November 8th, 2008
I’ve finished the base of the topography game. I’m now able to load an image and pan and zoom it. Also it calculates a clicked location from screen to image coordinates. This is not as easy as it sounds, mostly because I try to avoid floats due to the lack of a floating point processor. I wonder how they implement 3D games on the DS without floats and doubles.

If I get the WiFi working it should be relatively easy now to create a simple WMS client supporting panning, zooming and the getfeatureinfo request. Totally useless, but fun to do.
Although, really useless? We are involved in a project, EduGIS, to support geography lessons using e.g. WMS. Most students (or their younger brothers) will have a Nintendo DS.
Well, unfortunately it’s just a home project to play around with the DS, but I could imagine we could do more serious things. On the other hand, we will probably use the iPhone for that. I hope schaaltreinen will blog about that soon
I still don’t understand the policy of Nintendo towards ‘homebrew’ applications. iPhone development licences are a lot more clear.
Posted in Nintendo DS | 1 Comment »
November 5th, 2008
I started to port Geodan S&R’s succesful Topography Game to the Nintendo DS. Installation of the SDK, compiling samples, etc, was fairly easy. A lot more difficult is actually developing something useful. I have lots of years of C++ experience, but the architecture and libraries are quite different to what I’m used. Developing for Windows with Visual Studio, Boost, STL, etc looks incomparable to developing for ARM7 and ARM9 CPU’s. It’s more like embedded programming.
To keep track of what I’ve used:
The SDK, or toolchains as it’s called, is devkitPro: http://www.devkitpro.org
Some documentation about developing for the Nintendo DS:
http://osdl.sourceforge.net/main/documentation/misc/nintendo-DS/homebrew-guide/HomebrewForDS.html#andnow
http://www.dev-scene.com/NDS/Tutorials
http://www.drunkencoders.com/index.php?system_id=1&page=Tutorials
http://www.double.co.nz/nintendo_ds/
The Emulator I use is DeSmuME. Several emulators can be found at http://www.ndsemulator.com/
Just as an example, someone ported Doom to the NDS, cool :)

Posted in Nintendo DS | No Comments »
November 3rd, 2008
Like all families with children we own a Nintendo DS. Like the Lego Mindstorms it is of course not for me but for my sons
By accident I across a SDK, devkitPro . It allows for the development of Nintendo DS applications and also supports other platforms, like the PSP. It is a nice set of tools with a lot of examples.
I don’t know what to do with it now, but it could be interesting for example as a cheap device which communicates with the Touchtable. To stay withg games: a central application could run on the Touchtable (for e.g. Diplomacy, Catan, etc) while on the Nintendo DS every individual player can submit his orders, see his cards, etc.

