exporting them to python gives huge overhead, we had some laughs with that.
Reducing page-sized script to sometimes just a couple lines in python...
And you would still have to invoke their functions/factories. (or is that not an issue?)
Maybe parsing the python script to generate the fme script, then some manual post-labor?
Would require some work and experience
Manual conversion is the best, but also requires some work and experience.
Anyway at E sri they say their system has advantages..i see none though. Well, less memory intesive maybe. but then again creating lots of files to clean up after.
And of course model builder has improved quite a lot since i gave up on it...;)
Seldom use it though, sometimes to control mxd's.
Hi @annette2
Did you ever progress this further, "Can FME Server start an ArcGIS Server geoprocessing service?"
Thinking about this also and wondering could I call a geoprocessing service from either desktop or server...
Thanks
Ciara
Hi @annette2
Did you ever progress this further, "Can FME Server start an ArcGIS Server geoprocessing service?"
Thinking about this also and wondering could I call a geoprocessing service from either desktop or server...
Thanks
Ciara
Just tested it out and its quite easily achieved with HTTPCaller, I was able to set my parameters for the tool in the Query String Parameters.
Will be quite useful as allows us to automate the testing of the GeoProcessing Service as we run same tests across 3 environments, as well as load testing and then scheduling from FME Server
exporting them to python gives huge overhead, we had some laughs with that.
Reducing page-sized script to sometimes just a couple lines in python...
And you would still have to invoke their functions/factories. (or is that not an issue?)
Maybe parsing the python script to generate the fme script, then some manual post-labor?
Would require some work and experience
Manual conversion is the best, but also requires some work and experience.
Anyway at E sri they say their system has advantages..i see none though. Well, less memory intesive maybe. but then again creating lots of files to clean up after.
And of course model builder has improved quite a lot since i gave up on it...;)
Seldom use it though, sometimes to control mxd's.
Here's one advantage I can think of: The ArcGIS toolboxes work with the tolerance and resolution set in the feature class/dataset, where as FME works in full precision. This can sometimes give some interesting differences when e.g. comparing geometries where rounding differences can throw off the results.
I hope/suspect this will be much improved with the new tolerance settings in FME 2018, however.
exporting them to python gives huge overhead, we had some laughs with that.
Reducing page-sized script to sometimes just a couple lines in python...
And you would still have to invoke their functions/factories. (or is that not an issue?)
Maybe parsing the python script to generate the fme script, then some manual post-labor?
Would require some work and experience
Manual conversion is the best, but also requires some work and experience.
Anyway at E sri they say their system has advantages..i see none though. Well, less memory intesive maybe. but then again creating lots of files to clean up after.
And of course model builder has improved quite a lot since i gave up on it...;)
Seldom use it though, sometimes to control mxd's.
Not sure if that is an advantage to Esri's model builder.
Just tested it out and its quite easily achieved with HTTPCaller, I was able to set my parameters for the tool in the Query String Parameters.
Will be quite useful as allows us to automate the testing of the GeoProcessing Service as we run same tests across 3 environments, as well as load testing and then scheduling from FME Server
Hi Ciara, Good to hear from you and congratulations on your new job.
We have about 15-20 python scheduled batch jobs. Most we will port to FME.
The http will be an option for the jobs we cannot port to FME :)
