Meeting Replay Tool:

Since the hab crew were busy and didn't have time for a meeting, we didn't have to create a meeting replay. Instead this gave us the opportunity to work on some outstanding issues with the replay software.

Working with our colleagues at the Open University, we've completed and tested control of the replay from Compendium. This allows the RST chair to select a node in Compendium and cause everybody's replay to jump to the point in the meeting where the node arose. This seems to be working satisfactorarily now.

Meanwhile, we've been having battling with problems caused by the different versions of web-browsers, java and javascript - the replay client has grown in complexity as features have been added - problems often occur at the boundaries between these technologies.

Compendium:

Compendium mediated the second RST SOWG meeting, and without the meeting replay, we worked solely from the extensive mapset from the Crew (which Maarten will no doubt comment on in his report of Hab activities).

The XML exports and attachments from the hab crew were getting very large (~35mb), which on the one hand was great because they contained a tremendous amount of rich data, but on the other hand were causing each RST member to perform very lengthy downloads and imports into their individual Compendium databases. In addition, one of the RST members was not able to get Compendium going on their PCs and also not able to join the Webex meetings, so was only able to participate on the audio portion of the telecon.

We realized that there was no need for the RST members to have to download and install individually, because the RST Compendium facilitators could create graphical HTML exports of the whole database and all images and other attached files, and load them on a web server. Simon and Al cooked some examples of this during the RST session and gave the URLs to the RST members on the telecon. Everyone liked them and agreed that for review before the SOWGs, team members could just navigate through the maps on the web without losing time doing the downloads and imports.

This morning's session was also the first time we used summary maps created by the RST Compendium facilitators in advance (in the hour before the telecon started, via a transatlantic phone call, IM session, and Webex). This created a mini-"portal" of all the maps, images, analyses, web links, and other materials that the RST would need during the session. We did this very quickly before the session so not all the possible data and links were created, but we will be well positioned for next time.

The biggest time-lag in creating and publishing the materials for this morning's SOWG was in transferring the web exports transatlantically from Al's home PC, via DSL, to the KMI website. We realized it would be much speedier for Simon to create the maps, do the exports, and publish directly from his office at KMI, rather than losing time for the FTP upload.

It's clear that it would be highly desirable for an automated process to take the Compendium database and supporting materials and create and publish the web export without manual intervention. A good direction for future development. Even more compelling would be a closer integration with Science Organizer as the web publishing repositiory, so that Compendium web links and other metadata could become part of the SO repository without manual work.

There was the same glitch today as for our first telecon: the WebEx meeting booked by Maarten could not be started without him (as owner). We're still waiting for someone else to be given this authority. Luckily (again) Al was able to set up a WebEx meeting courtesy of Verizon. But if we had not had this luck, we would yet again have been stymied.

The CoAKTinG version of Compendium does screen-sharing via the Jabber protocol, but (a) not everyone has this installed, and (b) there is no question that we need to share an unpredictable set of applications with colleagues, so desktop sharing is the only answer.

Once the RST meeting got going, there was a high level of excitement and enthusiasm as we looked through Brent and Abby's extensively documented Compendium analysis maps, rich with images arranged in creative ways to commnicate their inter-relationships.

We all felt that the discussion, insights, and questions added to the maps by the RST took the discourse to a new level of engagement, and that the potential of easier/quicker RST pre-review of the maps via the web would help to ramp up even higher.

Speaking as Compendium developers and researchers (Al+Simon), it was very exciting to see many of the tools and methods we've been working on for many years coming together in today's meeting. Maarten's use of tags and templates, Brent and Abby's use of images, the ease of finding the right materials from Science Organizer to integrate into new summary maps, and the fluid way Simon, Shannon, Stacy, and Melissa were able to navigate maps and images, tag new content, interweave their analyses with the hab crew's work, and publish all the results as both graphical web content and a Word document immediately after the meeting -- not to mention that all the work of all the teams is available for further work in the Compendium database -- all worked together seamlessly.

As Shannon observed, these methods allow science teams to move detailed scientific work from individual observations to group collaboration, even over time and distance, immediately, without having to wait for "translations" between paper notebooks and other media, and without losing meaning and nuance. See her report for a more detailed view from the RST Lead's standpoint.

ScienceOrganizer:

During this rotation we are testing the ScienceOrganizer Communication Agent (SOCA), a new agent that represents the ScienceOrganizer system in Mobile Agents. The SOCA acts to store data collected by other agents during simulations in ScienceOrganizer. This preserves a memory of the simulations by generating a trail of the important data collected, all linked to the activities and plans for the simulations in ScienceOrganizer.

During the first EVA, which sent the ERA out to scout the terrain around the Hab, the SOCA performed flawlessly, storing knowledge about all the activities that the ERA undertook, as well as the panoramic photos the robot captured. This allowed the crew and remote scientists to get a lay of the land and plan subsequent EVAs. The remote scientists, in particular, were able to follow the collection of this data in real time by monitoring the flow of the data into ScienceOrganizer. This of course was very exciting.

During the second EVA, some of the collected data was not loaded into ScienceOrganizer and we are currently trying to determine the cause or causes. So far, it appears to be due to a combination of problems, and teasing out and rectifying these exact causes is an important and valuable task for the rest of this rotation. The remaining EVAs will undoubtedly benefit from this work.