Pre-conference seminars:An Overview of Developing ArcGIS Applications
Building Your Own Geography Network
Introduction to ArcIMS Configuration and AdministrationTechnical workshops:
ArcGIS General: Fundamentals of ArcGIS, ArcSDE, and the Geodatabase
ArcSDE: Administration for SQL Server
Statistical Methods for Spatial Data Analysis
Cartographic Design
Geostatistical Analyst
Real-World Lessons on ArcIMS Implementation and Development
3D Analyst and ArcGlobe
Designing Geodatabases
Installing and Configuring ArcIMS
Writing Your First VBA Macro
Installing and Administering ArcGIS Software
Building 3D Animations
ArcGIS Server Introduction
ArcGIS has come a long way in the two years since my last Users Conference: ArcGIS 8.3 is widely used (though our current installed version is still 8.1), and many of the sessions this year showed features of 9.0, due to ship in a few months. Government and industry users of GIS have developed the geodatabase server environments (based on ArcSDE) that were on the horizon two years ago, and many users are extending the architecture by programming in various languages and environments (VB, C#, VBA, C++, the .NET framework). The question for liberal arts colleges and consortia is how to leap aboard this moving train of GIS development, and how to link spatial data into courses and research and other activities.
Some of the things I had thought were on the horizon for W&L (and for ACS and NITLE) turn out to be, very simply, beyond our capabilities (ArcSDE is certainly in that category --and anything else that requires real database management), and therefore I must regretfully conclude that we and other liberal arts colleges have a reasonable prospect of using only a fraction of the power of ESRI's products. Consortia might be able to provide the requisite support, but I see no indication that they are ready or willing to develop the necessary external funding to establish or maintain the support staff positions.
The upshot: we can continue to provide desktop-level access to spatial data for presentation and basic analysis, but hosting Web services beyond the experimental level and developing spatial data servers for ourselves or for consortia just can't be done without support for professional-level RDB apps (SQL Server, Oracle, etc.).
The most hopeful directions seem to be
Geography networks are an important model for solving the "how-can-I-find...?" question for data and basemaps, and for distributing data to remote users. GeoData.gov ("your one stop for federal, state and local geographic data") and ESRI's geographynetwork.com are two prominent examples, tantamount to map services. Such services are generally built on ArcIMS, but multiple licenses (one for each server) are required to support anything beyond basic experimentation. It seems clear that an entity like NITLE could develop and host a geography network for the members of the several consortia, but equally clear that this prospect is in the distant future.
ArcIMS is the distribution technology that seems most likely to serve our needs in the short run, though all we can manage without much more support and investment is proof-of-concept experiments. ims2.wlu.edu houses Skip's experiments with ArcIMS, and this summer's R.E. Lee project on the Civil War is my maiden effort to set up and run extensions to what Skip began. Scaling up to a real service, to supply multiple users, would require multiple dedicated servers and ArcSDE --again, something that might happen at the consortial level, but couldn't be developed without substantial external funding and appropriate support. The security issues of experiments are bone-chilling to contemplate: see How Web Servers Help Attackers . Clearly we need to look into how well protected ims2 is, before we go much farther with experiments.