Link – Image Server
On Monday during the Plenary we will introduce a new type of Image Server. It takes a different approach to serving up large (TB) quantities of Images. The data is held in the raw file format (no need to pre-process, or load into a DBMS). The Server builds an index so that the files can be treated as a continuous image map. Services can be authored to process the images on the fly as they are served up to desktop or web clients.
My first reaction is cool, but I don’t see how this would replace me loading rasters into my ArcSDE as I do now. I’m guessing it costs less than ArcSDE and a DBMS, but how much less?
Damn, I guess I’ll just have to wait until Monday.
[update] OK, I’ve thought about this a little bit. How are they doing this without a DBMS? I’m guessing they are reading in the cell information from the rasters and then storing it somewhere (xml?). I guess Image Server processes the requests from clients, but how does Image Server know what to return?
I’m guessing this is an outgrowth of abandoning the Personal Geodatabase MDB format as we all have learned it pretty much doesn’t work as well as shapefiles (let alone Coverages, GRIDs and Tins). I’m actually excited about a new data format, but I’d like to know more about how we’ll be able to “interchange” the datasets. I assume it will be back to the venerable e00? 😉