Posting to newsgroup

I posted the following to microsoft.public.mappoint:

MapPoint needs to implement a WGS84 to OSGB/ODN Helmert transformation.

OK this is only relevant for the UK, and people using OSGB grid references.

See what I'm on about at:


Here are the replies...

Each is separated by a green bar.


you should send that comment as wish to [email protected] to ensure
that it is taken care of ...

mfg Klaus Oberdalhoff [email protected]




reading article 42 on mp2kmag shows, how to use Mappoint from C++, maybe
that one helps.

it seems to work, though it's not supported by MS.

mfg Klaus Oberdalhoff [email protected]



Last but not least, i "Googled" a litlle <g>

The spreadsheet includes a full set of Transverse Mercator projection
functions, which allows you to:

Convert latitude and longitude to grid eastings and northings for any
Transverse Mercator map projection, including the Ordnance Survey National

a.. Convert grid eastings and northings to latitude and longitude for any
Transverse Mercator map projection, including the Ordnance Survey National

b.. Compute local scale factor at any point on the grid, and compute grid
convergence, t-T and true azimuth between any pair of points on the grid.

c.. Convert latitude and longitude coordinates to cartesian XYZ
coordinates and vice versa.

d.. Convert latitude and longitude coordinates in any coordinate system to
any other coordinate system. For instance, GPS coordinates may be converted
to Ordnance Survey National Grid coordinates with an accuracy of about 5
metres, anywhere in Great Britain. Please read the guide mentioned above to
learn about the limitations of this type of datum transformation. For GPS to
National Grid conversion requiring better than 5 metre accuracy, you should
not use this simple transformation - use the National Grid Transformation
OSTN97 and National Geoid Model OSGM91

e.. Reformat latitude and longitude coordinates between: Degrees, Minutes
& Seconds; decimal Degrees; and Degrees & Decimal Minute formats.


a cheap conversion program


UK National Grid.
UK Ordnance Survey maps are constructed using the OSGB36 Geodetic Datum, but
IGC requires all Flight Recorders to be set to the WGS84 Datum. A conversion
program to convert between UK Ordnance Survey Grid References and Lat/Longs
(both OSGB36 and WGS84 Datums) can be accessed from .


Here's a repost of an article I put in this group last year.

The British National Grid is the prime geo-reference system in the UK. All
mapping produced by the Ordnance Survey (the UK national mapping agency) is
it, including the more than 200,000 sheets that cover the whole country at
scales of 1:1250 and 1:2500. These are nearly all available in digital form
now (and nearly all captured and maintained using Laser-Scan mapping

It's on a scaled Transverse Mercator projection (Britain as a country is
aligned with its long axis North-South).

Most transformation software that can do Transverse Mercator can therefore
do NG, given the appropriate parameters (see below).

The OSGB pamphlet of 11/94 "The Ellipsoid and the Transverse Mercator
Projection", is the definitive description, and contains algorithms for
transformation to Lat/Long. It's available from:
Control Sales, Geodetic Survey, Room C325, Ordnance Survey,
Romsey Road, Southampton, SO16 4GU, England.
Tel +44 1703 792519
Fax +44 1703 792250

OSGB have recently published a pamphlet "National Grid/ETRF89 Transformation
parameters" with an empirical piecewise transformation to GPS coordinates.

Here is some info on the UK National Grid that I submitted to the
GeoTIFF working group.


Nation: Great Britain (GB) (Northern Ireland which is part of the UK has
own grid).

Date (approximate) of Implementation: 1936

Projection Type(s): Transverse Mercator

Datum Name: OSGB36

Ellipsoid Name: Airy Ellipsoid

Documentation References: (1) JB Harley 1975 "Ordnance Survey Maps -
A Descriptive Manual, p21., HMSO.
(2) OSGB pamphlet of 11/94 "The Ellipsoid and the
Transverse Mercator Projection", OSGB

Projection Params, as appropriate:

Scale factor =0.9996012717
True origin of projection is Lat 49 N Long 2 W.
False origin of coordinate grid is is -400Km E, +100Km N, from true origin,
ie Lat 49 46' N, 7 33' W.

Paul Hardy (PGH), Product Manager (was Chief Programmer), Laser-Scan Ltd,


mfg Klaus Oberdalhoff [email protected]


I emailed Microsoft. Microsoft MapPoint Developer Wish List


A couple of bugs...

MapPoint needs to implement a WGS84 to OSGB/ODN Helmert transformation.

OK this is only relevant for the UK, and people using OSGB grid references.

Parsing error in TL square

The import data wizard doesn't like Cambridge, but is happy with Milton Keynes.

This CSV is parsed correctly:


but this isn't:

10:40:47,TL4705060256. It ignores the TL.

Reducing to 10 metre accuracy and the wizard's happy:


See what I'm on about at:


I got a reply!

Thank you for your email.  I will share it with the development team.

That's better than a poke in the eye. It is much valued. Well worth pasting it here.

I guess it's an automated thing. The I is probably a old 286 running some mail reader thing on Linux. We Brits get a bit cynical, especially when we are 120 metres from where we think we are.


Then I got another reply:

----- Original Message -----
From: Microsoft MapPoint Developer Wish List
To: [email protected]
Sent: Tuesday, November 13, 2001 5:50 PM
Subject: FW: Problems with mappoint 2002 (OS Grid problem)

I wanted to follow up on your email.  ADO is not  Microsoft , but we use  do use it.  ADO  thinks that the column with OS grid references is a number instead of a string. Using XLS (instead of CSV) or putting the column in quotes (quotes around each string the in the column) will make it work.
Oh, and  just so you know the OSGrids in MapPoint were written by  British developers, so the comments on your website are misrepresenting. 
Hope this is helpful.
MSDN MapPoint

To which I replied:

Dear MSDN MapPoint
Thanks for your reply.
In your reply you mentioned "ADO". I'm not sure what connection ADO has with MapPoint. (ActiveX Data Object [ADO] is a Microsoft acronym!)
Would you clarify this?
Your reply doesn't address the Helmert transformation error within MapPoint. This leads to a significant error of 120 metres. Is this going to be fixed?
>>...the OSGrids in MapPoint were written by  British developers, so the comments on your website are misrepresenting.
The comments on my web site represent my distorted view of the world. (Distorted and displaced by 120 metres!)
No offence is intended!

I ended up submitting an article, Accuracy With OSGB Grid References, to MP2K.


Someone else noticed!