So now I have a new project (#12.561): to create a Nintendo DS application. First I will try to make one of the most succesful applications of Geodan S&R: the Topography Game, available for many platforms and very suitable for the DS.
Tags: DS, Nintendo
Posted in Tabletop | No Comments »
September 12th, 2008
I’ve recently purchased a Lego Mindstorms NXT. Of course for my sons, but as they are a bit too young for it now, I will have to use it.
It’s also for my work actually. The Lego Mindstorms is an affordable tool for different subjects we are researching, such as:
- OGC Sensor Enabled Web
- Location based services
- Command & control
- Automonous behaviour
OGC Sensor Enabled Web
The OGC Sensor Web Enablement is quite obvious: The Lego Mindstorms supports different sensors. Standard e.g. a light sensor, measuring light intenstities, and a sound sensor measuring Db and Dba. You can also buy additional sensors like temperature, RFID and gyrometer sensors. The sensors can be used for the different components of the OGC SWE. The first components which will probably be used are:
- Observation and measurement (O&M) to encode the sensor data
- Sensor Model Language (SensorML) to describe the sensor systems
- Sensor Observation Service to request, filter and retrieve sensor data.
Later the planning and alert services could be added.
Location based services
Also location based services are quite obvious. The built robot has a location and thus could respond to it (e.g navigate, react on POI’s, etc), or other applications could react to the location of the robot. Of course we first want to show the location of the robot on the Touchtable.
Control & command
On above mentioned touchtable the user could give commands to the robot. Eg to move to another location, to send measurements etc. The robot could act as a mobile field user who receives commands from a control center. Because a PC, like a UMPC can be attached to it, it could mimic the behaviour of field user ,like measurements teams during a disaster, in a training situation.
Autonomous behaviour
Contrary to the C&C mentioned above, you can give the robot an order (e.g. to move somewhere and drop something). Also the robot has the abilities to respond to situations itself. These kind of autonomous behaviour is for example examined in disaster management but normally requires an expensive system.
Conclusions
The Lego Mindstorms is an very affordable tool to investigate different geo related subjects in a extremly easy (and cool) way. It has the potential due to its combination of sensors and intelligence to investigate even further and implement trully geo-innovative ideas for only minimal costs.
Posted in GIS, Mobile, Open Standards, Tabletop | 1 Comment »
June 29th, 2008
As project #546 I am writing a routeplanner based on Openstreetmap, targeted at pedestrians. It will consist of three main components: a route engine, a graph builder and most interesting for me a special pedestrian routing client using not only the OSM roads but also POIs, parks, etc. The first step is now (almost) finished: An implementation of Dijkstra’s algorithm in C#. It can handle a large number of data but I need to test it now with some real data instead of auto-generated locations and edges. Of course these data will be extracted from Openstreetmap data using the graphbuilder.
Posted in GIS | No Comments »
May 23rd, 2008
The WFSHarvester is now an Open Source project on SourceForge.
Project website: http://wfsharvester.sourceforge.net
SourceForge site: http://sourceforge.net/projects/wfsharvester
There is not a lot to find though. At this moment I am adding comments to the source files, cleaning up some stuff, writing a few lines of documentation. When this is done I will put it in the SourceForge.
After that I will create some sample files and a simple client for testing.
Posted in GDI, GIS, Open Source, Open Standards, WFSHarvester | No Comments »
May 22nd, 2008
Because the WFS Harvester is so simple and might be useful for other organisations as well I’ve decided to create an Open Source project. Apart from making the harvester available to other people, I also hope someone might improve it
Posted in GDI, GIS, Open Source, Open Standards | No Comments »
May 22nd, 2008
For a project it is necessary to harvest data from several distributed WFS’s and put the data in a central WFS. A simple and configurable WFS Harvester is developed to perform this task. Because a few easy to configure configuration files are used it can be used relatively easy for different WFS’s.
The WFS Harvester requires 4 parameters:
1. XML file
2. URL of WFS
3. URL of WFS-T (Transactional WFS)
4. XSL file
The XML file contains a GetFeature request. This request is sent to a WFS. The XML result of this request is transformed into a Transaction request using the XSL file. This transaction request is then sent to the WFS-T.
Of course it could be improved by requesting for the capabilities of a WFS and so on, but now it’s an easy simple library. The XSL file allows for example to copy WFS data from one datamodel into different datamodel and store it in another WFS.
Posted in GDI, GIS, Open Source, Open Standards | 1 Comment »
May 14th, 2008
Months later than hoped TomTom gets the approval of the European Commissioner for Competition, Neelie Smit Kroes, to take over Tele Atlas. Neelie Smit Kroes ordered an additional research to investigate the consequences for the map/navigation market. The investigation established that enough competition remains.
The main concern for the EC was that TomTom/TeleAtlas might raise the price or restrict access to digital maps for other PND (personal navigation device) manufacturers. The EC finds that this is unlikely as there is a competitor, Navteq.
Posted in GIS | 1 Comment »
April 15th, 2008
Yesterday, 14th April, KML 2.2. became an official OpenGeoSpatial (OGC) standard. KML is of course well known as it is Google format for data to be visualised in Google Maps and Google Earth. A huge amount of KML files is currently available. Searching in Google Search with the term ” filetype:kml” results in 3.670.00 hits, “filetype:kmz” (compressed kml) 1.870.000 hits. So in total more than 5 million KML files.
The KML XML Schema can be found here: http://schemas.opengis.net/kml/2.2.0/ogckml22.xsd. The geometry types are now derived from GML 2.1.2.
It offers two ways to add you own data or datamodel: Either by creating a KML Application Profile or by using the ExtendedData element. I will write some post about it soon. Especially the last one can prove very useful to visualise data as well as use it for further processing. Also I think I’ll write something about my opinion of the KML standard in the OGC set of standards.
Posted in Uncategorized | 1 Comment »