[Gdal-dev] problems with OSR

[Frank Warmerdam]

Gary Bishop wrote:
> Thanks,
> 
> I figured out that the version in CVS would work. I didn't interpret the 
> messages as warnings since they included the string ERROR...
> 
> The demo is a lot better with a game controller with rumble feedback. A 
> good sound card helps too.
> 
> We render the vector data using OpenGL and do our point and area queries 
> using OpenGL selection. For access to the vector data (all we support 
> right now) my grad student found a library to read shape files. I forget 
> its name. We can download shape files from the National Atlas and our 
> colleagues over at the Ancient World Mapping Center can easily ship us 
> there data in shape files.

Gary,

Perhaps that's my shapelib library ... Shapelib.   It is also incorporated
into OGR which provides Shapefiles and other formats.

> I'm struggling to learn about map projects (messy stuff!). The maps from 
> the National Atlas appear to be in NAD83 (which is something like 
> latitude/longitude as I understand it). I want to produce a normal 
> looking map (say of NC) so it looks to me like I can use the functions 
> in OSR to get into another coordinate system that I haven't figured out 
> yet. Then we can work with the map in that coordinate system. I've read 
> about "State Planes". I'm thinking that is likely most appropriate for 
> state maps.

The North Carolina state plane coordinate system can be auto-created
as EPSG:32119.

eg.

import osr
x = osr.SpatialReference()
x.SetFromEPSG( 32119 )
print x.ExportAsPrettyWkt()

NAD83 is a datum (or Geographic coordinate system).  Geographic coordinate
systems used degrees of latitude and longitude.  The NAD83 aspect indicates
some detailed information about how it is tied to the ground.  Difference
datums are often offset from each other by a few meters depending on how
the original surveying was done.

Projected coordinate systems, like the north carolina state plane zone,
are measured in meters (or feet in some cases) instead of lat/long degrees
and generally give a truer reflection of the shape of things in their
locality of appropriateness.

Let me know how you do on the reprojection.  I can help as needed.

> Thanks for the interest in BATS. Right now, kids who are visually 
> impaired are in 45 public school systems in NC. Almost all of them have 
> NO access to maps of any kind. When the sighted kids are doing a map 
> activity the kids who are blind have NOTHING to do. That drives me 
> crazy. We're trying to provide some small step toward fixing that. We're 
> going to give our software away with as many maps as we can stand to 
> make. Then we're going to try to make some tools to make it easy for GIS 
> cognizant teachers and school kids to make maps for kids who are blind. 
> Hopefully some map publishers will pick up the idea and provide textbook 
> maps in an accessible format.

Hmm. It might be challenging to get much commercial interest given the small
size of the market, but it does seem like a good academic / open source
project.  There are lots of good map data sources available and with some
work it should be possible to bundle the software with locale appropriate
map data (as well as world datasets of course) with some local effort.

My son actually goes to a elementary school with a large number of visually
impaired students.  As things progress with BATS perhaps I can bundle it
with Ontario data and offer it to them for a trial.

Do you have an announcements email list or something like that I could join?

Best regards,

-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent