February 2008


An aspect of the open source software nirvana has been the end of Microsoft’s dominance. Well, anyone who sees Linux, OpenOffice or another OSS project/product as the death nail in Microsoft’s coffin is not living in reality. It pains me to say this; it really does.

When you see Microsoft borrowing ideas from the OSS movement, it’s probably best stop smelling the roses and pay attention. For instance, Microsoft’s Sam Ramji has an interesting post on how OSS has influenced Windows Server 2008. Six areas Microsoft has learned from include:

  • Modular architectures
  • Programming language agnostic
  • Feedback-driven development
  • Built-for-purpose systems
  • Sysadmins who write code
  • Standards-based communication

Sam writes:

“Overall, we’ve learned and continue to learn from open source development principles. These are making their way into the mindset, development practices, and ultimately into the products we bring to market. As all of the different organizations in IT continue to evolve, we’ll learn from each others’ best practices and make increasingly better software. As in science, this incremental improvement will move all of us forward.”

InfoWorld’s product review of Windows Server 2008 scored it an 8.5/10 and described it as an essential upgrade:

“Microsoft’s slimmer and stronger server OS, bolstered by virtualization, networking, and security advances, is an upgrade that IT can’t refuse, a 200-pound gorilla that eats commercial Linux”

Throw eggs at me if you like. But this should scare any OSS proponent. It seems like the folks at Redmond have been busy while the OSS movement has been prematurely readying Microsoft’s eulogy.

I hope I’m wrong. But Microsoft simply appears to be meeting the challenge of OSS better than OSS appears to be meeting the challenge of displacing Microsoft.

PS: The term ‘Freetards’ is used with attribution to FSJ.

Many of you know that OpenLogic offers consolidated technical support, updates and indemnification for hundreds of open source projects. Earlier this year, OSS luminaries such as Shaun (of JBoss fame) and Rod (of SpringSource fame) questioned whether customers could receive the same level of technical support from vendors such as OpenLogic versus the OSS project owner.

Well, OpenLogic just released news on the state of their business:

  • OpenLogic tripled its number of customers. OpenLogic now has 3 companies in the Fortune 10 and many more in the Fortune 500.
  • The top 5 projects that OpenLogic provided consolidated support for in 2007 were PostgreSQL, JBoss, Tomcat, Apache HTTP Server, and Hibernate.
  • In 2007, the OpenLogic Expert Community grew to 160 members. Field experts who OpenLogic pays on a per-incident basis helped resolve almost half of OpenLogic’s support issues in 2007, with the remainder being resolved by OpenLogic’s in house experts.
  • Projects in OpenLogic’s Certified Library more than doubled to 380 projects.

It’s interesting to see that customers are going to OpenLogic for JBoss and Hibernate support versus using Red Hat.

As OSS products become more mature, what value proposition will OSS project owners offer to balance the needs of customers seeking one number to call for the 10s of OSS projects in use at their company? Don’t get me wrong, I doubt that Red Hat or SpringSource is going to close up shop anytime soon. But, there is a case to be made for consolidated support providers such as OpenLogic.

As with most things, there’s likely room for both models…

I’m finally getting a chance to read up on the Microsoft interop news today. It really sucked to see the headlines all day, but not have a chance to read them until now.

While the words “open source” are mentioned in the announcement and on Bill Hiff’s post, this is really about open APIs, not open standards or open source. (Maybe ‘not’ is too harsh; maybe I should say ‘and significantly less about’).

Obviously Microsoft is opening up their APIs because it makes business sense. Google, Facebook, Amazon, etc. have already made the business case for open APIs to closed software. If you ask me, the interop announcement virtually eliminates (on paper) the argument that Microsoft will lock you and your data up. Hmm…have we seen this playbook before? Maybe from OSS vendors? “Use our product and you won’t get locked in. You have the source code so you can always go elsewhere if you’re not satisfied with us.” Understandably, Microsoft’s position isn’t as eloquent. However, it does counter a potential barrier to remaining a Microsoft customer. Let me be clear, I do not think Microsoft’s interop announcement actually protects against lock-in. It seems like Cote would agree:

“Overall, the thrust of the press release is that Microsoft is going to make the documentation and access to its software better (easier? Doubtful. See Joel’s piece), that is, more interoperable.”

But, it doesn’t matter what I (Joel or Cote) think. What matters is how Microsoft can market the anti-lock-in angle and how customers will respond.

I have never thought that OSS would be the end of Microsoft or the commercial software market. The fact that Microsoft is borrowing (in their own unique way) from the OSS playbook should scare OSS “freetards” straight.

Sometimes it pays to be late with “breaking news” ;-) For those of us that weren’t at JBoss World 2008, here is a great presentation that explains the RHQ announcement:

“RHQ is a combined effort to provide management infrastructure

It is not a management product

Management technologies utilize the same types of infrastructure that are rebuilt over and over

  • Inventory
  • Common agent infrastructure
  • Fine grained security and audit
  • Integration APIs
  • Plugin extensions for new product support
  • Reporting

RHQ delivers these core infrastructure services in an open source model as building blocks to be utilized in management products”

RHQ (which is GPL’d) will be used by Red Hat/JBoss products such as JON. However, any vendor could build a RHQ plugin. For example, vendor XYZ’s plugin would “go deep vertically” in terms of managing/administering their own product. Additionally, a management product that supports RHQ would provide horizontal views/management of multiple products, including the product from vendor XYZ.

