Slashdot Mirror
Oracle to buy JBoss (and others)
tfritsch writes "According to a story at News.com it looks like Oracle's shopping spree is to continue. The JBoss acquisition could be big - what does it mean for the future of the JBoss Application Server?" From the article: "Oracle makes the majority of its revenue from its database and applications business. And it has its own line of Java middleware, which competes with JBoss' software, and a set of Java developer tools. However, Oracle has been warming up to open-source products, including Zend's PHP development tools, over the past year because its corporate customers are increasingly using open source software, according to company executives. "
They lay people off to buy JBoss.
Ryan - http://www.thecosmotron.com/
being as guy running jboss was the former vice president of most of oracle's app server development team (bluestone/hp), this is kind of a funny and somewhat welcome suprise.
Since JBoss is private, Marc would have to be willing to sell JBoss to Larry. Larry can't just write a check, get regulatory approval and be done. There is no way Marc will sell his baby... I think he is much more interested in JBoss someday being bigger than IBM & Oracle. The world is moving toward software as a service. JBoss is positioned to be the king of that world. Marc knows this. 10 years down the road, no one will be paying for enterprise software licenses. Marc sees this and won't let even $400 million get in the way of JBoss being king of that world. I probably sound like a Marc loving lunatic. But we have to be honest. Marc created a virus that's exponentially eating away at Oracle, IBM, etc's business models. That virus can only be stopped if Marc sells. I've seen the smile on his face when he talks about the virus he created. By the time JBoss is public and purchasable by Larry, even Larry won't be able to afford it.
(Reposted from my comment on Javalobby)
Brush up your resume.
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
Don't believe for a minute that Oracle would purchase JBoss to "help it shift customers to a subscriber-based model". Oracle already has a superior J2EE server based on Orion technology. Far more likely is that Oracle wants to pull another PeopleSoft aquisition. They'll buy up JBoss, kill the company, then let the product die on the vine. All while pushing how "Open Source Friendly" they've become.
Javascript + Nintendo DSi = DSiCade
Anyone out there using Oracle's OID to do LDAP, and their servlet container, in a single app? Anyone using OID + JBoss? Any idea how that app could be improved by integration inside Oracle?
Know any docs on switching existing LDAP/servlet installations to Oracle with OID, to prepare for Oracle's apparently increased servlet support?
--
make install -not war
I've been pleased with Oracle's JDeveloper; writing an extension for it has been interesting and the Oracle folks have been quite helpful.
The Army reading list
I've long operated under the assumption that any decent (or even semi-decent) piece of "freeware" (free as in beer, but not as in speech) for Windows will eventually sell out and become "shareware" and/or conventional commercial software. Likewise, I've assumed that any decent piece of "shareware" will slowly go the route of full commercialization. This assumption has served me fairly well. (Examples of this pattern: PowerArchiver used to be freeware; now it's shareware. Paint Shop Pro used to be shareware; now, it's being sold in stores.)
Am I now going to have to start assuming that any decent OSS/FS project will eventually sell out?
With spending like this, exactly what are "conservatives" conserving?
Hey Oracle the 90's called they want their bubble back.
"This isn't a study in computer science, its a study in human behavior"
No not a troll. Anybody that has had to use any Oracle software other than the DBMS will tell you that it is garbage. Maybe they want to use open source software so they can sell complete solutions that actually work.
Please sign petition to restore sanity to our banking system!!!
http://financialpetition.org/
And we all remember the last time a company went out and bought up a bunch of companies trying to hack together a bigger brand and comprehensive product lineup, rather than take the time to properly acquire and integrate their product lines...
Oracle is a quite good company producing quality database applications.
The problem with them? They don't give a rat's ass about security. 600+, 800+ days of unfixed exploits? Who cares! Their security track record is much worse than that of Microsoft's.
The people who fork out a lot of cash to Oracle could rightfully demand that they receive quick fixes for these things.
