Red Hat to Acquire JBoss

tecker writes "Redhat.com has a banner and press release that states that it will be Red Hat that will buy JBoss and not Oracle as previously thought. The press release states "the world's leading provider of open source solutions to the enterprise, today announced that it has entered into a definitive agreement to acquire JBoss, the global leader in open source middleware. By acquiring JBoss, Red Hat expects to accelerate the shift to service-oriented architectures (SOA), by enabling the next generation of web-enabled applications running on a low-cost, open source platform." Could it be that a one company server package that will rival Microsoft's Windows Server 2003 and ASP will finally emerge?"

Comments

But what are the terms?

[liliafan]

Although I think this is an important development for java developers, I can't really see it really being a rival to Server 2003 and ASP, don't get me wrong I hate ASP and M$, but the simple fact is they have a huge market share, that just doesn't want to move, additionally they have legacy.

I would be interested to know more about the terms of the takeover, I remember reading recently that Marc let the Oracle deal drop because if/when he sold out he wanted his terms and conditions to be met.

Re:But what are the terms?

[ajakk]

Having been burnt on his first startup, I suspect that Marc got pretty good terms on the deal. JBoss has been running in the black, and their connections with some big clients could help Red Hat get more service contracts. I think that the acquisition makes sense, because it will help push Red Hat into the high end service area even more (i.e. where the real money is). I do wonder how well Red Hat will be able to manage the diverse group of people working for JBoss. I am sure that not all of them will be happy with the buyout. Considering the international nature of JBoss workforce, I suspect the Red Hat might have some difficulties managing them.

Re:But what are the terms?

[/ASCII]

PHP has an even bigger market share, by your logic they should have bought Zend. Well designed systems that are harder to master, like Rails, JBoss and ASP.Net won't drive the trivially easy web languages like ASP and PHP out of business, but there's room for everyone.

Redhat to aquire JBoss

[vv2]

So they are not buying Oracle then - settling for JBoss must be a bit of a dissapointment.

Would it...

[Natrone]

... give RedHat an instant "in" on the application server market so coveted by BEA and IBM? This seems like it could be an intersting fit, and would certainly save JBoss from extinction by Oracle (as seems to be the trend).

Re:jboss

[ajakk]

Red Hat wants the support contracts that JBoss has. That is where these companies are trying to make money. I bet that Red Hat will start offering a consolidated support contract that will offer support for both JBoss and Red Hat when you are running JBoss on Red Hat. People who are paying money for JBoss support will be more than willing to push out a couple of bucks for Red Hat support as well.

Red Hat couldn't create their own support group for the JBoss application server because of the complexity of the technology and the lack (and cost of acquiring)of people with the Java skills to understand it in-depth. Also, Red Hat didn't have the reputation of providing world-class support for Java. Now it will.

Makes sense for a service organization to do this

[Gunfighter]

Red Hat already had some enterprise Java stuff, but the middleware component just puts the icing on the cake. I think Red Hat is simply using this purchase to officially add this to their portfolio. By portfolio, I don't mean "software products", I mean their service offerings. The software has been, and will continue to be, free. It's the brains behind the operation that cost companies money. In fact, Red Hat probably already had engineers who were paid to support customers running Jboss, but now they are the "unofficial official" place to go when you want enterprise, corporate support for Jboss.

It's past time to stop looking at Red Hat as a software company and start looking at them as a service organization. This isn't surprising considering the success their RTP neighbor, Cisco, had as a service organization (and you probably thought they were a network hardware vendor all this time).

JBoss Microsoft Agreement

[ajakk]

I wonder how this will play with the JBoss and Microsoft agreement that was made in September. That deal was for Microsoft to work with JBoss so that JBoss can run better on MS servers. Clearly, having JBoss run better on Microsoft servers is against the interests of Red Hat.

Re:jboss

[LnxAddct]

Red Hat has a good history of doing nice things for open source projects, or proprietary projects that they bought and made open source. If a big supporter of open source didn't pick up JBoss, Oracle would have killed the project eventually (they have experience doing these things). One cool thing about this is that Red Hat develops GCJ (Gnu Compiler for Java) and they've got it compiling Eclipse and the Java portions of OpenOffice.Org, so I'd venture to guess that this increases the chance of JBoss running natively too which would be interesting.
Regards,
Steve

gcj

[Micah]

I assume this is good news for GCJ and/or Classpath, given Red Hat's committment to free software. Surely they will now devote many resources to making JBoss work reliably on Free Java, then we all win!

