null
The EclipseZone Interview Series lets you get to know the individuals and projects behind Eclipse. In-depth and personal interviews deliver back stage access to the people that make Eclipse possible.
Recently EclipseZone had the pleasure of speaking with Todd Williams, currently a member of the Eclipse Foundation Board of Directors and one of the founders of Genuitec. Genuitec is best known for MyEclipseIDE, a powerful environment for Java client and server development layered on top of the Eclipse IDE. It combines open and closed source with a reputation for excellent service and a very inexpensive (less than $30) yearly subscription.
[Todd] Genuitec was launched in 2001 by its three founders (Wayne Parrott, Maher Masri, and myself) as a consulting company that specialized in building large-scale systems in several industry verticals. However, upon the release of Eclipse we saw the potential for a tremendous shift in the technology market. So we quickly adapted to the change by becoming a dedicated Eclipse technology company.
Now, Genuitec builds and sells products based on Eclipse while helping other companies do the same through our consulting services practice. Our principal location is in Dallas Texas, but the company operates as a distributed team, with employees located in several different states and countries.
Well, we certainly didn't sit down one day and decide that geographic dispersion would be the secret of our success, since distance traditionally makes things more difficult. But when we refocused the company around Eclipse it quickly became evident that experienced developers would be very hard to find. It was only by opening our search to the world that we could find people with the skills we needed. So, it was the hiring of talented employees that forced us into a decentralized corporate model. However, in retrospect it really has worked quite well and our employees view the ability to work from a location of their own choosing as a benefit. With all the communication technologies now available, distributed models are definitely workable. They just take a bit more effort, by all parties, to keep the communication flowing appropriately.
We were doing a lot of J2EE consulting when Eclipse 1.0 was first released to the public in November 2001. Our consultants looked at Eclipse as a possible replacement for their current IDE and Eclipse was the logical successor given IBM's decision to discontinue VisualAge for Java.
We began writing plug-ins in early 2002 for our consulting work, primarily to launch and debug application servers from Eclipse. We called them the Eclipse Application Server Integration Environment (EASIE) plug-ins. Since we weren't in the tools business at that time, we simply gave them away from our website. After we served 100,000 downloads in only a few months, it became obvious that there was an emerging market that was not being served. So, we reevaluated our business model and began the initial planning for MyEclipse at the end of 2002, which we then launched in mid-2003.
While we were building some Eclipse plug-ins in early 2002, Wayne Parrott and I had a long discussion that even though Eclipse was only thought of as a Java IDE at that time, it contained an extensible framework that could be used as the architectural base of a wide array of desktop applications. Based on some early experimentation, we found that with specific code modifications on Eclipse 2 we could indeed remove all of its IDE artifacts. So, if the idea of using Eclipse to build other types of applications caught on, there would be ample opportunity for consulting services. Currently all of our consulting work is Eclipse related.
You're absolutely correct. There wasn't a market at all; just the germ of an idea. Fortunately, around that time Marc Erickson, who was running marketing for the old Eclipse consortium, gave me a call to see if I had any ideas for articles to help publicize Eclipse. I gave him a couple of concepts, but the one he jumped on was the idea of using Eclipse as the basis for general purpose applications. So, I wrote the initial draft and after a few editing round trips we'd finished the article in a couple of weeks. It was called The Case for Using Eclipse Technology in General Purpose Applications" and Marc then set out to get it placed in one of the Java tech magazines. That's when it got interesting.
At the time, Marc was an IBM employee and when he shopped the article around IBM, before getting it placed in the press, it wasn't well received at all. I believe it had something to do with clouding the way Eclipse was being positioned at the time. In any event, Marc gave me a call and asked if I'd strike his name from the article and disavow any involvement on his part. I did as he requested and we distributed the article from the Genuitec website beginning in June 2002. Genuitec tried to get the article published subsequently, but the feedback we kept getting from editors was that the subject wasn't mainstream enough.
Mainstream or not, the article was heavily downloaded from our site and the calls from clients began to come in. Consulting has been going well ever since. In fact, if you search Google for "Eclipse application framework", the article is still in the top five links.
Eventually the article did get published. I officially declared the idea mainstream when Dr. Dobbs Journal (among others) reprinted it in 2004. If you follow that link you'll notice that Marc is properly attributed this time since he has retired from IBM.
