Comments on the Concept

From Vic Loudon 20081218

Standing on the platform

First of all, thank you for letting me see this revised outline, and for inviting further thoughts. Thinking about your project in more detail, I am impressed, though not totally clear about its overall implications (HRAs are allowed to be confused). I like what you have written, but worry slightly about what BGS might do with it.

1. I am not sure how you interpret ‘platform’. ‘Platform’ can carry a whole range of meanings. It is sometimes used in a broad political sense, signifying the set of principles, assumptions, ideas and policies on which a party is standing, and how they propose to implement them if elected. In IT, it sometimes seems to mean something as narrow as an Applications Program Interface. I would like to think you have the broader meaning in mind.

2. If taken in the broad sense, the platform could be a basis for one of the more important sea-changes in the history of BGS. The scoping study might establish the context for placing BGS geological cyber-information firmly in the wider panorama of Earth systems science and e-science; for exploring what is possible, what is desirable, and what is not; and recommending how the necessary evolution and migration could best be encouraged.

3. On the other hand, if ‘platform’ is interpreted too narrowly, the study could be an unfortunate diversion from the mainstream. The development of IT in BGS is littered with the corpses of initiatives intended to bring comprehensive answers to the problems of the day (examples: Rokdoc, G-Exec, the Experimental Cartography Unit, Mimer, WordPerfect, and shelves loaded with highly specific standards that came unstuck). They all had value at the time: in opening up new ideas, helping to organise work, and moving users along a learning curve. But their advocates, both developers and users, tended to fall in love with their creations, and could not accept that they should be thrown away when they had served their purpose and reached their sell-by date. This had the malign effect of periodically leaving BGS to spin off into comfortable backwaters, move away from their customers, and lose the momentum of mainstream development. In the end, the most successful implementations seemed to come from widely accepted systems with strong commercial backing and customer acceptance (examples from: Oracle, Microsoft, ESRI, Google); followed by continually moving on and migrating legacy information whenever a better or improved product achieved wide acceptance.

4. If we start from a top-down view and keep in mind the e-science aspects of the cyber-environment, then BGS can establish the underlying systems that will be needed as we move towards a service-oriented knowledge utility (SOKU). The ‘cloud’ of computing services would then automatically solve the difficulties mentioned in 3 above. To take your example, the modules for generating 3D geological models would be identified and standard interfaces (between data and procedures) would be defined. Any software that met the interfacing requirements could then be used. As a general open platform develops, GSI3D, for example, could be used wherever in the Environmental Sciences it met the specifications. As the users are in charge, with the ability to select modules from any compatible source, their decisions in aggregate will drive the Darwinian evolution of the system. Open source software could thrive in that environment. All that is required is creation and acceptance by the global community involved in Earth and environmental science of a well-defined comprehensive and extensible systems model, and the necessary standards to accompany it. That should be easy for BGS to push, much easier than in 1835 having to push global acceptance of the concept of the geological map.

5. Nevertheless, identifying future products and directions, consolidating progress, and in general navigating the waters of fast-moving science and technology, demands an understanding of developments. This has to be based on thought, investigation, and experimental trial (in order of increasing cost). I entirely agree with John Laxton’s point that there is a need to identify specific process models as test-beds, but, as he points out, a high-level conceptual model is needed first. Before looking at models of geological processes, the underlying scientific needs must be clarified as comprehensively as possible. For example, understanding and creating models of geological processes and events are an essential step in bonding geology to other aspects of Earth systems science. But in designing, for example, process models that are relevant in the context of past global warming episodes, we may have also to consider how they can be built into a system for comparison with Earth systems events at the present day.

6. I hope that your scoping project might lead to follow-up studies to create mechanisms to establish and maintain the form, content, and development of a modelling platform which is:

  • compatible with a comprehensive view of wider aspects of the geological cyber-strategy, with other current initiatives (examples: OneMap, GeoSci ML), and with legacy information across geology and related sciences
  • compatible with a geological cyber-environment (infrastructure providing end-to-end support in geological investigations) seen as a well-specified, coherent part of the whole-Earth systems cyber-environment
  • extensible to accommodate new approaches that unleash the potential of IT rather than just replicating existing procedures (examples: multi-resolution heterogeneous models, emergent systems)
  • well-defined to provide global standards to guide software developers and encourage global compatibility and interoperability in geological investigations
  • flexible to keep in step with progress in e-science (examples: Grid, ‘cloud’, SOKU).
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License