Apache Marmotta is now a Top Level Apache Project
We are excited to announce that with the recent resolution of the board of directors of the Apache Software Foundation (ASF), the incubating project Apache Marmotta becomes a top-level project of the ASF.
The idea behind Apache Marmotta of providing an Open Platform for Linked Data has been many years in the making. Therefore I would like to take this opportunity to provide a short history of Apache Marmotta.
Short History of Apache Marmotta
The early days: KiWi
The core work on the technology essentially started with a EU funded project called KiWi (Knowledge in a Wiki) in 2008. KiWi investigated the idea of “Semantic Wikis” and aimed at combining easily editable web content with semantic web technologies to provide data to both human and machine users. KiWi was a very successful applied research project: the founding team of Marmotta – Sebastian Schaffert and Thomas Kurz – presented the idea of KiWi and Linked Data at a series of conferences. Highlights have been the 6th European Semantic Web Conference in Heraklion, Greece, in June 2009 where we won the “Best Demonstration Award” and the demo at the International Semantic Web Conference 2010 in Shanghai, China.
In the current source code of Marmotta you can still find some of the components that originated in the KiWi project, most notably the triple store together with its rule-based reasoner and versioning – of course now with several iterations of improvement and code rewriting to make it more stable and reliable so that now it can handle billions of triples with ease.
Teenage Years: Salzburg NewMediaLab and the Linked Media Framework
The next big step after KiWi was the “Salzburg NewMediaLab” starting 2010, where the goal was to bring these technologies from research into real use with selected industry partners. Being a typical research project, KiWi featured many very advanced and experimental ideas that were cool to show but not stable enough to be used by thousands of users. So we decided that we need to take apart the complex KiWi system and take out the technologies that were stable enough. Also, we realised that most companies already have their content management system in place, so it does not make sense to replace it with a new system but rather augment their existing system with semantic technologies. Born was the “Linked Media Framework” or LMF.
The Linked Media Framework took from KiWi the mentioned triple store and the semantic search functionality, and replaced the experimental content analysis of KiWi with the more reliable Apache Stanbol, which was anyways developed in the same team. The result was a sophisticated Linked Data application server, which is now in production use at several partners, e.g. the search engine of SalzburgerNachrichten.
At around this time (summer 2012), we also found out that there is some conceptual work going on at the World Wide Web Consortium regarding the specification of a “Linked Data Platform” that does not only allow read access to a Linked Data server but also would support creating and updating data. In a way, we realised,
they were writing a specification about something quite similar than we had already developed and running. So we decided that it would make sense to separate the Linked Data part from content analysis and search again and publish it as a separate project, as an implementation of the Linked Data Platform.
Growing up: The Apache Way
Realising that this could become a useful tool for a wide range of users and developers out there and seeing the success of Apache Stanbol, we decided that the best way would be to start a second project at the Apache Software Foundation, and contribute our existing code as a starting point. We proposed it as “Apache Linda” (for “LINked DAta”) at the ApacheCon Europe in Sinsheim (Germany) in November 2012. Sergio Fernández presented there our idea, getting good feedback, so a couple of weeks later we finally submitted the proposal for incubation.
While we thought that Apache Linda is the greatest and best name for a project ever, we quickly learned that things are never so easy: over the short history of Computer Science, there have already been several projects called “Linda”, and some of them apparently still raise negative feelings when thinking about them. So we had to find a new and better name. Now, anyone who ever had to think about a good name for a project knows how hard this can be. Long story made short: in the end we decided for “Marmotta”. Why? Because it is the Italian name for an animal that lives in the mountains around us. It digs big tunnel systems underground. It whistles. It is very social. It is very cute. Many Apache projects use animal names. And best of all: there was never a project called “Marmotta”!
So it came that Apache Marmotta became an incubation project at the Apache Software Foundation (ASF) in December 2012. We found great mentors in Andy Seaborne (of Jena), Fabian Christ (of Stanbol) and Nandana Mihindukulasooriya and started moving and polishing our code and clarifying legal issues to conform to the strict ASF guidelines. And we also got our first external committers: Raffaele Palmieri and Peter Ansell (of Sesame).
After 7 failed vote attempts, we finally managed with the first release (3.0.0-incubating) of our software under the roof of the ASF on 24th April 2013. Luckily, the second release (3.1.0-incubating) in October 2013 went much smoother and our mentors were insisting more and more that we leave the podling and graduate as top level project. 😉
Graduation: Apache Marmotta, MSc
After the 3.1.0-incubating release, things suddenly went on quickly. We decided it would be time to attempt graduation, as there were no further issues left to be clarified. We voted for Jakob Frank as Vice President, because he is most familiar with all the guidelines and with his formal background would be a good representative for the project in the ASF board. And Jakob prepared a proposal for the incubator board, which was voted on unanimously. He submitted our proposal for graduation to the ASF board, which decided on 20th November (again unanimously) that we are ready!
The Road Ahead
Plans for the future? Well, after graduation the “the serious side of life” starts. Apache Marmotta is looking for volunteers to join the growing community. Contributions in bug-reports, usage hints, documentation and of course code patches are welcome on any of the project’s mailing lists.