The ability to use the Gtrans engine and parameter files is tremendously useful for me. I also use Gtrans natively, and then to create special parameter files for tricky operations. However, the FME implementation of Gtrans needs improvement.
Here are some current issues.
1/ Gtrans parameter files (*.tf / *.tfi) that contain mutliple steps fails.
Sometimes, a one step transformation is not possible within Gtrans. The solution is to create a transformation path via an intermediate common projection.
Such tf files fails when used in FME. No error message is given, and FME reports success even though the workbench has failed with no output.
2/ The process indicator fails when Gtrans is used.
I have a custom transformer for a recurring Gtrans operation. When I use that, the progress percentage indicator fails, even though the process runs as it should. This is a problem for lengthy operations with large raster data sets.
3/ The Gtrans transformer should be able to read *.tf / *.tfi files from the native Gtrans archive. Now I have to either point Gtrans to the FME Gtrans files (which are a bit dated) or manually copy new parameter files from Gtrans to FME. A simple search path in the Gtrans transformer settings could remedy this.
Maybe fellow users have found workarounds for the above?
Regards, Mats.E