Open Source Solutions

Clarise and I met with Gabriele Ruffatti and Grazia Cazzin, from Engineering Ingegneria Informatica S.p.A. to discuss, and as Clarise wrote, get demos of SpagoBI [a full BI suite], and see the new Spagic EAI tool that was announced at OSBC2007, and the Spago framework [think Spring, only different].

Once he returned to Padova, Italia, Gabriele sent out some thoughts that came to him after participating in the open source business conference. He kindly gave me permission to publish them here.

"I attended Eben Moglen’s speach at the last Open Source Business Conference in San Francisco (May, 22nd). Hearing so many times: “stand up for freedom”, I thought: "I’m here now to promote free (in Europe we say libre as well) open source projects in different domains (Spago, java framework; SpagoBI, Business Intelligence; Spagic, SOA environment; Spago4Q, a Business Intelligence domain specific solutions) at an event mainly presenting commercial open source". While there, I’ve promoted new enhancements of SpagoBI, the Business Intelligence Free Platform, comparing it to other commercial open source choices such as Pentaho and Jasper and I’ve thought again: "what is the key differentiator? Stand up for freedom…"

"A participant said to me: “Guy: you’ve made a mistake. SpagoBI is not the Business Intelligence Free Platform; it’s the Free Business Intelligence Platform”. Was he right? I think no. This is the key: SpagoBI is a free platform, commercially supported, offering a new choice despite many other commercial open source products claiming more effectiveness to the market.

"What I mean by "commercial open source product" is: a solution claiming to be open source, claiming to have a community supporting it, but offering closed add-ons for enterprise adoption with a proprietary approach to the market (i.e.: acquisition of projects and IPs to strengthen its stack, dual licensing approach, aggressive marketing). Is it effective for the BI domain? Yes, probably it is.

"What I mean by a "free platform" is: free design, free collaboration, free assembling, free adoption. Not just because the license (SpagoBI adopts the GNU LGPL license), but because the efforts are in the software development improvement direction rather than in marketing proposition, in collaborations with different projects and solutions instead of acquisitions, in integrations with many free, open source, and also closed solutions to achieve the most effective solution for the user, maintaining a totally free code base core, instead of closed specific add-ons. Is it effective for the BI domain? I really don’t know, but the market, or a very new market knows the answer.

"Stand up for freedom. Is it “against” a wide commercial adoption? Is it “against” the growth of a strong business ecosystem? I know that Europe, Asia and South America are looking for free/libre software; Public Administrations are looking for free/libre software; new domain specific applications can be built thanks to free/libre software (Spago4Q is a just a first sample of it). At OSBC in USA, I’ve heard not only of commercial open source packages, but also of “building the right solution with an assemble mindset, opposite to a buy mindset”. It’s just a new choice: with an open mind, you can build open applications offering freedom to invent, to share your own knowledge, to assemble the right solution for people’s needs."

-- Is freedom the right approach to the Business Intelligence domain? from Gabriele Ruffatti, Director - Architectures & Consulting, Engineering Ingegneria Informatica, S.p.A.

F/LOSS - Free/Libre… Open… "Free as in speech, not as in beer" is well known. But how open is open? This is a question that continues to be debated in terms of whether or not a company is truly open source, pureblood or mudblood. Andy Astor puts forth the rationale behind EnterpriseDB's strategy around openness [NB: EnterpriseDB is a company that is sometimes cited as a company using open source without being open source]. We tend to be fairly liberal in our acceptance of what is an open source company [including EnterpriseDB], while the market for enterprise open source defines itself and "commercial open source" has more denotation than connotation within that market, and I rather like the direction being taken in Commercial Open Source, cited above, by Carlo Daffara. Others may be more stringent in accepting a company as truly F/LOSS or not.

Most open source projects avoid locking their users into their product by following or providing open standards and open APIs, and/or using a plug-in, or framework, stack or platform architecture. Such an architecture allows the users and the projects community to more easily customize the product to suit their unique needs, and to more easily give-back to the project, without necessarily becoming an approved committer to the core code. We're currently following 47 open source projects that provide components that can be used to build solutions for data management and data analytics, data warehousing and business intelligence. We're following 5 open source BI suites, most of which use or can use the same core components. JasperReports for the reporting engine and Mondrian for the OLAP engine are common but not all-pervasive. All five suites provide a convenient stack of components that work together "out-of-the-box", but provide varying facility for including additional or replacement engines and components. We're still working out what is the real differentiator among all these tools, but Gabriele has a very good point: the more freedom, the more openness, the better.

Personally, I've been working in data analysis for nigh onto thirty years. The reason that we've been looking for open source solutions for decision support, data warehousing, data mining, BI, GIS, EAI, BPM, [insert latest hyped term here], etc, is that this area requires more customization and user involvement than any other enterprise application. The beginning of this millennium saw a trickle of open source projects addressing BI needs, and 2005 saw an explosion in the number of projects. Open source BI provides the flexibility that is always needed in satisfying the user needs for data analysis, visualization and sharing. The greater the freedom, the greater the openness, the greater the flexibility… the easier it is to satisfy the users. This is why data management, analytics and BI is going to be one the most successful markets for enterprise open source, bar none.