[Proj] Finnish KKJ in proj 4.4.9 - Linux vs. Windows

Roger Oberholtzer roger at opq.se
Thu Dec 15 02:46:31 EST 2005


On Wed, 2005-12-14 at 10:11 -0500, Frank Warmerdam wrote:
> On 12/14/05, Roger Oberholtzer <roger at opq.se> wrote:
> > I have sorted out using proj 4.4.9 to obtain KKJ X/Y values from WGS84
> > LAT/LONG values.
> >
> > To be more precise, I have this sorted out on Linux. I compiled the same
> > 4.4.9 code on Windows using MS VC++ 6.0 and the makefile.vc in the proj
> > source. The problem is that I do not get the same results on the two
> > systems! The error is in the neighborhood of 100 meters.
> > Has anyone used proj on Windows and had any such problem? I have used
> > other similar libraries on both platforms before, so I did not expect
> > this.
> 
> Roger,
> 
> I haven't run into this.  If you can demonstrate the results you get on
> linux and win32 using the "proj" or "cs2cs" commands, then I could
> try and see if I get the same problem here on win32.

I will get something out later. I want to go through all the steps again
to be sure I have not done something dumb.

> > I was thinking of trying the 4.4.6 binary that is available for
> > downloads, but it does not contain a proj.lib file, making linking
> > problematic. Any reason there is not a windows binary of 4.4.9, and only
> > of 4.4.6?
> 
> Producing binaries is a hassle, and so I do not make it part of
> my release regimen.  The FWTools releases do include up to date
> PROJ.4 binaries though I'm not sure if I include the PROJ.4 include
> files and libraries for those wanting to build their own applications.

No problem. I just wanted to be sure that there were no specific issues
with 4.4.9 on Windows. As it is a calculation library, I would have been
surprised if there had been.

In the windows bin archive, a few (not all) of the files in the nad
directory have DOS EOL sequences, while the source archive contains
files with non-DOS EOL sequences. So:

1. Why don't all the text files in nad have the same EOL sequences on
each platform.

2. Is the code really written so it makes a difference? I think it may
be, as when I changed the EOL sequences in the esri file (which I use)
the results changed. This seems unnecessary, as a text file can be read
correctly no matter if a small bit of care is taken. Is a patch needed
for this? Should I make one?

Another Windows question: if the program is compiled to run on Windows,
perhaps it could also look in the registry for the location of the
various runtime files. If this is acceptable, I can supply a patch.
Current behavior would not be altered.

It is very simple to make a nice install for the windows package via a
free program called Inno Setup Compiler. It only requires a script to
describe the package. I can provide details if it is interesting.

Gee. I sound like a Windows user. Nope. Pure Linux. But, out clients
keep us working in the strange World of Windows for libraries and
things.

>   http://fwtools.maptools.org/
> 
> You might want to see if you get the same results with the FWTools
> PROJ commands as you get from the ones you built yourself.

I will add this to my list of tasks.

-- 
Roger Oberholtzer
OPQ Systems AB




More information about the Proj mailing list