James Fee’s report of a shift from ESRI to Open Source for GIS server technology has stirred up some interesting discussion.
The challenge I see for ESRI is of the chicken vs. egg variety.
ArcGIS Explorer (AGX)’s ability to run geoprocessing tasks appears to be a big advantage over Google Earth. However, until there’s a critical mass of published Geoprocessing services, there’s little incentive for people to download, install and learn AGX. They’ll likely just stick with GE. Likewise, until there’s a critical mass of AGX seats installed, there is little incentive for people to purchase an AGS license, develop and publish geoprocessing services.
I suggest that ESRI provide a hosting service, perhaps similar to Amazon’s EC2. By hosting AGS for 3rd parties, ESRI could have a more flexible licensing model. Instead of charging per cpu, ESRI could charge based on some combination of cpu cycles, network i/o, disk storage. Flexible pricing would encourage experimentation, people could develop geoprocessing services and quickly publish them. If the gp service catches on, the developer could buy more cpu resources from the ESRI datacenter. Perhaps ArcSDE databases could be available within the Datacenter and provided to the geoprocessing services on a fee basis. That way, if I have a geoprocessing service that buffers all the Starbucks and intersect it with available office space, I could do it without going through http (just a straight SDE connection).
Remember, a chicken is just an egg’s way of creating another egg.