Tom,

I’m sorry about this. The problem is that MGET is not compatible with ArcGIS 10.2.1 yet, and I have not had a chance to release compatible version. I am occupied preparing for a series of meetings that occur throughout February but anticipate releasing a compatible version in March.

Meanwhile, I have had time to partially diagnose the incompatibility. There are two issues that I know of:

1.    For the first time (going back to ArcGIS 9.0), ArcGIS’s internal version number includes a third numerical component, the .1 in 10.2.1. Silly as it seems, this additional component of the version number tripped up MGET’s code that detects which version of ArcGIS is installed. This is the reason why all MGET tools will appear to run successfully in very short time, with no additional messages reported. The version detection code is failing before MGET can even initialize its connection with ArcGIS, so it cannot report any messages.

2.    ESRI upgraded the version of the Python numpy package that is included with ArcGIS to a version that MGET is not yet compatible with.

I have temporary workarounds for both of these issues, if you are interested. For #1, I can send you a patched MGET file that fixes the problem. This is straightforward and low risk.

For #2, you may rename the numpy directory (I can provide detailed instructions) for the numpy that ArcGIS installed, and MGET will fall back on its own copy of numpy that it includes in case the user does not have numpy. This workaround is more complicated and higher risk, in that parts of ArcGIS itself might need the newer version of numpy. My guess is that it will not be problem—that ESRI just updated their copy of numpy to keep up with the latest, which is consistent with their past behavior. If it is a problem, it will manifest as certain Python geoprocessing tools included with ArcGIS itself will fail. If that happens, functionality can be restored by simply renaming the numpy directory back to its original name.

If these workarounds interest you, let me know and I’ll email you the details. I have heard from another user that they seemed to work fine with MGET, and have not heard of any numpy-related problems with ArcGIS’s own tools.

If this is not suitable, I’m sorry but you’ll have to wait until the new version of MGET comes out in March.

Best regards,

Jason

P.S. Some might ask, why didn’t this problem come up with ArcIGS 9.3.1, which includes a third numerical component to the version number? It is because ArcGIS’s internal version number (the one in the Windows registry and elsewhere) for 9.3.1 is actually 9.3 SP2. I guess ESRI Marketing decided to rebrand 9.3 SP2 as 9.3.1 at the last minute, or something like that. In any case, MGET was not tripped up by 9.3.1 because the registry lists it as 9.3 SP2.

Hi,

I am running ArcGIS 10.2 with the latest version of the Python package.  I followed the instructions on the website to the letter.  The Python package and Toolbox installed without error and I added it successfully to ArcMap.  I am trying to populate a point shape file with elevations from an Aster elevation raster formatted as a .bil.  When I run the tool, “Interpolate Raster Values at Points”, the process window appears and it all seems to run without errors.  However, the target field in the points layer formatted as Float remains unchanged.

I tested similar tools, “Find Nearest Features Listed in Field” and “Find Nearest Features” with the same result, no errors but no results either.

I tested the tools “Create X Coordinate Raster” and “Create Y Coordinate Raster” which ran without error but failed to create new rasters.

All of the above tools where used on shapefiles and rasters within the Windows directory structure, not from within a file or personal geodatabase.

Is this a known bug with the new release for ArcGIS ver 10.2.1?

Thanks for building and providing this valuable toolbox.

Tom

______________________________________________________________

Tom Bell Ph.D. Professional Geologist