@Shakyamuni 985110 wrote:
Most likely problem is different cell devices using different Datum’s or representation of the shape of the earth, you have to convert between them when using different data sets.
I ran into this problem myself recently whilst trying to use a freeware GIS app from some German developers to plot the estimated and actual range of a small radio transmitter. To satisfy the requirements of my countries Communications Ministry (Ofcom UK) as well as global good practice of avoiding harmful interference to other users I do need to keep some kinds of reasonably accurate records of where such equipment is being used.
Ofcom and other UK public service organisations (as you might expect) use the British Ordnance Survey grid.
This useful projection splits up the UK into convenient squares of 1km that can be further subdivided into 100m or 10m areas, with the added advantage that distances between two grid references can thus be worked out using basic arithmetic. For a small country like the UK any position error from “flattening the earth” is less of an issue – it does seem to be accounted for in the projection itself although the maths involved is more than I would want to be dealing with, thankfully most handheld GPS units and mobile phone apps will do it accurately.
The OS to be fair do explain all the datums they use and the conversion but it gets more complicated as there now appears to be more than one depending what sort of maps you are getting from them, the required accuracies and distances being measured (which is fair enough now we have access to GPS/GLONASS/Gallileo navigation satellites).
The freeware app uses Openstreetmap overlays and has some setting where you are supposedly able to use UK grid references, which apparently is embedded in the openstreetmap overlay.
Alas, whatever numbers are in there are wrong (and I lack the maths skills to correct them) as using this GIS my waypoints are all shifted about 100m from where they should be – I double checked the grid references with two devices running different apps, a “dead tree format” Ordnance survey map I bought in 2006 which is now completely obsolete due to many buillding developments, and another mapping software I got for tracking cycle routes (which does have licensed OS map data but with some limitations compared to a fully fledged GIS).
Similar position errors (but of a greater magnitude) appear on the cat tracker site (which also uses an odd mashup of Openstreetmap data on a Google map) across Northern Europe although this might be the US prof deliberately using Google Maps (who can afford the extra traffic) to avoid crashing OSM with waypoint data of 1 000 000 cats – and pet cats often roam as far as 1km from their homes increasing the margin of error :laugh_at:
