not good at communicating with the community and it's not doing it
  • Oracle is adopting subspicious tactics about the way in which some seats will be renewed at the JCP elections

In particular, for the latter point, there's an interesting post by Stephen Colebourne. Stephen thinks that there's a risk that Oracle is trying to have a friendly corporate to be nominated at JCP in place of Doug Lea, that stepped out. Stephen might be right or wrong, but I think he's right in raising a yellow flag. In the meantime, Doug explained that he stepped out of the JCP with a critical attitude and he doesn't believe any longer of the openness of the JCP.

Frankly, I wouldn't rend my garments if it was true: it's an election
and elections are a matter of politics. Each party has got the right to
try to get elected allies exercising the power that he has got
(actually, Stephen is acting on behalf of the community's interests);
the only unfair play would be to cheat the vote count, which is not
happening.

This makes very sense with the idea that I have about Oracle policy and that I've expressed in my previous post: Oracle is just trying to grab as much as control it can on Java. This doesn't mean that Java is more closed than before; Oracle's attitude is not a complete U-turn from Sun (let's recall that the Apache debate started with Sun) and indeed at least now the strategy is very clear and not ambiguous. We're still having some freedoms from OpenJDK licensing (see again my previous post) and we have just to pragmatically evaluate whether they're enough or not for us. If you don't believe they are, you shouldn't be reading here any longer, since it's a Java blog and you'd rather start working on a migration plan and exit the Community.

If you believe they are enough, you're probably enjoying the fact that Oracle, by exercising a strong stewardship, is assuring to Java a bright future and it will be probably successful in applying the roadmap that, at last, will move the stagnant waters that Sun left us in. The point is now how to maximize the community's benefits. And so, we're back to Ian's post: how do we manage the relationship with Oracle? First, it's business, that is it's a matter of quid pro quo: we can get some benefit if we can offer something back. Sun thought that there were many good opportunities with the Community, that's why - with some exceptions, such as Apache - the relationship was usually good. Now read my lips: Sun's business model failed. They created a bubble in which the community prospered, but they didn't get enough back. On the contrary, Oracle has done very well in the past and it's doing well right now. I think they think they don't need the community, and they could be even right. 

That's why I find it naive that we suggest Oracle to "get a clue" about their customers and forecast doomsday scenarios with their customers if the community is not involved enough. If a corporate makes lot of money, it means that it's managing customers very well. I think that a top manager, in the unlikely case he's reading a business prediction by an engineer, just laughs.

That's why this blog ends with a question (as usual). What we really have to offer to Oracle? That's the first thing that I'd really like to know before discussing whether we have to go at war, or make peace (and which kind of peace) with Oracle. Unfortunately, I don't see answers so far.

PS Yes, the community proved able to run self-sustaining businesses on top of Java: see just as an example SpringSource (Rod Johnson made the point several times, including not going to JavaOne this year), RedHat, the Apache ecosystem. But these are independent worlds from Oracle: Oracle doesn't get money out of them, but doesn't want or can't kill them. They can probably evolve in mutual independence, not caring each the other. But keeping a minimum of cohesion between the "traditional business" à la Oracle, IBM and the "community business" would be important for keeping Java strong. On the contrary, I think we've much to lose than Oracle.

Related Topics >>

Comments

As a developer the community

As a developer the community was a major value add to Java. Without that, why should i choose Java over .NET now? Either way i'd be limited to a single vendor and their business strategies/models.

I guess your answer would by that my manager will choose for me based on whoever have the glossiest brochure. I mean, who cares what developers think?

@dirtyqwerty Almost there.

@dirtyqwerty Almost there. Indeed there are corporates whose managers are "community-educated". But they are a minority. For the rest, right, developers aren't those who make decisions.

@scotty69 My point is not that we really don't have a value. Wade's response has got some insights, maybe we have some. But it's something to think carefully to guess whether it's just our perception or something real. For instance, Wade mentioned higher education. I could counter-argue that, given the status of many universities, Oracle might probably keep the Java push by just giving a few money to them, in some form. Etc...  For sure, we tend to overestimate us. Understanding which is our real potential will make us understand what we could really expect from Oracle.

What's the point of your

What's the point of your final question? If Oracle is the tough company you described, then they will do with Java whatever they want, no matter what the community thinks it may has to offer. The answer is irrelevant if Oracle is not interested in listening, and so is the question. It's solely up to the community to decide whether to go the Oracle way or not.

