Hi Guys, Just wanted to give you a heads up. In my wrapping up of SIS-3, I've committed some contributed code from Nga Chung, a student from my CSCI 572 Search Engines and Information Retrieval class at USC this past semester, Summer 2010. I've also removed large portions of the existing LocalLucene port, as the code I just committed does the same thing and it's a bit more understandable to me. It's also probably more limiting, and it doesn't contain the Tier stuff that Patrick originally wrote. My feeling though with that is let's have SIS start small as a computational library for distance, spatial data representation (e.g., WKTs), for storage (spatial data structures like QuadTrees), and for projection. When the time comes to implement the projection part, we'll pull Patrick's stuff back in as needed.
Nga is on the SIS mailing list (so feel free to speak up, Nga!). As part of her final project in my course she developed a fully functional implementation in Java of the QuadTree described in the Spatial Algorithms Hamet book here [1]. It's truly awesome. On top of that, she added a location web service that handles point-radius, and bounding box interrogation of the underlying QuadTree, and then the ability to take the results and throw it into Google maps. Patrick had done some similar stuff with KML and polygons and we might want to bring that in at some point. For now I didn't include the Google maps html page, but I included everything else. I'm debating whether to include the HTML page on a wiki, or as part of the documentation for SIS. Either way, I'll file an issue for it and attach the patch in JIRA. Beyond this implementation, I also refactored SIS into a set of Maven modules, starting with sis-parent (providing the parent POM), sis-core (provides the core distance, geometry, storage and eventually spatial projections), and sis-webapp (the WS layer for SIS). At some point, we'll want to create an sis-app, which probably will contain the SIS command line interface as well. I'm going to add in some issues for things like: (a) a common SIS spatial type (probably with subclasses like point, linestring and multi-linestring), as well as time; (b) spatial reference systems; (c) well known types and conversions; (d) command line interface for SIS. I think we're getting close to a 0.1-incubating SIS release. If we can put some documentation up on a wiki, and then add in the CLI, and maybe some unit tests, I think it'll be good enough to push out for a first version. Feel free to chime in with thoughts. BTW, I've been reading the PostGIS book [2], and it's *really* good and has me thinking about all sorts of awesome places to take SIS. Yay. Cheers, Chris [1] http://s.apache.org/aa9 [2] http://www.manning.com/obe/ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.mattm...@jpl.nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
