I was just reading on Adrian Coyler’s blog that Spring’s dm Server is moving to Eclipse.org. Cutting through all the blah blah blah, this is what I came up with as the reason for this move:
“At SpringSource we know that open source development and community involvement can play a huge role in evolving simple, pragmatic solutions that enable a technology to bridge from early adopter to mainstream usage. We know because it is a path we have successfully taken many times. In creating the Virgo project at Eclipse.org, we seek to accelerate the journey of the dm Server and of enterprise OSGi along this path.”
What I extrapolate from this is that they want to accelerate adoption. I have a simple fundamental question:
Does it really matter if you host your project at an OSS organization like Eclipse.org or Apache?
I’ve been living in the Java OSS community for about 9 years now. IMO, based on my experiences with JBoss, the only thing that Eclipse.org or Apache really gives you is an established brand to promote your project. I’ve always felt that a move to Apache or Eclipse, while may be beneficial in the short run with a bump in your adoption curve, in the long run it is bad for both your project and ultimately your business. Why? Because you lose control of both your brand and the governance of your project. Brands cost money and time to build. Governance introduces the inefficiencies of any bureaucracy (see my previous blogs on problems at Apache.org). So, with both of those disadvantages, I question the advantages of moving dmServer to Eclipse.org. JBoss has never had a problem starting a new project and building new communities on our own. Neither has Spring for that matter. Our brands are strong enough so that we don’t need the bump of an Apache or Eclipse to drive adoption.
The only real benefit I see to move to Eclipse or Apache is if you’re trying to build a consortium of companies that build off of a base core technology. I really don’t see that happening with dmServer, even if that is there goal. I know we wouldn’t use the technology. Would Oracle? IBM? SpringSource is no longer the little guy. VMWare is a serious competitor to all of us.
Standardization more important
IMO, a better route to grow adoption is standardization. Hibernate got a huge bump in adoption by aligning itself under the JPA banner and EE 5. We are seeing the same with Seam->CDI->Weld and Bean Validation. Standardization drives adoption because it frees competing companies from having to depend on a competitor’s code base. Since the APIs are public, vendors can provide their own implementations and value add on their own terms. At the same time, standardization gives an easier entry point for vendors that want to enter into the space and gives users the peace of mind that they have less vendor lock-in. Just because something is open source doesn’t mean that it is immune to lock-in. The lock-in here is to the implementation. Standardization breeds multiple implementations.
With dmServer, yeah, it is built upon the OSGi specification (and OSGi implementations), but that is just not the same. You have to bring your innovations to a standards body to create the ecosystem.
What do you think?
So please help me optimize my personal algorithm for professional open source. Do you think it really matter if you bring your project to an established organization if you yourself are already established? I say no…
Bill the Plumber 
Jan 12, 2010 @ 16:04:23
I don’t disagree with your main thrust, but I’m not convinced that standardisation helps that much. Hibernate classic worked, hibernate-under-EJB5 was subtly different from both Hibernate and bits of the EJB spec, primarily due to bits missing from the spec. I’m a fan of stable code, and test suites -no need for standards bodies to get involved. JUnit, for example. And the OSS right-to-fork deals with the lock in problem. That said, if Hibernate use went up from being standardised, then it was the right action from a business perspective.
-I believe Jetty went into Eclipse because they wanted a bit more legal backing, Eclipse suited them better than the ASF.
-No idea about dm Server.
Joining an ecosystem helps if you want tighter integration. That’s why multiple database-atop-Hadoop projects are in Apache Now: HBase and Cassandra. Yet there are differently licensed things outside, and everyone uses JUnit.
What’s interesting about Eclipse is how it regularly tries to move out of an IDE into some platform play. Previously that’s never gone far. No idea about this time, but like you say, nobody wants to be owned by VMWare.
Jan 12, 2010 @ 16:38:37
Seems like a good strategy, in my opinion. Its a form of marketing by getting your brand all over the place. People who know Eclipse but aren’t necessarily acquainted with Spring will be introduced via Eclipse.org. Also, it gets Eclipse.org developers more interested in contributing to Spring. Standardization may be more important, as you have noted, but even better is to follow both paths. I think JBoss would be wise to make a similar move. I have been a fan for a long time but I see community interest waning.
Jan 13, 2010 @ 14:05:17
REx, then why not do the same with the entire Spring codebase? Bring the entire thing to Eclipse?
Jan 13, 2010 @ 14:58:02
Because it might be perceived as Spring trying to strongarm the Eclipse community. They are obviously looking at this as a way to form a partnership (albeit mainly for show).
Jan 12, 2010 @ 20:32:21
I disagree with you.