By now, we saw Oracle killing the OpenSolaris and the OpenOffice community, but not killing NetBeans or Glassfish. We saw a disgusting lawsuit against Google, the only one of the big boys in the Java camp who is really capable of radical innovation (Android = mobile Java done right, GWT = web Java done right), a further delay for Java 7, and a ridiculous nomination of Holo-whatever for the JCP. That's not very encouraging.

On the point about the

On the point about the community, I think Oracle needs it just as much. The community is where developers eventually originate whether it higher education or self taught. If Oracle is going to monetize Java, they have to have developers who can write in Java or will accept positions to do such things. I believe that means all levels such as small to medium sized businesses, and not just larger corporations but certainly including them, must be the target audience. I think if they go with the top 2000 or 1000 or so, then the end result will be Java begins to fall behind C/C++ which is still widely used. Oracle Forms can't compete with the projects using Java for all the things being done.

There are only a certain amount of developers who can be hired by these companies, and if some other technology similar to Java begins to take hold among small to mid-sized because Oracle shifts the language and JDK/JRE itself from the role it has served, then it will be a major game changer. Oracle is a good database, but for most things it is replaceable. Java too is replaceable. However, Java is used and is chosen for many projects using no other Oracle related products at this time. It is in many mobile devices. Take another technology not encumbered with a badly chosen strategy offering language simplicity, similar features and keywords, and you have your replacement offering mild reworking to port many existing libraries and projects.

Those burned by some bad strategy around Java then decide other things are replaceable such as hardware, operating systems, databases, etc. Then a number of acquisitions become wasted money essentially becoming a death nail. Oracle won't win with a Win/Lose mentality (Seven Habits of Highly Effective People).

I think Suns business model fell down around their understanding of how to organize certain systems into tiered revenue generators; some money on smaller levels, more on mid, and most on the larger side. They kept changing their plans etc. It was always something different. Too, there were too many ambiguities with licensing etc which led to the whole Apache thing.

All these pointing to no real strategy as to what of their many assets was going to generate revenue and how, and what was going to help them engage, grow, and support a community who could build and support systems on their offerings. That community doesn't just happen by itself. The JDK, an IDE, and JEE, being available for all to get, use, and learn made the community. Being able to get those things and start small companies and use them allowed companies to get big enough to buy other services etc. They just didn't know how to get those people once they were ready to move from small to mid to big.

It is good to have certain areas of business shared with the community to help foster the next generation and help model the base pieces of the overall strategy which is something they did; JDK/JRE, NetBeans, GlassFish. But, they had many other things and working standards, and they pushed them as the next big thing instead of letting that organically happen and focusing and solidifying something before moving on ultimately chopping the legs out from under folks starting to use those technologies they were pushing. They could have remained focused on their core businesses such as systems, Solaris, commercial Java licensing, and other successful Java projects where they could have monetized support or licenses (NetBeans Platform comes to mind; support and training stream) or other types of revenue.

Another large area they missed the mark was the web space and fostering hosting services etc. They could have been showing organizations how to host their products best, and in the cases where it wasn't an issue of know how, but an issue of technical reworking, they should have focused on that. They could have monetized JEE in many ways. I'm sure they probably made some money with dedicated devices, but I know nothing of those details, so will not speculate on that.

OpenSolaris: big missed deal. They could have done so many things there with the number of years they threw money and time at it. For desktop they could have worked in the ability to install capabilities which would allow them to use Windows drivers perhaps. They had deals with MS. They could have worked it out if they really thought about it. They could have worked out ideas with MS on how to best MacOS in a nix environment. They had all the hardware know how in house.

They could have done Android years ago.

Of course hind site is 20/20. This is just a mild sampling as well. But, there were many more things related to why Sun failed at what it did than the technology it created and community interactions. There just seemed to be a large lack of vision at a higher level; some of there projects were very visionary on the project level.

If Oracle is smart, they'll create an internal think tank devoted to honestly answering the question of "With all the technologies Sun had,why couldn't they pull it together, and why did they ultimately fail?" instead of trying to throw a one size fits all business model on all they acquired. At the end of the day, they now have so many things more than Oracle DBs (MySQL, Berkley, Oracle), App servers (WebLogic, GlassFish), and Java. If they can't figure out that essential question, they'll either be selling off some of these acquisitions or failing in the same areas. My guess is they are working on figuring that out; there is just a lot there to figure out, and they have a lot of new potential.