Open Innovation at the Open Source Junction

This guest post has been contributed by Ross Gardler of OpenDirective. Ross is Vice President of Community Development at The Apache Software Foundation and a mentor at the Outercurve Foundation. Ross has been active in open development of open source software for over ten years.

Over the last couple of years OSS Watch has run a series of three events called the Open Source Junction (OSJ). These events aimed to bring together academia and the commercial sector to foster communication, collaboration and open innovation and were kicked off by Gabriel Hanganu when I was a member of the OSS Watch team. The second two editions were held after I had left to start OpenDirective. Having understood OSS Watch’s vision for these I continued to participate.

With an explicit goal of building a network to surface new opportunities for collaboration between academic and commercial participants OSS Watch had decided to tackle some significant challenges. Consequently, the OSS Watch team planned to work between events to help build on these opportunities. Each subsequent event sought to broaden these participation networks even further.

Since OpenDirective was created with the express intention of helping to take research outputs to market through open innovation the OSJ objectives are well aligned to our own strategic goals. In this post, at the request of OSS Watch, I will revisit these three events and explore some of the initiatives that can, at least in part, be credited to one or more of the Open Source Junction events. In so doing I hope to demonstrate the value of the Open Source Junction events. In a subsequent post I’ll make some recommendations for future events of this type.

The first Open Source Junction was held on 29-30 March, 2011 and had a focus on “open source cross-platform mobile apps” and “how to manage the co-production of cross-platform mobile apps in an open development context”. The structure of the event was a fairly comfortable one for workshop regulars. Presentations were interspersed with occasional “interactive” sessions that focused on introducing participants interests and skills. Being a two day event the evening social activities were important in building a sense of participation, but in the main this event sought to build a base level of knowledge about one another as well as a basic understanding of common collaboration practices in open source software development. OSS Watch’s own event report concluded that “given the enthusiastic response to the event and the firm prospect of future collaboration, the community’s life force is already looking strong.” Certainly there were a number of attendees who indicated they had new opportunities to explore.

By the time the second OSJ came around on 5-6 July 2011 some of these opportunities had progressed and small sub-communities was forming within the larger OSJ network. The second event focused a little more tightly on specific areas of interest as identified by participants in the first OSJ. The organisers defined the focus as “context-aware mobile technologies”. The OSS Watch report for this event says “a huge array of services [were] offered and requested… aided by the format of the event, with numerous … interactive sessions built in”. Indeed this event was more interactive than the first. This was achieved by having more sessions designed specifically to get people thinking beyond their self-defined boundaries of expertise.

By the end of the second event the stage had been set, the main actors had been identified and preparations for the third event began. It was at this third event the script would be written and roles cast.

Open Source Junction 3 was held almost a year after the first on 20-21 March 2012. This event, once again, built on previous events. However, this time the goal was not to narrow the focus further but instead to introduce a new, but related, topic and potential collaborators who specialised in this topic area. The third event therefore had a broad focus of “mobile technologies and the cloud”. It brought together the mobile skills identified in previous events with a broader set of skills around cloud based delivery systems. The goal was to bring a new angle to the existing community which would offer up previously undiscovered opportunities.

In order to realise the OSJ objectives this final event was significantly more interactive than earlier editions, with eight fully interactive sessions compared to two at OSJ1 and four at OSJ2. The first day consisted of a series of ice-breakers and introduction sessions leading to a second day that was designed to generate concrete ideas for collaboration. These sessions were carefully managed to ensure that where possible, the right people connected to the right people. Whilst there was plenty of opportunity for chance meetings the OSS Watch team had pre-identified overlaps between many participants and worked hard to help individuals discover and explore these potential touch points.

OSS Watch’s event report concluded that “OSJ3 built on the solid foundations of the past events with connections that had been made at previous OSJs, such as Cloud4All and Webinos, taking concrete steps forwards. With the ever increasing focus on interactivity at this event, many new connections, such as linking MAAVIS and Cellularity, were formed. Over the coming months OSS Watch will seek to assist in the further development of these relationships in order to ensure that they continue to feel that sharing early, sharing often is both comfortable and productive. For starters, OSS Watch is working with the OMELETTE and WSO2 teams to explore whether MyCocktail should enter the Apache Software Foundation alongside the Wookie and Rave (which themselves benefited from previous OSS Watch support).”

So it would seem the three events built up a number of promises of collaboration. In researching this post I decided to visit the opportunities identified in the final OSS Watch report. My concern was that events such as these often result in potentials that are not followed up on. In reality full in-boxes quickly force out new plans despite our best intentions. However, on this occasion I’m pleased to report that many of the hottest opportunities have indeed taken measurable steps forward.

The potential for MyCocktail to enter the Apache Software Foundation alongside Wookie and Rave (both of which were represented at all three OSJ’s) was examined by OSS Watch and WSO2 (a participant in OSJ3). At the time of writing this has not progressed significantly, initial explorations indicated that although the project is very interesting there is little motivation within the MyCocktail team. The MyCocktail code is open source and, in theory, could be reused easily by others. However, in practice, long term maintenance is important and it would seem that in this case long term maintenance is unlikely to be present. This is the only one of the three concrete opportunities identified in the report that has not led to a success.

Cloud4All and Webinos have engaged in a number of strategic discussions. These have resulted in the Cloud4All team exploring the use of an Android node.js port delivered by the Webinos team. This work is still at the experimental stages but implementation was started at a recent Cloud4All hackathon event in Austria. Unfortunately Webinos team members were unable to be present at this event although it is still hoped that active cross-pollination, as opposed to passive code-sharing, is possible. This kind of reuse, particularly if active collaboration is undertaken, will bring significant benefits to both projects. It is clear that the initial discovery of this potential collaboration was initially identified and explored as a result of representatives of each project attending all three OSJ events. Indeed, OSS Watch facilitated a “connect session” at OSJ3.

MAAVIS was also represented at all three events, however their emerging partnership with Cellularity was a result of the final broadening of scope in OSJ3 since Cellularity was a newcomer at that time. The Cellularity approach to building a form of private cloud infrastructure using custom hardware and open source software sparked ideas for delivery of the MAAVIS product. A subsequent demonstration of Cellularity’s commercial hardware offering, facilitated by OSS Watch, allowed this concept to be further examined. Also present at this meeting were representatives of the JISC funded DataFlow project.

DataFlow was not present at any OSJ event but synergies were subsequently identified by OSS Watch and OpenDirective staff when analysing event outputs. Both the MAAVIS and DataBank opportunities are still being explored and OpenDirective have submitted proposals for product prototyping to the Technology Strategy Board as a result of these discussions. If successful this proposal will see the prototyping of a new open hardware and software framework that could integrate all three projects in a final marketable product. Watch this space for updates in the coming months.

It is likely that other opportunities were identified by participants at the OSJ events, I’ve only explored the ones I am aware of. However, even if this is not the case and the successes identified above are the only direct results of this work by OSS Watch I think it is safe to say that the Open Source Junction events were a significant success. As OSS Watch enters a new funding period in the coming month I hope to see more events like these.