Pages

Wednesday 25 February 2015

CSS Workshop is Live on GeoServer User Manual

Quick thanks to Travis and Mike for porting my FOSS4G CSS Workshop into the GeoServer User Guide (and to Eva Shon who helped with the initial workshop).

Here is the direct links to the latest user guide:


The GeoServer community would love a hand testing GeoServer 2.7-RC1. It has an all new GeoTools based css extension and we need your feedback!

Thursday 19 February 2015

First meetup of 2015 with Victoria GeoGeeks

Our previous Victoria GeoGeeks event had such a great turnout we decided to try again.
  • Paul treated us to his first ever CartoDB demo (ranging from shapefiles to screen scraping crime data out of the local police department). Good SQL comfort food, and a lot of questions from the crowd.
  • Intermission: an actual paper map with recycle locations
  • Terry Curran shared his work with Marine Data Centre / Pacific Salmon Foundation. A  GeoServer / GeoNetwork / PostGIS stack. Used a tool called jNAP to directly edit the metadata. Check it out at soggy.zoology.ubc.ca.
  • An intro to the BC Geocoder (soon to be an open source project). Real impressive setup with 1.8 million locations an hour, REST API, Batch GeoCoder etc... Some of my old pals from Refractions were on hand to answer the technical questions.
  • I ran a preview of next weeks FOSS4G-NA talk Comparing OSGeo and LocationTech. Comparison are always a real tricky topic, even with a supportive crowd.
Thanks to Greg for helping me arrange the "Sticky Wicket" venue. This time out we did not specify the 10 mins "GeoRabble" format - resulting in much more in-depth presentations. I think I like the shorter format, enough to start a conversation and keep the pub atmosphere going.
OpenSource on the BC Geocoder Roadmap

Friday 6 February 2015

A good test for OGC and OSGeo collaboration

One thing we ran into when going through GeoServer incubation (with OSGeo) was the interesting case of how to handle test data (from the CITE project) and application schemas (from every second standard the OGC produces).

The solution for GeoServer was was to point to the software notice. Here is an example from the cite wfs 1.1 README:
The CITE team engine does not offer a specific license for re-distribution, as such it falls under the general guidance provided by the OGC legal page (http://www.opengeospatial.org/legal/) and the following license:
OSGeo incubation is not especially strict. Since OSGeo and the OGC have an "Memorandum of Understanding" so we figured we could talk through any confusion.

It looks like pycsw has found just such a confusion! The pycsw is a catalog implementation written in Python that is super close to finishing OSGeo incubation.

The project has been rejected from debian distribution ... because the OGC Schemas are not clearly open source. Here is the formal response from November 17th:
Dear Maintainer,
unfortunately I have to reject your package.
One result of the discussion about tinyows was that OGC schemas don't fall  under the Software Notice but the Document Notice. This makes them non free  (no modification) and tinyows had to move to non-free.
I am afraid that pycsw has to do this as well.
  Thorsten
Sounds like a great opportunity to leverage OGC and OSGeo collaboration. I am not quite sure what the answer is (schemas are open to modification, but doing so does render them incompatible with others). I guess we have the same question when distribution the EPSG dataset used for interoperability.

If you are interested in this sort of thing check out http://lists.osgeo.org/listinfo/standards .

If you are an open source developer and would like to avoid these questions consider OSGeo Incubation.

Update

From email Discussion: OGC XML schemas and FOSS4G software distribution
There is now a wiki page to brain storm suggestions for Carl to take to the Technical Committee meeting. Deadline is  March 9th 2015 if you are in position to contribute.

Based on the above conversation it sounds like OGC is going to reach out to FTP Masters and resolve this issue.

A big +1 to OGC - thanks for keeping it real.