David De Roure started his presentation with recording what he called the eight signs of time:

• Everyday researchers doing everyday research: e-Science - should not we talk about e-Research instead? David De Roure said - is about everyday researchers doing everyday research, not just about the "heroic science". Everyday researchers are using whatever tools are available to do their work better. For instance one chemical Lab is using bogging software as a lab diary.
• A data-centric perspective, like researchers: research is more and more about acquiring data with all kinds of equipment, from repositories, digital libraries, and creating new value from that. Hence work flows and provenance of data is important. This is not "anti-computation", but the design of the research process is around data.
• Collaborative and participatory: the social process of science has changed in the digital age, with new collbaorative tools like wiki's and blogs.
• Benefiting from the scale of digital science activity to support science: collaboration websites like OpenWetWare and myExperiment bring together many researchers that share and collaborate.
• Increasingly open: There are open journals, open access to data, science commons, that provide easier access to scientific information
• Better not Perfect: Researchers are looking for a tool that helps them at this moment. They are not looking for the perfect ICT tool that promises them even greater improvements in the future.
• Empowering researchers: The success stories come from researchers who have learned to use ICT. It does not work if you provide them with a tool that takes away autonomy.
• About pervasive computing: e-Science is about the intersection of the digital worlds and the physical worlds. Hence there is a great interest in integrating sensor Grids and mobile devices into the e-Science process.

The problem with the Grid is that it does not fit these signs-of-time observations.

The "heroic Grid" infrastructures like the UK eScience Grid, and TeraGrid are not being widely adopted as the heroic Grid community expected. This type of Grid is also mainly about computation and not about data. Instead of focusing on collaboration and participation, the heroic Grid is deeply routed into a service provider mindset. Also the heroic Grid is looking for a perfect well-engineered solution according to David De Roure. It is also difficult to get on the heroic Grid. You have to become member of a virtual organisation, and acquire a Grid certificate in a not so easy procedure. This gives the heroic Grid a feel of institutional control. The common access to the heroic Grid today is through a portal, which is not quite right for the next generation.

The problem with the heroic Grid is that it also sees the flow from heroic science to mass adoption by researchers as a long (15 year) process with one way flow from computer science to the mass research. David De Roure says what is needed is not only a roll-out of technologies, but a roll-in of researchers in the process so that a service delivery process is turned into a knowledge co-production.

On the other hand, the Web 2.0 design patterns very well fit the e-Science requirements.

Everyday researchers doing everyday research is covered by the "Long tail" of Web 2.0. A data-centric perspective, like researchers is covered by the "Data is the Next Intel Inside" pattern. Collaborative and participatory is covered by "Users add value". Benefiting from the scale of digital science activity is covered by the "Network effects by default". Better not Perfect by "The Perpetual Beta", software is continuously being improved, not in long cycles. Empowering researchers is covered by "Cooperate, don't Control". About pervasive computing by the "Software above the level of the single device" pattern.

So there definitely is room for Web 2.0 in the e-Science process. But where does the Grid fit in? One proposal is to have the Grid controlling the resources and having the Web 2.0 as a kind of user friendly interface. However, David De Roure proposes a more moderate role for the Grid: the heroic Grid is in fact a rather tightly coupled distributed system. Other resources are better interfaced by HPC computing, cloud computing, etc. So in his view the Globus based Grid is only one part of the resource controlling eco system. Web 2.0 provides an interface to all of them.

One specific development that makes new every day research possible is, according to David De Roure, work flows. Work flows automate routine processes. As an example he mentioned the desktop client Taverna that runs on a PC, but can automatically contact a lot of data bases and other services.

What is the main message the audience should take home? David De Roure said one should remember:

• e-Research is about doing new research
• Grid is just one part of the solution. It works fine in a limited application area
• Users are not just consumers of infrastructure, but they can contribute to the developments
• Web 2.0 is a set of design patterns. They fit the new e-Science developments
• Think about Web 2.0 coupling Grid and other services
• Work flows make e-Science easier, and Web 2.0 makes work flows easier.