OpenGuides 0.45 has now been released, and now supports geolocation worldwide. It is available from your favourite CPAN mirror, or
http://www.larted.org.uk/~dom/computing/code/openguides/OpenGuides-0.45.tar....
md5sum: 218b27cc4df3c6de3e8295adc9c0ae26 OpenGuides-0.45.tar.gz
Changelog:
Made the geolocation stuff work worldwide. Squeeeeeee! You can now choose between doing your distance calculations with the British National Grid, the Irish National Grid, or a UTM ellipsoid. If you wish to use anything other than the British National Grid and you have pre-existing location data then you will need to save an edit of each node with location data before distance searches will work. In less exciting news: Fixed bug relating to lat/long representation. Removed debugging warn accidentally left in last release. Fixed some HTML validation errors.
Debian packages are also available from the usual place.
Dominic.
This one time, at band camp, Dominic Hargreaves wrote:
Debian packages are also available from the usual place.
Yay! Lat/Long works!
Can you explain this setting? What are the options? Do users entering coordinates need to know? I'm a little wooly on the detail of this geo stuff.
# Which ellipsoid do you want to use? (eg 'Airy', 'WGS-84') # This is only used if you select geo_handler = 3 above ellipsoid = International
On Wed 01 Dec 2004, Rev Simon Rumble simon@rumble.net wrote:
Can you explain this setting? What are the options?
You can use any ellipsoid accepted by Geo::Coordinates::UTM.
Do users entering coordinates need to know?
No.
Kake
This one time, at band camp, Kake L Pugh wrote:
Do users entering coordinates need to know?
No.
I'm not up on all this geo stuff, so stop me if I'm being stupid. Though pointers to places where I can get this knowledge would be appreciated. 3D spatial reasoning is always the bit I don't do well on in IQ tests, so this stuff doesn't come easily for me.
Doesn't the ellipsoid used have an effect on the Lat/Long? Doesn't this mean we need to specify which one is expected? In other words, does a single point on the globe have a different Lat/Long depending on the ellipsoid in use?
From my brief reading of the README for Geo::Coordinates::UTM, this seems to be the purpose of the module.
Je 2004-12-01 12:36:48 +0000, Rev Simon Rumble skribis:
Doesn't the ellipsoid used have an effect on the Lat/Long? Doesn't this
http://www.gps.gov.uk/additionalInfo/coordinateSystems.asp
P
On Wed 01 Dec 2004, Rev Simon Rumble simon@rumble.net wrote:
3D spatial reasoning is always the bit I don't do well on in IQ tests, so this stuff doesn't come easily for me.
I found this page very helpful once I'd read it enough times and lain down on the sofa with my eyes closed once or twice: http://www.vterrain.org/Projections/
Doesn't the ellipsoid used have an effect on the Lat/Long?
There are two issues with ellipsoids, and this is the reason why back on the list in June or whenever it was zool and Ivor were disagreeing about whether WGS-84 is global or not. It is, and it isn't.
The issue you are thinking of is related to the earth not being quite a sphere. So yes, you're right that latitude and longitude vary depending on which approximation to the shape of the earth to work them out. However, the earth is near enough to a sphere for the purposes of locating a restaurant - see http://openguides.org/mail/openguides-dev/2004-June/000462.html for some calculations that Dave did.
The other issue is the one of once you have latitude and longitude, turning them into square coordinates, and this is where we really do care about the ellipsoid.
First consider a Mercator projection - not a transverse one. Imagine the earth is made of a thin rubber shell, and the lines of longitude are wires embedded in that shell. So the wires curve and meet in two points at the poles. Now detach the wires from each other at the poles, and straighten them out. The rubber will stretch to allow this.
Near the equator, there will be very little stretching. Near the poles there will be a lot. So near the equator, distances will be preserved but distances will change more the further you get from the equator.
Now consider a transverse projection; keep hold of our straightened-out globe and turn it on its side. What *was* the equator is now some line of longitude, and it's near that line that distances are preserved. So if the countries we mainly care about are round about there, then hurrah, it's all going to work nicely.
If they're not, then rotate our wire thing around a north->south axis (change the ellipsoid) until they are.
Does that help explain it?
Now as for which ellipsoid you should choose being as your guide is worldwide - your guess is as good as mine. However it is possibly not as vital for your distance calculations to be incredibly accurate. Depends what you want them for really.
Kake
On Wed, Dec 01, 2004 at 12:55:51PM +0000, Kake L Pugh wrote:
Now as for which ellipsoid you should choose being as your guide is worldwide - your guess is as good as mine.
If you check with the local Ordnance Survey types, they'll tell you what reference ellipsoid is in vogue for their area. e.g. The Irish Grid uses a Modified Airy Ellipsoid http://www.osi.ie/gps/overview/irishgrid.asp
Stephen
On Wed, Dec 01, 2004 at 01:15:30PM +0000, Stephen Stewart wrote:
On Wed, Dec 01, 2004 at 12:55:51PM +0000, Kake L Pugh wrote:
Now as for which ellipsoid you should choose being as your guide is worldwide - your guess is as good as mine.
If you check with the local Ordnance Survey types, they'll tell you what reference ellipsoid is in vogue for their area. e.g. The Irish Grid uses a Modified Airy Ellipsoid http://www.osi.ie/gps/overview/irishgrid.asp
*sigh*
Learn to read, Stephen.
Stephen
Excellent explanations all-round folks. Thanks very much! I'll whack something in the openguides.org FAQ, for others with the same confusion as me.
Right, I've done that. Please have a read and improve it if you think I've got it horribly wrong:
http://openguides.org/dev/?node=FAQ
openguides-dev@lists.openguides.org