Oracle teaming with PHP? The worst security nightmare ever. PHP is absolutely craptastic from a security viewpoint (insecure default configuration, etc.), for example the mail() function makes it the favorite of spammers, because you can use it to spam a lot with it - because the mail() function's broken implementation allows spammers to send out mail in the thousands. Working around it is possible, but cumbersome - 99% of the people using the function doesn't even know about the issue, so its a spam-haven.
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
Am I now going to have to start assuming that any decent OSS/FS project will eventually sell out?
Maybe this is just the way of business, who knows. People do want to make money, even from their labors of love. But the question I pose is simple: can't the "sell-out" software simply fork at the point of the acquisition? It's not like you can put open source software back in the can. All you can do is restrict it going forward.
Let's take JBoss as an example. What's to prevent JBoss developers (or anyone) from coming out with "JHoncho" based on the pre-Oracle version of JBoss and fork from that point? Oracle ends up buying the JBoss name and not much more, unless the developers want to work for them.
we'll all be yelling "Geronimo.....!!!!!"
(rimshot)
Wah wah "evil corporations" "poor workers" "outsourcing" blah blah blah.
One thing that seems to be overlooked is that with productivity rises, it takes fewer employees to do the same amount of work. The same is true after a merger, where it's redundant (no pun intended) to have two shipping departments or two sales forces.
I've been laid off several times in the last six years (once on Christmas Eve), and it's never been a big deal. I'm not saying it's been "fun" but if you have a rational savings plan to build a contingency fund, you should be able make it during the times you're laod off. I have sympathy for folks who are losing their jobs, having been there myself, but I also know this isn't the end of the world. I hope they do, too.
You can look at a layoff as a crisis or as an opportunity. Your choice.
Interested in a Flash-based MAME front end? Visit mame.danzbb.com
Oracle's Ellison girded investors for an open-source push at a Feb. 8 conference in Santa Monica, Calif., saying, "We are moving aggressively into open source. We are embracing it. We are not going to fight this trend. We think if we're clever, we can make it work to our advantage."
Yea by buying them out, that's a real bear hug or embrace.:-)
Scott McNealy to Michael: "Suck my Sun!" Michael Dell to Scott : "Lick my Dell!"
Cost of making competitor to JBoss and launching it into the market > $400M
You can build and market a hell of a lot of software for that...
http://jfin.org/jFin pure java open source financial library
...more than it is about the App Server. It's interesting that with just one shot, Oracle will practically "own" EJB3 persistence.
And I guess the JBoss guys (Fleury at the forefront) are more than happy to finally cash in, big time.
This was at one of the Sun conferences in Toronto and Oracle was one of the major sponsors. Well anyway.. it seems that Oracle and Sun are working closely for the EJB3.0 implementation - especially the database abstraction layers and ORM tools. So maybe this has something to do with Hibernate?
_Vishal www.squad9.com
If this happens, I'm ditching JBoss. Period. Oracle is lousy at software. I've used their license of Orion before and at 10k versus 1200, they are shady at best.
They cannot seem to drop a piece of code. They just keep adding and adding and adding until they are shipping you a stack of CD's two inches thick.
So someone, tell me, what's the next open source EJB server in the line?
You are checking your backups, aren't you?
Breakfast served all day!
I am not sure why another company would want to buy JBoss.
In its time it was very innovative with two things. First, making EJB type properties available to POJOs (properties like security, transactions, remoting). Second, they pioneered the business model of selling services based on a free product, which encouraged very wide-spread adoption. Both of these were controversial at the time and JBoss should be applauded for showing us the way.
However, the problem is now many other companies do the same thing. Big application server companies give away free copies, at least for development teams. Java itself is moving toward making EJB type properties available to POJOs. On top of all this, over the last few years there has been a clear trend to move away from EJBs, favoring instead something like a Tomcat/Spring approach for J2EE applications, and, in other cases, the even lighter LAMP stack.
It seems to me a few years ago JBoss would have been a great purchase, but right now I am not so sure.