It’ll be interesting to see if RHQ gets traction beyond RH/JBoss. And how it integrates/competes/etc. with the integration and interoperability goals of the Open Management Consortium.

When I read Dana’s post about leaving Jonathan & Sun alone, I couldn’t help but side with Dana. But then I read Amanda McPherson (Marketing director at the Linux Foundation) original post and now I’m torn. Then I read comments on Amanda’s blog from Sun employees, and I’m even more confused. (BTW, I urge you to read Amanda’s post…she makes some great points).

The core of this discussion boils down to whether it is “okay” for Sun to compete against Linux with OpenSolaris. I have written in the past that I believe it is perfectly valid for Sun, or any company to compete with Linux or any OSS project/product. I am a fan of competition. I believe it helps all parties raise their game.

How is the OpenSolaris vs. Linux discussion different from MuleSource vs. WSO2, JBoss App Server vs. Apache Geronimo or JasperSoft vs. Pentaho, etc?

In the past, Mike Dolan (a fellow IBMer) commented that the issue isn’t whether Sun should compete, but why they don’t collaborate:

“Instead of “competition”, think about what could happen if Sun worked proactively and in a contributory manner to help fight “Bug #1”? (See Ubuntu bug database…). If Sun’s so relevant and its technologies are so great, why waste time fighting a community that consists of HP, IBM, Dell, Oracle, SAP, Cisco, Google, BEA, and thousands of others? Why not join in that community and provide your value-add? In my opinion, competing for OS lock-in is “so 90’s” (as my youngest sister would say)…”

I can’t argue with Mike from a pure “OSS religion” standpoint. But, regardless of what Sun wants us to believe, pragmatism is behind their OpenSolaris vs. Linux strategy more than “OSS religion”. Few vendors would be able to walk away from the significant revenue that Solaris drives to Sun. And remember, Linux grew largely at the expense of Unix (Solaris was the #1 Unix OS vendor). It’s a tough position for Sun to be in…..especially since they want OSS to spell SUN.

Personally, I encourage Sun to continue competing against Linux. Let the market decide.

Everyone knows that I’m an IBMer in the software division that competes with JBoss. I have a lot of respect for the folks leading JBoss, and hope that has come through in my public interactions with guys like Bill, Sacha & Shaun.

I’ve said it before, and I’ll say it again. JBoss is good for the middleware market. They (and other OSS middleware vendors) keep all the big vendors at the top of our game.

But as I read story after story about JBoss targeting a 50% share of enterprise middleware workloads by 2015, I was left scratching my head. First off, most stories just seemed to accept the “target” without questioning the likelihood. About the only story that separated aspiration from likelihood was by Dan Farber at BTL. Kudos Dan!

Red Hat’s “Enterprise Acceleration” initiative, which underpins this 50% goal, consists of:

1] Comprehensive Enterprise Middleware Portfolio
2] Enterprise Products – From Community Development to JBoss Enterprise
3] Enterprise Acceleration Center

JBoss has been executing on #1 & #2 on this list for some time. The only “new news” would be #3. Come on Sacha, I wanted to hear something new out of JBoss World! ;-) What’s the Enterprise Acceleration Center really about? According to Red Hat, key elements currently planned include:

  • Performance Tuning Lab – performance benchmarking, testing, best practice guides Interoperability Lab – interoperability testing with other environments and products
  • Live Certification Center – ISV and customers can test their applications on JBoss Enterprise Middleware and proactively adapt to new releases
  • Migration Lab – processes, partners, services and best practices to transition from other software to JBoss Enterprise Middleware

Will the combination of #1, #2 & #3 drive the conversion from users to customers?

Anywho, I wish the folks at JBoss the best of luck towards this 50% goal, just don’t expect to get there without a fight ;-)

Roberto recently spoke to SourceForge Community Manager Ross Turk. Ross gave a rundown on what SF has learned with SF Marketplace:

“First, we learned that people are interested in the idea. People are responding to it in pretty large numbers; growing numbers, in fact, and I think that’s good.

Second, we learned that there are a few types of transactions that people seem to want to do that our system doesn’t support. For example, people who want to sell services by the hour are working around the lack of that ability by creating listings for a single hour of service and dealing with the discrepancy in purchase price with the buyer directly. Adding the capability to have per-incident, per-hour, and per-project pricing would be useful to a lot of people.”

When I first heard of SF Marketplace, I couldn’t think of what services I would want to purchase from the developer(s) of a product I found at SF. Obviously I’d get support for things like JBoss or Spring via the appropriate vendor, but the majority of ‘stuff’ on SF isn’t of the Spring kind. For example, would I really ever want to pay for technical support for Gallery 2? (I’m technical enough and can Google for answers with the best of them).

But I wasn’t thinking broadly enough. I may not need support for Gallery 2, but maybe I want to integrate phpBB & Gallery 2 and throw in CAPTCHA on my site. Google can help, but maybe I’m in a rush. Or maybe I need a custom site/app developed using various projects on SF. Well, it looks like SF.net user faisi (and many others) can help.

I had imagined developer X, a contributor to project XYZ would offer services for project XYZ via SF Marketplace. That is one option. It also appears that developers A, B & C, who may or may not be contributors to project XYZ, are offering services around project XYZ. This could position SF Marketplace as a competitor to Guru.com or Scriptlance.com down the road. Interesting…

I hadn’t considered SF Marketplace growing in this fashion. But hey, if users are happy, well done SF!

Next Page »