ScienceOrganizer:
Replicating scientific data

Dan Berrios: During Rotation 29, we are testing the capabilities of a system to automatically replicate data from an installation of ScienceOrganizer in the Hab to a publicly accessible installation of the ScienceOrganizer at NASA Ames. In real exploration missions, a replicated system such as ours could prove invaluable. Remotely located scientists are unlikely to have direct access to NASA networks, but will need to be able to access scientific data regarding explorers collected samples, notes, theories, and plans.

So far replication of this data has gone well. Last night, we realized that the Hab ScienceOrganizer database had started to block all replication requests from the NASA Ames ScienceOrganizer system. This is due to the somewhat tenuous nature of the communications connection to the Hab and to an anti-hacking feature of the database system we are using. The internet connection from the Hab intermittently goes down now and then. This immediately severs any replication connections initiated by the Ames machine The Hab machine stores knowledge about these broken connections, and, after a certain number of them, automatically assumes the Ames machine is a hacker attempting to break into its database (since broken connections are failed connections). We have adjusted the configuration of the Hab machine to allow for a larger number of such broken connections before blocking a host machine attempting to connect. We will have to examine further how many such broken connections are likely to occur in real exploration missions, and recommend appropriate database configurations.

Compendium:

Simon Buckingham Shum, Al Selvin & Michelle Bachler: Following discussions at the last RST SOWG meeting, we generated a 'portal map' for today's RST meeting. The idea is that it has distilled just those elements needed to prepare for this meeting: briefing notes from the RST Leader, the links to WebEx and the Meeting Replay, and critically, the web export version of the Crew's Compendium maps. This has proven to be a big hit for the RST, who previously, on waking up in the early hours on the West Coast, had to waste valuable time wading through the email torrent flowing onto their desktops, to find key links, and import the Compendium database from the Crew. Now it is collated in a single URL, clearly marked Portal Map in the subject line:

Figure 1. A Compendium 'portal map' which links the key resources needed by
an RST member to prepare for, and attend, a virtual meeting. - (Click for Detail)

Another development in the use of Compendium was that the scientists in the Crew responded to the annotations that the RST had made to their maps, providing answers and clarifications to the RST's queries. Use-cases such as these provide clues to the ways in which the scientists find it most natural work with the tools, and clarify requirements for improvements which will make exchanging materials, and interpretations of them, as fluid as possible.

Figure 2. RST Crew computer dialog - (Click for Detail)

Using Compendium as a shared canvas for collaborative analysis between the scientists on 'Mars' and their remote support team on 'Earth'. First the scientists in the crew laid out photos of rock samples, and analysed them. The RST reviewed this and raised queries, linking them into the map, which accompanied anotehr map containing their summary report. The crew then responded (yellow highlighted nodes).

A minor bug was detected in the ScienceOrganizer Reference nodes that were being generated by Brahms agents in EVA plans (broken links), which was quickly patched by Mike Scott.

Meeting Replay Tool:

Danius Michaelides & Kevin Page: For various reasons the video for Mondays crew planning meeting wasn't available in time for us generate the replay for the RST meeting the following morning - which demonstrates the sensitivities of a process spread between so many people and time zones! To get a replay out as soon as possible we've started generating a preliminary replay with a simplified timeline - with only the Compendium nodes - before later creating the more labourious speaker markup.

Meanwhile, RST members have been having some problems viewing the replay - on the whole everyone has been successful, but there are still some issues. We are trying to track these down, but there are so many variables...

Miscellaneous points:

Simon Buckingham Shum: Ironically, we again had difficulties with the most established technology that we're using, WebEx for desktop sharing. Maarten gave Simon his login details enabling him to create meetings, but every time he set up a new meeting, his browser inexplicably crashed. He was only able to log in to the meetings as a 'normal' participant, which prevented him from pasing desktop sharing capability to anyone. The RST was, however, still able to operate, navigating the web exports of the crew's plans linked from the portal map, coordinating verbally rather than via a shared display, and was able to send its feedback.