FREE - Java, J2EE and Ajax Audiobooks for Software Developers - www.DeveloperAdvantage.com
BEA? I think parent poster meant "JRockit", not "JBoss". Which, by the way, is a damn fine JVM for running servers. (BTW, the sun linux-amd64 jvm sucked big time - had to resort to running a 32-bit chroot with the 32-bit JVM to get a decent performance out of eclipse...)
A great opportunity for hypertension, strokes and heart attacks.
So I don't understand people's fear. If Oracle buys it and turns it into crap, you can always fork the latest source prior to the acquisition, right? Am I missing something?
I note in TFA that JBoss is unprofitable. So Marc might sell because he doesn't want to put anymore money into it.
If it takes the 10 years you are talking about, where is he going to get the investors?
It reminds me of a joke about a Vermont farmer who wins $1 millon in the lottery.
When asked what he will do, he says:
"Well I guess I'll keep farming until the money runs out."
If you are losing money, you can't keep paying the developers forever.
With J2EE, every year or two there is a new version of the standard, so JBoss
would not keep its value without continued development.
"We can't solve problems by using the same kind of thinking we used when we created them." -- Albert Einstein
JBoss is a business. It has VC backing that wants to make money to put back in its fund. As long as the money is right, it is there to be sold, and there is nothing wrong with that :)
Dion,
I am pretty sure that JBoss did not give controlling interest to the VC. So it's not really up to the VC.
-James
Spring and Hibernate, yes. Struts, yes. Velocity is not widely used.
And let's also note that it's not about "no J2EE", it's about "no EJB 2.x". Java EE consists of many important frameworks: Servlets, JSP & JSTL, the WAR/EAR deployment model, JMS, JCA, JTA, the various JAX api's, etc.
-Stu
I can unequivolcally state that I am the leading Slashdot poster with Batman in my name. That statement doesn't generate revenue or otherwise help me in any useful way.
This is something I wish more people understood. Being the "leader" doesn't really mean much for business unless it gives you power of some sort. That power is either influence or money.
The Linux Kernel team and Apache Foundation, for example, have power through influence. Redhat, on the other hand, arguably has power because of its position as the Linux revenue leader.
JBoss has neither revenue nor major influence. Most uses of open source J2EE only require a servlet container, which JBoss doesn't provide! Hence the real influential power in the open source Java space is Apache's Tomcat. And generally speaking, I find JBoss is out there but not very prevalent in existing IBM or BEA customers. And I find people usually rely on it for EJB, not JMS or XA transactions, or JCA.
Unless you've got some hard numbers to back up your assertions, I'm afraid that the massive, OSS, J2EE market simply doesn't exist.
In terms of revenue, you're absolutely correct. But if you count usage, I would say that Apache Tomcat likely beats every commercial or OSS vendor by a wide margin.. but then again, it's only a JSP/servlet contianer. (Note I work for BEA but I don't speak for them)
-Stu
I mean granted, Hibernate is newer and cleaner, but TOPLink is the old faithful ORM with great features. If they merged the two projects that would be good. Most of TOPLink's team still remains in Ottawa, I believe, so they're independent from the rest of Oracle's middleware team.
-Stu
You're right. I got it wrong.
That's what I get for reading the artcile. The BussinessWeekOnline article stated that JBoss was unprofitable.
But searching for other sources backs up you're point that they have always been profitible.
"That's very different. Never mind."
E. Litella
"We can't solve problems by using the same kind of thinking we used when we created them." -- Albert Einstein
Has anyone given thought to the notion that a motivating factor for this purchase is to acquire and control the Arjuna distributed transaction control infrastructure that JBoss just acquired [sorry for the PDF] and plans to Open-Source this quarter?
meh.
Gee--the two most difficult support organizations join forces. I'm glad I let my JBoss support contract lapse... .
Boss has fairly high revenue for a 150 person company, but since it isn't public knowledge it is hard to say. They are profitable with 175 employees, so you know JBoss has to have at least modest revenue. Not IBM numbers, but still.