Genuitec was a founding member of the Eclipse Foundation and joined the initial IBM-led consortium in early 2003. I was elected to the Board of Directors, to represent the Add-In Provider community in 2004, was re-elected in 2005. Currently I chair the membership committee.
Since I represent the add-in provider community I use my board position principally to:
Like the vast majority Add-In Provider members, Genuitec does not provide any full-time committers to Eclipse. However, we contribute in many other ways such as writing articles and whitepapers, evangalizing Eclipse technology, and serving on various Foundation councils, committees and working groups.
Eclipse is unique amongst open source communities because it explicitly and consciously fosters commercial adoption of its technologies, and works towards creating a commercially profitable ecosystem by providing an extensible architecture and flexible licensing model. At the present time, Eclipse is currently comprised of about 50 projects and sub-projects that cover a wide spectrum of technology that are designed to be used as the basis of commercial products.
MyEclipse begins where the Eclipse SDK ends and is one of several commercial tier 1 J2EE IDEs that run as an Eclipse SDK product extension. Currently, MyEclipse is comprised of about 200 additional plug-ins, excluding Eclipse itself. A portion of the infrastructure of MyEclipse comes from a few of the non-platform Eclipse projects, for example EMF / GEF / Draw2d, just like they do for most of the other Eclipse-based IDEs. However, the majority of our product is architected, designed, and developed by our internal team.
That's a great question because I think there's some confusion about both MyEclipse and our business model. The base difference is that MyEclipse is an integrated product, from its architecture through its design and implementation. Distributions, like Yoxos and others, are bundles of many "point solutions" that happen to be available in open source. Our approach to developing MyEclipse is actually the opposite of bundling.
We design the features of MyEclipse by listening to our users and by utilizing the expertise we garnered through our J2EE and Eclipse RCP consulting practices. When we're about to go into the implementation phase on a feature set, we research what is available commercially and in open source to see if anything will give us a head start in the implementation phase. Sometimes we find helpful open source components, but often we don't. In either event, we then build out the features as designed. As you can see, this is the opposite of a bundling approach since the open source research happens as a side-effect of our development process, not as the central theme of our business model.
In summary, a bit over two years ago, we set out to incrementally build a highly-competitive J2EE IDE and make it available at a price point that any developer in the world can afford. As a result of our development and business models, we're able to build highly-innovative features into MyEclipse while making it available for an extremely low yearly fee ($29.95), including all updates and support.
This is a perfect example where common sense prevailed over market research. The management team had a gut wrenching decision to make in 2002 as we were trying to reposition the company towards Eclipse development. As I recall, Maher, Wayne and I were sitting in a Starbuck's fleshing out the ideas behind MyEclipse and the subject of price came up. Earlier in the conversation we had already agreed that MyEclipse would be different from other J2EE IDEs because it would be affordable for individuals, not just big corporations. But, that didn't give us "the number". For that we had to personalize it. How much would we pay as individuals? How much would we have paid when we were 25? How much would our employees pay out of their own pocket?
The more we thought about it, $29.95 seemed too low, if anything, but certainly not too high. In retrospect, it seems it was just about perfect so we've left it right where we initially set it over two years ago. Our customers appreciate it and we can reinvest more resources into the product since we're profitable at that price point. It's definitely win-win.
As a small company, we have a different cost structure than our competitors in the J2EE IDE space. And, given our pricing we can sell in volume all over the world, not just in the US. Additionally, we have highly motivated and dedicated development and support teams that listen to our users and genuinely want to help them solve problems effectively. [See Rick Ross's recent experience with MyEclipseIDE support in this article. -Ed] Our focus has always been, and will always be, doing our very best to continuously add value to ensure that our users remain pleased with their purchase and become increasingly productive over time. Doing the right thing by our users is the key to our success; it's no more complicated than that.
Todd Williams is Genuitec's Vice President of Technology and is responsible for the organization's Eclipse technology direction and consulting practices. He has twenty years of experience in the development of computing infrastructures, large scale distributed software architectures, and the optimization of development processes and techniques. In addition, Todd is Genuitec's representative to the Eclipse Foundation and currently holds a seat on the Board of Directors and chairs the membership committee.
You are viewing a mobilized version of this site...
View original page here