<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html lang="en">
<head><meta http-equiv="Content-Type" content="text/html;">
<title>Scalix message content</title>
<style type="text/css">
<!--
.Normal { text-align: left; text-indent:0pt; margin-top:0pt; margin-bottom:0pt; font-family: Arial; font-size: 12pt; }
.Default_Paragraph_Font {font-family: Arial; color: Black; background-color: white ; font-size: 12pt; font-weight: normal; font-style: normal; font-variant: normal; vertical-align: normal; text-decoration: none;}
div.ltTOCtitle { font-family: Verdana; font-size: 8pt; font-weight: bold; text-align: center;}
div.ltTOCl1 { font-family: Verdana; font-size: 8pt; font-weight: bold; margin-left:32pt; text-indent:-32pt;}
div.ltTOCl2 { font-family: Verdana; font-size: 8pt; font-weight: bold; margin-left:40pt; text-indent:-32pt;}
div.ltTOCl3 { font-family: Verdana; font-size: 8pt; font-weight: bold; margin-left:48pt; text-indent:-32pt;}
div.ltTOCl4 { font-family: Verdana; font-size: 8pt; font-weight: bold; margin-left:56pt; text-indent:-32pt;}
div.ltTOCl5 { font-family: Verdana; font-size: 8pt; font-weight: bold; margin-left:64pt; text-indent:-32pt;}
div.ltTOCl6 { font-family: Verdana; font-size: 8pt; font-weight: bold; margin-left:72pt; text-indent:-32pt;}
-->
</style>
</head>
<body>
<div class="Normal"><span style='font-size: 10pt; '>Hi Paul</span></div>
<br>
<div class="Normal"><span style='font-size: 10pt; '>just want to let you know
that it works now. I had to change 4 lines in the pj_transform.c file (some
wgs84 hardcoded stuff for shift grids) and then changed the definitions in the
epsg file as suggested. And the best thing is, that it works with MapServer too
(vector and raster data). I'm now able to do an on-the-fly equalization
:)</span></div>
<br>
<div class="Normal"><span style='font-size: 10pt; '>Thanks</span><br><span style='font-size: 10pt; '>Stefan</span></div>
<br>
<div class="Normal"><span style='font-size: 10pt; '>> -----Ursprüngliche
Nachricht-----</span><br><span style='font-size: 10pt; '>> Von: Paul Kelly
[mailto:paul-grass@stjohnspoint.co.uk]</span><br><span style='font-size: 10pt; '>>
Gesendet am: Freitag, 3. April 2009
14:58</span><br><span style='font-size: 10pt; '>> An: Ziegler
Stefan</span><br><span style='font-size: 10pt; '>> Cc:
proj</span><br><span style='font-size: 10pt; '>> Betreff: Re: AW: Re: AW: Re:
[Proj] Problems using grid shift
files</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>>
Hello,</span><br><span style='font-size: 10pt; '>> I think it would help if
you give an example of one of the
</span><br><span style='font-size: 10pt; '>> transformations
</span><br><span style='font-size: 10pt; '>> that gave the incorrect results,
with the exact cs2cs </span><br><span style='font-size: 10pt; '>>
command-line you </span><br><span style='font-size: 10pt; '>> used and test
points.</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>> On Thu, 2 Apr 2009, Ziegler
Stefan wrote:</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>> > Mmmmh, I think this is
beyond my mental skills ;-) So I </span><br><span style='font-size: 10pt; '>>
start with some</span><br><span style='font-size: 10pt; '>> > questions:
what is a pivot datum? And how can I chain
datum</span><br><span style='font-size: 10pt; '>> > transformation
(without piping in the shell)? How does proj does
a</span><br><span style='font-size: 10pt; '>> > re-projection? I assumed
something like: projection A -> ellipsoid
A</span><br><span style='font-size: 10pt; '>> > ----> here comes the
datum transformation ----> ellipsoid B
-></span><br><span style='font-size: 10pt; '>> > projection B. So if I
know the accurate relation between
</span><br><span style='font-size: 10pt; '>> ellipsoid A
and</span><br><span style='font-size: 10pt; '>> > B (like the +towgs84
params) I'm able to reproject more or
</span><br><span style='font-size: 10pt; '>> less
exact?</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>> No, there is generally an extra
stage:</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>> Source Datum --> datum
transformation A --> Pivot datum -->
datum</span><br><span style='font-size: 10pt; '>> transformation B -->
Target Datum</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>> Only when either your source or
target datum is the same as the pivot
</span><br><span style='font-size: 10pt; '>> datum is there only one datum
transformation. For all other </span><br><span style='font-size: 10pt; '>>
cases you need </span><br><span style='font-size: 10pt; '>> two sets of datum
transformation parameters, for the source
</span><br><span style='font-size: 10pt; '>> and target
</span><br><span style='font-size: 10pt; '>> datum to some intermediate pivot
datum. Parameters for use of </span><br><span style='font-size: 10pt; '>>
WGS84 as a </span><br><span style='font-size: 10pt; '>> pivot datum are well
defined for most datums. But in your
</span><br><span style='font-size: 10pt; '>> case when you
</span><br><span style='font-size: 10pt; '>> want to use CH1903+ as a pivot
datum instead you need to </span><br><span style='font-size: 10pt; '>>
calculate a new </span><br><span style='font-size: 10pt; '>> set of datum
transformation parameters.</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>> As of yet I'm not sure exactly
how to calculate the new </span><br><span style='font-size: 10pt; '>>
parameters but if </span><br><span style='font-size: 10pt; '>> you post some
details of the conversion you're trying to do with some
</span><br><span style='font-size: 10pt; '>> sample points, maybe someone
will be able to work it out for
you.</span><br><span style='font-size: 10pt; '>>
</span><br><span style='font-size: 10pt; '>>
Paul</span><br><span style='font-size: 10pt; '>> </span></div>
</body>
</html>