Agree on this count - my point wasn't that they weren't a viable company, it was that they're so small that their revenue makes them effectively irrelevant. To put this in perspective, BEA is considered a pip-squeak compared to the big 4 (SAP, Microsoft, IBM, and Oracle) with it's $1.2 billion in annual revenue. BEA's headcount is considered relatively low by industry standards for its revenue & license runrate (it's around 3200 employees). Compared to JBoss, this is almost 20x larger.
Their power is mostly that of influence, which admittedly has been historically high, but now I feel is something they've squandered to great lengths through a combination of the attitude of their public personas and the astrotufing compain.
On the other hand, JBoss has been the dominant player in the enterprise Java space this last year. Not only does JBoss have #1 market share among application servers
Disagree. Market share is share of MARKET. Market implies revenue first and foremost. Secondly, "usage share" (for lack of a better term) must be something emperically measurable with reasonable repeatability. JBoss downloads are not a reliable indicator of share, nor are magazine surveys or analyst surveys.
The major exception to this rule has historically been free web servers, as Netcraft (and anyone for that matter) can crawl the Internet for this. Since JBoss doesn't ship a servlet container, this doesn't work.
they have been the major contributor to the Java EE 5 (EJB3) specifications which are the most important enterprise Java thing happening now.
Disagree. EJB 3, while a useful standard, is competitively irrelevant -- it does nothing to differentiate vendors. Most of the influence on EJB 3 came from two sources -- the popularity of Spring-like dependency injection (which is from Interface 21, who have nothing to do with JBoss) and Hibernate.
Yes, the Hibernate folks like Gavin are JBoss employees and major contributors to the persistence spec , but firstly, this has nothing to do with the JBoss Server, this has to do with the popularity of HIbernate. Secondly, in my opinion this whole persistence thing is (to quote Don Box) the Java world's Vietnam. They can't stop throwing resources into the black whole when other communities (.NET for one) don't even bother and are just as productive.
The most interesting things happening in the Java world right now, from my perspective, are things like Azul, Eclipse, AspectJ, Spring, WebLogic Real Time, JBI, JSR-168, the SIP servlet stuff, etc. EJB3 is important but it's catch-up with existing alternatives.
And I'd also say the the even bigger trend in the industry goes way beyond Java - it is the focus on integration, SOA, portals, etc., which JBoss really doesn't have much influence on. Building isolated web applications connected to databases is boring, the market is saturated, and there likely will be better/more productive ways of doing that soon. The real pain will be service-enabling legacy systems, high speed, high reliability messaging, mainframe integration, data transformation, etc.
I think it's fair to say that JBoss is one of the top few companies shaping the Java landscape now - and they are certainly more of an influence than their competitors (IBM and BEA).
This is completely contrary to my perspective on the marketplace. Whose influence are you talking about? Spec leading? Sales? Marketing? I just don't agree. From my perspective, JBoss is widely regarded as company with some interesting innovations in their server and some very imporant innovations in other products (HIbernate) , tarred by unethical and arrogant business practices. This is by both developers (read Hani's Bile Blog -- and note
-Stu
While I tend to prefer Hibernate these days, I have had a lot of success with TOPlink through 2002 (when I stopped using it) in several projects. THe mapping workbench always sucked, but it wasn't a big deal to create the mappings in XML or Java (which is what you'd do in HIbernate anyway). I've found that TOPLink usually generates quite good SQL. You have to poke it with a stick at times, but you have to poke HIbernate a lot too.
TOPLink is also completely non-intrusive to one's class hierarchy (at least on the projects where I've worked on it) -- it used reflection to scan for changes back in the day. HIbernate is admittedly much faster here.
One anecdote on this..... I did an informal poll of a number of sr. financial services architects & developers in London last November (including some very public figures & bloggers), several of them rated TOPLink and Kodo above Hibernate in terms of overall features and performance. I admit, even I was surprised at this (I would peg Hibernate as an equal or superior). So, your dev VP is in good company. TOPLink was the only game in town until Kodo and Hibernate, really.
-Stu
Wrong thread, scumbag.