[Proj] Re: SRS stuff broken in GDAL and PROJ

[Maciej Sieczka]

Frank Warmerdam pisze:
> Maciej Sieczka wrote:

>> Out of the 2865 common EPSG codes, proj4 strings of 133 of them are
>>  different enough between the epsg files of PROJ 4.5 and 4.6 to
>> result in re-projection errors of few centimeters most of the time,
>> about a meter in several cases to as much as kilometers in few. I
>> would call this a critical issue.

> I am presuming that the ones with large differences are due to the
> fact that PROJ 4.6.0 does *not* apply a change-of-ellipse shift when
> no datum to wgs84 mapping information is available.

Right. After making correction for that, 120 SRSs in the epsg file of
PROJ 4.6 still have errors significant enough to cause an error of few
decimeters, compared to epsg file of PROJ 4.5.

> I would suggest you isolate 2-3 examples out of your list that you
> are confident are significant issues and we can discuss them in more
> detail.

All 120 are significant. Please pick for testing the one you like. I'm
attaching a list of test points I used - 120pts.txt. They are in
ll/WGS84 coordinates. The file is organized as follows:

---
espg_no easting northing
...
---

Respective proj4 strings, derived from the epsg file of PROJ 4.5 and
4.6, followed by cs2cs result obtained with both proj4 strings per each
pair of coordinates is attached as 120errors.txt. The file is organized
as follows:

---
epsg.45:<epsg_no> proj4 string derived from epsg file of PROJ 4.5
epsg.46:<same_epsg_no> relevant proj4 string from epsg file of PROJ 4.6
result.45: re-projection result from ll/WGS84 using the 4.5 proj4 string
result.46: re-projection result from ll/WGS84 using the 4.6 proj4 string

...
---

> Please ensure that you have worked out whether the significant
> difference is in the PROJ.4 string generated for the epsg file, or
> whether it is the behavior of PROJ.4 between 4.5 and 4.6 that you
> think is responsible.

I used exclusively PROJ 4.6 in the test. Only the epsg file alone was
taken from PROJ 4.5 to compare it with PROJ 4.6 one.

> As you likely know Andrey has committed a fix in the last day or two 
> for CPLAtof()

That's excellent. There is no mention of it in the relevant Trac ticket
tough [1].

Cheers
Maciek

[1]http://trac.osgeo.org/gdal/ticket/2036

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 120.zip
Type: application/zip
Size: 6077 bytes
Desc: not available
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20080506/a668a2bc/attachment.zip>