Recent

Support Wikipedia

Calendar

Navigation


On a New Road

Having an unreasonable amount of funThursday January 19, 2012
For the past several months I've been having an unreasonable amount of fun. I've been spending almost all of my time coding, with very little time wasted on meetings and corporate rubbish. I've built a publish/subscribe cloud repository that deals with some of the oddities of dealing with streaming telemetry. All the usual replication/redundancy stuff. And I've been building a rich client for doing visualization. I'm totally in love with the WorldWind library from NASA. Totally solid and filled with a pile of features that make building complex apps (almost) dead simple. I did have to do one bit of low level hacking: their line drawing primitives work well for a few hundred lines that don't need level-of-detail support. I needed to handle lines with 10s of thousands of segments with level-of-detail handling. Not hard, just fun. I've done a lot of library work that it probably makes sense to open source sometime.

I have actually been doing a lot of blogging, just not here on my personal blog. You'll find it at the PacX challenge site which I and several other folks contribute to.

Comments:

Glad to hear that you are avoiding corporate rubbish! I don't seem to be able to get to the PacX blog ... your link is funky and the blog link on the PacX webpage also seems to be funky, but in a different way ... it's probably me, but just in case I am not funky ... Lisa.

Posted by Lisa on January 19, 2012 at 09:42 AM PST #

Just for information purpose, since your project seems to involve geographic information... The Open Geospatial Consortium (OGC) - http://www.opengeospatial.org - defines many international standards, some of them being adopted by ISO (International Standards Organization) and INSPIRE (an European commission initiative - if you have clients in Europe, you may be asked to be INSPIRE-compliant by European laws). The OGC standards are mostly oriented toward Web Services. However there is one OGC Standard Working Group (SWG) proposing a set of standard Java interfaces: http://www.geoapi.org. The intend is to achieve for geospatial information what JDBC achieved for databases on implementation neutrality. GeoAPI 3.0 has been adopted as an OGC standard last April and GeoAPI 3.1 is under way. For now, GeoAPI focus mostly on metadata (ISO 19115) and coordinate operations (ISO 19111) services. But work on temporal (ISO 19108), geometries (ISO 19107), coverages (ISO 19123) are under way.

Posted by Martin Desruisseaux on January 19, 2012 at 10:47 AM PST #

(continuing on previous comment) I remember that you already found a map projection library from JHLabs. Note that in my understanding, this project has been transferred to the OSGeo foundation: http://trac.osgeo.org/proj4j. If you were using GeoAPI interfaces, you would have the choice between different implementations (currently the Proj.4 library from which JHLabs were derived, the NetCDF geolocation package or the Geotoolkit.org library – but it should be easy to create JHLabs wrappers), would use an API closer to ISO standards (so no need to learn a new conceptual model if you have to read the XML files mandated by laws), could start with a lightweight implementation and switch more easily to a more complete library when needed (e.g. if you need ISO 19139 XML marshalling, map projections built from the EPSG database, Well Known Text (WKT) parsing and formatting, n-dimensional coordinate systems, projection derivatives, etc.).

Posted by Martin Desruisseaux on January 19, 2012 at 10:52 AM PST #

(continuing on previous comment) Maybe more important, for whatever GeoAPI implementation you would choose, you would automatically inherit the GeoAPI conformance test suite for verifying the quality of the selected implementation. This test suite is based among others on the Geospatial Integrity of Geoscience Software (GIGS) - http://www.epsg.org/gigs.html. For example the GeoAPI conformance tests have revealed that the "Lambert Conic Conformal (2SP Belgium)" (EPSG:9803) projection method in the Proj.4 library has an error of more than 700 metres. I would guess that JHLabs has the same error since it was derived from Proj.4. OGC organizes 4 meetings by year (schedule at http://www.opengeospatial.org/event?category=ogctcpc). I usually attend at some of them. It would be nice to meet you there if you have interest in geospatial applications in Java. Otherwise please feel free to contact me if you would like more information (every softwares I mentioned are open source).

Posted by Martin Desruisseaux on January 19, 2012 at 10:57 AM PST #

Sounds like a fun job James. I notice you are using various gauges in your app, it might be worth you taking a look at Gerrit Grunwald's SteelSeries library, he has some great looking gauges there: http://harmoniccode.blogspot.com

Posted by Steve Webb on January 19, 2012 at 11:53 AM PST #

Post a Comment:
Comments are closed for this entry.