Best Answers

I’m trying to run some workspace through Task scheduler with FME Form. I’m using an Emailer package and when I run the task as my user, everything is fine. But When I run it as local system it always fails as it can’t find the package. I tried defining Shared folder and setting up environment variables, but couldn’t make it work.As for setting up shared connections it works fine, but not with packages.What would be the right setup for this?

Before I file a bug report, I thought I would post here.  I used to be able to right-click in the Data Preview and get the ‘Zoom to Selected Feature’ for the CAD geometry.  The last few versions of FME have not had that available.  What gives?Please see the attached.Currently using the latest build:  FME(R) 2025.2.2.0 (20260106 - Build 25827 - WIN64)

I’m trying to convert those attribute values whichever has comma into an array.Attribute A has values - "ABC" and "ABC","DEF" and "ABC","DEF","GHI".Desired output -“ABC”["ABC", "DEF"]["ABC", "DEF",”GHI”]Any inputs are appreciated. Thanks very much!Tagging ​@danilo_fme , ​@david_r 

We setup an OAuth client in GCP that we’re trying to get working with FME Flow. I have installed the safe.google-bigquery package on FME Form and configured the web service with our unique client information from GCP. See below I then created a web connection based on this web service and uploaded this and a workspace utilizing this connection to our FME Flow instance. The workspace succeeds initially, presumably until the access token expires, but is failing with the following error after an hour or so:  It works again if I manually re-authorize the connection so I believe Flow is failing to retrieve a refresh token from GCP: Any help would be appreciated. Thanks!

FME 2025.1.3.0 Build 25643I downloaded the new ArcGIS Feature Service Package and set up a connection using the Esri ArcGIS Online OAuth Web service. When I attempt to insert a new feature into an existing AGOL feature service using the default values of the Esri ArcGIS Feature Service format, I get a warningWARN  |ArcGISFeatureServiceWriter: The field 'GlobalID' has a NotNull constraint, but it is NULL, missing, or empty. Skipping feature...The feature is not committed but the FeatureWriter or Writer (I tested both) shows success and the number of features written as 1. Without reading through the warnings, I wouldn’t know there was a problem.Since I’m inserting a feature I don’t have a globalid! And the field is noneditable in the feature service, so it seems like an odd error.Messing around with the settings I found that if I made sure the GlobalID field was exposed in the User Attributes and in the Advanced Parameters “Use GlobalID” was set to Yes, then the feature would insert. Bu

In the answer to this questionEsri ArcGIS Feature Service - Reads Dates as Epoch Milliseconds | Communitya URL is given to the FME hub which doesn’t match the address I’m used to seeing for FME Hub, crucially the info on this package is all quite different too This is what i’m familiar withhub.safe.com/publishers/safe/packages/esri-agolEsri ArcGIS Connector - FME Hub and this is the link providedhub.fme.co/publishers/safe/packages/esri-agolEsri ArcGIS Connector - FME Hub

Greetings everyone,I have a large number of 2D polygons for which I want to determine the Z value. I also have a DTM (digital terrain model).Now I have observed the following: If I take only one polygon and, using the FeatureReader, read in only exactly those triangles of the DTM that are touched by the polygon in 2D, and then determine the Z values of the polygon using the SurfaceDraper (always carrying forward only z_max and using that as the Z value via the 3D_Forcer), I get different Z values than when I first create a buffer around the polygon and then pass the buffer to the FeatureReader and include the triangles around the polygon in the SurfaceDraper.The triangles from the DTM do not touch each other.There are deviations of several meters, depending on how strongly the terrain height varies around the polygon.It seems as if the SurfaceDraper somehow takes the surrounding terrain into account, even if it does not touch the polygon.It cannot be due to the tolerance; I have alread

We have multiple workspaces that write to ArcGIS Online with multipart geometries. The old ARCGISONLINEFEATURES writer works and correctly writes the geometries. The new (2025) Esri ArcGIS Feature Service Writer returns an error rejecting the features and the following warning:ArcGISFeatureServiceWriter: INSERT error for a feature in Project Stats by State. The error code from the server was '1000' and the message was 'Adding/updating geometries that result in SQL GeometryCollections is currently not supported (SQL Server performance).' Using a Deaggregator seems to fix this in most cases. However, we want to retain the multipart geometries. In other cases, I see a similar warning and the following error after using a Deaggregator:ArcGISFeatureServiceWriter: INSERT error for a feature in Terracon Offices Closest Areas. The error code from the server was '1000' and the message was 'Data is Null. This method or property cannot be called on Null values.' I’ve tried a NullAttributeMapper a