With any large OSS project there are also the legal issues of IPR (i.e. patents), copyright and openness to consider. Both Apache and Eclipse come with well defined rules and practices for these topics. Most big corporations are comfortable with both working inside these communities and/or depending on products developed by this community. Having your companies OSS product under the umbrella of either organization means that your company and Eclipse/Apache are committed to providing an open development eco system and compliance with either organizations strict rules on IPR and licensing. Not a small matter.
Secondly, standardization is a slow process that can take multiple years. Simply growing a defacto standard around an open system and an ecosystem that any stakeholder can participate in is both much more efficient short term and long term increases the likelyhood that any technology implemented actually ends up being standardized (because for most real life standards, an open reference implementation is required/desirable).
So especially if your goal is to standardize, you should consider doing this from an established & well respected OSS community.
Neither standardization nor open sourcing grows adoption by itself. Adoption is grown by providing some added value.
Jan 12, 2010 @ 23:55:21
No, I don’t think so. If you want standardization then you cannot have that without governance. If you have established a great product by innovating inside a closed brand then the next logical step is to move it into the public domain.
Moving it to Eclipse should free Spring resources so that they can move to the next chapter of innovation. And sure that worked since already people from Tasktop have jumped aboard to help move dm into mainstream usage. Innovation and an established name are all great ego boosts but don’t cause a wrinkle in the spacetime continuum. When things really are ready for the big stage then you must either be MS or IBM or join a large scale open source community.
Jan 13, 2010 @ 14:17:59
Standardization efforts need governance, IP sharing rules, etc. An OSS project cannot function efficiently within a democracy. Imagine if your governance model shut down your contributions because the governing body decided your company was dominating the project. (This happened in Geronimo).
Also, this idea that resources are going to magically appear and maintain dm Server just because it is at Eclipse is poppycock. dmServer is currently already open source. Are you saying that SpringSource doesn’t accept or try to recruit independent contributors?
And I totally disagree that you have to be MS or IBM to drive something and make it mainstream. I know because I lived it. I’m sure SpringSource feels the same. (And WTF is Tasktop anyways).
Jan 13, 2010 @ 15:02:56
Bill, one big thing you missed is the license change. Going from the GPL to a weak copy-left license like the EPL is a smart move. First, it will help with adoption as there are people out there that don’t want anything to do with the GPL. Second, it will open up other companies to contribute and potentially build on the core dmServer code with their own proprietary extensions. It’s a smart move in my opinion.
Let’s see in a year how much it pays off.
Oh, Tasktop is the guys behind Mylyn
http://tasktop.com/
Jan 13, 2010 @ 20:44:13
Re: Standardization efforts –> My point exactly, thanks for confirming
Re: Resource going to magically appear –> Did you even look at the project proposal? 20% of the contributors have magically appeared.
Re: Totally disagree –> Yeah well, congrats, most projects are not that lucky. And I suggest that you lookup Tasktop.
Jan 13, 2010 @ 22:30:01
@Chris: License is totally orthogonal to to where the project is hosted. Plus, I think the number of users that won’t touch GPL code is very very very very small. Otherwise Linux (and other FOSS projects) wouldn’t be so popular and RHT wouldn’t have a business (a rather good business I must say).
@Wim: I did look up tasktop. I’m supposed to be impressed or something? I’m sure that most marginally successful open source projects can name at least one impressive user.
BTW, I hope you guys don’t mind arguing, but so far I don’t see a lot of proof for your statements. I can name a bunch on my end.
Jan 13, 2010 @ 23:05:59
Re: supposed to be impressed or something? –> No. It was a reply on your argument that resources would not magically appear (right where you said poppycock). However, ppl from Tasktop stepped in right away. You want proof? Go check the project proposal.
Re: BTW, I hope you guys don’t mind arguing –> Au contraire, I love it.
Jan 13, 2010 @ 23:22:21
Debate is fun ;)?
“License is totally orthogonal to to where the project is hosted.” ?
Oh, that’s why you see GPL projects at Apache.org and Eclipse.org? It’s not as orthogonal as you think.
Also, you have been shielded within Red Hat if you think companies don’t care about the GPL. That said, Red Hat has done very well for itself given the GPL and I commend and respect them for that.
Jan 14, 2010 @ 13:42:02
The fact that Apache.org or Eclipse.org refuses GPL, LGPL, or even ASL licensed projects is their problem, not the open source project’s. My point was that adoption bumps from licensing are orthogonal to adoption bumps from hosting at an OSS Org. For example, I’d bet you’d get a small credibility bump from being hosted at codehaus.org, but they don’t give a crap what license you use.
Again, only an insignificant amount of companies care about GPL (or LGPL) for that matter. If it was significant RHT wouldn’t be a $650+million dollar business. IMO, its the Apache blowhards that pump up this myth. In reality it just doesn’t exist.
Jan 25, 2010 @ 19:54:12
There is a question on whether a commercial organization like JBoss or SpringSource want to host something that isn’t core to their business. IT makes sense to send those out and collaborate on them.