Hi James

There are some useful web sites for converting the odd reference (e.g. http://www.obliquity.com/astro/osgb.html, http://www.obliquity.com/astro/osgbapplet.html or http://gps.ordnancesurvey.co.uk/convert.asp - I haven't tried the batch conversion tool).

But I guess what you want is to do it directly from the web service . I was planning to do something similar (but haven't had the time) and I would use the constants and calculations in http://www.ordnancesurvey.co.uk/oswebsi … ations.xls to be able to write code to do the transformations. The important bit to note is that the spreadsheet clearly states that "the code can be copied..." so as far as I understand it, there should not be copyright issues but you may still want to check with OS :)

Hope that helps

Cheers

Nick

Hmm.

Have you got decent libraires for conversion from WGS84 to OS (and vice versa) in JS - or are you using an approximation ?

My understanding is that as they have different datums - an accurate conversion requires a table of fudge factors - or some very complex maths (even the OS GridInQuest DLL uses a table of fudge factors for different Lattitudes I believe)

I'd certainly be interested in doing it client side - as on occasion I have a server trip simply to do a coord conversion

Thanks
James

Many thanks for that I'd not seen a js implementation - that could be useful.

I agree entirely about the accuracy for species records - especially those with legs when they are recorded in a 10km square !

However some of the things I've also been playing with are overlaying polygon boundaries etc - and if the viewport isn't quite accurate with its corner coordinates everything can get offset (I've been playing putting species points onto google maps along with site boundaries)

Having said that - it might well be worth using both the methods so at least all the species data is translated client side

Thanks
James

Hi,

Just to note that that particular javascript library doesn't take into account the datum transformation between WGS84 and OSGB36:

"Note that GPS is based on WGS84/GRS80, which can vary from OSGB36 by as much as 120m or 6" or arc (OSGB36/Airy is a better fit to the UK geiod than the geocentric WGS84 which covers the entire world). The latitude/longitude given here is OSGB36; I may look at converting between OSGB-36 & WGS-84 sometime in the future."

It's actually not that complicated to transform between them using the Helmert transformation given in the OS GPS paper. The projection stuff is more complicated imho.

andyb