Re:$350m

[tppublic]

Actually, $350M isn't all that much. A financial analysis looks something like this:

Assume JBoss is growing at a rate equivalent to the S&P 500 (10.5%) - I'm trying to be conservative here and not get overblown about growth (since values are very sensitive to growth).

Assume RHAT wants to at least maintain its return on equity of it's stock, currently 19%. So the earnings rate on the purchase is 19% - 10.5% = 8.5%

At $350M, that means JBoss has at least $30M in profit ($350M * .085) for this to make sense.

If JBoss is growing at 20% per year and you want a 5% risk premium (accounting for uncertainty in the future of the market for middleware), then the earnings rate becomes 4% (19% + 5% - 20%), which means $14M in current income at JBoss to have it make sense for RHAT.

You can see how growth causes leverage in a price ... since:
value = earnings / (required return rate - growth rate)
... this division is part of the reason why stocks who have high growth expectations are very hard to value (at least using this method, especially when the denominator becomes negative) and why they fall so quickly from high stock prices when their earnings slow. This is why other (more complicated) models may use a higher growth rate in close years, but force the growth rate to slow in later years to the market rate - it helps to avoid the crazy value multiplication that can occur in the simple models.

Re:jboss

[LWATCDR]

"i have to admit i never liked jbosses model, give the users a nice piece of [censored] without proper documentation and then charge for the books and support. the software itself was great when i had the look at it, but the fact that you had to hack around german forums to find out some nice tricks for free, wasn't so tempting."

Exactly what piece of open source sofware have you found that has really well writen documention?

For that matter what piece of closed source software have you found that comes with really good documentation?

Oreilly makes most of it's money by documenting other peoples software.
I don't see any real difference. Heck I spent a good part of friday looking for a fix for Asterisk@home. I found it on a forum on sourceforge after a few hours of searching.
Of course I added it to the wiki but WTH didn't anyone else?

You show me any program that comes with complete documentation, tutorials, and troubleshooting guides please? I would love to see it.

Re:Will JBoss go the way of CCVS

[DdJ]

Red Hat purchased another opensource project/product a while ago called CCVS( Credit Card Verification System ) and converted it to their proprietary license before later killing the product couple years later.

Heh, that's not exactly how it went.

The original product implemented communications protocols that were owned by financial institutions.

These protocols were under heavy NDA. As a result, there was never a release of CCVS under any open soruce license. Red Hat wanted to open up the whole thing, but that would have been a violation of our contracts with those financial institutions.

In addition, there was a rigorous certificaiton process required for any software that did this stuff -- if anyone did modify the software we distributed, it would have been in violation of the finanical institutions rules to actually use it without going through a rigorous and time-consuming certification process for basically every single change to a line of code.

How do I know? Basically, I'm the guy who wrote it.

(There was more than one of us, but I designed the whole thing, and wrote the infrastructure parts, all of the telecom modules, and some of the protocol modules and language adapters. Other people wrote some protocol modules that plugged into my code, some of our language adapters, and one guy wrote our database layer.)

Some CCVS trivia:

• I ported it to PalmOS over a three day weekend once, so we'd be able to actually show it while walking around at trade shows.
• On the floor of a trade show (ALS '98?), at our booth, I ported it to the Corel Netwinder in about 15 minutes. (Yes, that was trickier than typing "make" -- if I can ever open up the source code, I can show you why.)
• Our first customer ever, and I think the only one to use the original "1.0" release of the software, was LinuxMall. Remember them? If you ordered stuff from them over the web in the mid-to-late 90s, odds are your credit card number went through my code.
• It wasn't strictly a Linux product. We had it running on SunOS, SCO, AIX, DigitalUnix, you name it. Internally, we even had it running on NeXTstep, and on Apple's "prelude to rhapsody", and on beta versions of MacOS X. If Red Hat had taken slightly longer to cancel it, we might have owned the (tiny) MacOS point-of-sale market. But it never ran on Windows.
• We made our APIs available for: C, TCL, Perl, PHP, Python, and Java. Yes, we were commercially supporting financial transaction processing APIs for all of those languages back in the 90s.

(You'll have to pardon me for going on like that. I'm kinda proud of what our little company managed to accomplish.)

Which reminds me: anyone from Red Hat (or with contacts at Red Hat) reading this? I'd love to get that source code back!

I believe I know how to make it open source today, and I'd like to take a stab at it -- and at porting it directly to today's 2.5G and 3G cell phones.

But, legally, Red Hat owns that source code, and I do not have the legal right to try to open it up without their say-so. I have been able to get responses from the folks at Main Street Softworks, but they don't have the CCVS source code or rights to it either.