Building an "imadoki', or "nowadays" application was the topic of the panel discussion.
While the contrasting views on Openness and Open Source didn't lead to any passionate debate, it was interesting to see the two companies' differing approaches. Miyagawa-san is a big contributor to Open Source Perl projects, and while IBM also invest in open source projects, they have a corporate manual for engaging with OSS, and the projects they run themselves are subject to different licences.
The one project that has been raising it's head recently is Project Zero. Project Zero is a Java-based engine for creating RESTful web applications, that can be programmed against in either java or PHP, and is designed to give PHP and Java folks the simplicity and productivity of a Rails like experience. The project is not open source, though, and fall into IBM's Community Driven Commercial Software Development (CDCD) model. This means that the code is open for anyone to inspect and suggest corrections or improvements for, but IBM are the the only committers, so they retain control over exactly what makes it into the final build.
This approach allows them to provide a high level of warantee for their products, but I feel that the whole idea of community drive development is that in a way, the strength of the community provides the warantee, and in a more responsive way that having a central control bottleneck. If I have a specific issue with the code, then I can suggest that IBM fix it, but it still remains up to their discretion to fix it. Similarly, say that the IBM powers that be decide that they no longer consider the product "strategic" and decide to orphan it. How does that work if the community needs to continue using and supporting it?
Next Yohnemochi-san presented some slides showing how IBM planned to support what he described as an explosion in the growth of applications.
US Research presented describes what I call the "Geek Spectrum" of application developers, from hard core J2EE coders, through scripting language based Web Developers, to business users who have traditionally put together macros, but can now create mashups and hook in to other services by assembling apps from building blocks.
To enable each of these segments of the spectrum, IBM has various tools, and the aforementioned Project Zero is targetted at the upper end of the Web Developer layer.
Finally they both talked about Open Source, and there were questions about how it worked in Japan verses in the west. Miyagawa-san said that feeback for open source projects comes much quicker overseas, probably because people have more willingness to state their minds. Yonemochi-san raised some embarrassed chuckes when he said that Japan had been compared to a black hole for Open Source ... everything gets sucked in, but nothing ever comes out again. Harsh! He also said that due to the difference in geographical size, some companies in the US retain onsite engineers, where in Japan they might not, simply from the practical stand point of physical response time. Most places in Japan are reachable within a few hours, but that is not so in the US.