Hello,I am testing out the Esri ArcGIS Feature Service writer that is in the Esri ArcGIS Online Connector package (Esri ArcGIS Connector v3.13.2). I’m using Form 2024.2.2.0 (20250114 - Build 24820 - WIN64). Screenshots are edited to remove confidential information. I have noticed that the “Target WKID” cannot be left blank even if you have selected “Feature Service Handling: Use Existing” If I am understanding the documentation correctly, this parameter is only needed if the writer is creating a Feature Service. However, as the default is set to WKID 102100, if you don’t change this to the correct WKID, it will write the data out to the wrong location. I initially made this mistake before I realized I had to input my data’s WKID. For the old Esri ArcGIS Portal Feature Service Writer, you were able to leave this parameter blank: I think this parameter being a required parameter is unnecessary when Feature Service Handling is set to “Use Existing” and I like the way it was in the older w

Hi everyone I have been having some issues with the new ArcGIS feature service writer. When using this writer in combination with a change detector, it doesn't seem to update or insert any of my changed records and i have been getting weird behaviour all over the place. Let me preface this by saying that everything works as expected on the old portal feature service writer, it's when switching to the new feature service writer that weird things start to happen.  Under the advanced parameters I have enabled the following fields to adress the first problem I encountered (namely that the writer no longer seems to care that global id is a read only field and skips it for writing):  Afterwards, I ran into the next set of errors:All global id's for the update actions are filled and added as an attribute. I don't really get what i am doing wrong, but furthermore I really don't understand why the behaviour seems to be so different in the new arcgis feature servicewriter. Why do i have to jump

Hi,I’d like to read a folder of documents and a set of points (say from a csv) and then join these together (filename of documents might match an attribute on the points), then write them all to an ArcGIS Online feature service with the documents as attachments.  I would ideally like to do this in one workspace and not have to hardcode anything.I don’t think this is possible without using one workspace to write the CSV to feature service, and then another to read it and add attachments using the ArcGISAttachmentCOnnector, as that requires the feature service URL, which may not be known at the time of creating the feature service.Thanks,Marc

I have a csv spreadsheet of arcgis server feaureservice urls with the featureclass, name attribute , and projection attribute. This list is long and has multiple servers and I want to pass this list into my workspace and have the FeatureReader transformer read the url and featureservice with a spatial filter for my area of interest (AOI) and then clip out the data I need inside the AOI and only keep the attribute name field in the result, run some additional analysis, then write out the clipped results to a GDB, then move to the next record url in the spreadsheet.. My plan is to reproject the AOI to each projection listed in the csv and then pass the AOI as the creator to the FeatureReader transformer to only read based on the spatial filter. This works great if I do separate feature readers for each URL but that makes for a gigantic workspace and isn’t very dynamic as the spreadsheet and AOI changes frequently. Is there a way to pass a csv list to the FeatureReader in this way and acc

I have a workspace where the DateTimeCalculator is failing for all records. The rejection code is always INVALID_INPUT. And yet, when I view the data in Inspector the two input fields I’m using both list as Date under Data Type. The dates look fine fine (e.g. - 20251216). It fails for any mode - Add or Subtract, Calculate Interval, etc. Any ideas why it’s failing and what I need to do to fix it? (FME Form 2025.1.0.0)

Hi,I need to use many hosted GDB from the web. They aren’t in the same root folder, they all have sub folders.Quebec Hydrographic Network GeobaseIs there a way I could use Multiple URL instead of having to use 15 FeatureReader, one for each GDB?Or a way to download everything from the root folder ?Thank you.

Hi All,I have a complicated question. I created a custom transformer which can read any table from our Postgres/PostGIS database and write in into a list of desired formats (like Geopackage, Excel, Csv, FileGeodatabase).I have an issue when writing to Excel. A certain attribute is shown in a scientific notation. Ik know how to solf it in an none dynamic way. I could use an ExcelStyler setting the cstom number format to 0, or add an attribute xlsx_row_formatting or apply formatting to a attribute by adding an attribute <attrName>.formatting. I just can't figure out how to do this in a dynamic mode. I wish there was a excel writer to disable scientific notation, but there isn't. Any help or suggestions are appreciated. Arno

Hi everyone!I’m learning about dynamic flows in FME and I’ve encountered two issues that I haven’t been able to resolve yet. I’m reading multiple .gpkg files from a folder, but in the output, all the data is saved in a single .gpkg file. I need FME to generate a separate output file for each .gpkg read. How can I set this up? When I insert the Attribute Manager and rename the attribute from acm_id to id, the values of that attribute (acm_id) turn null after the renaming, and the new attribute (id) does not appear in the final output file. Does anyone know how to prevent this from happening? Thanks in advance for any help or suggestions!

I have used a Directory and File Pathnames reader to generate an attribute holding file paths for the .xlsx files that I need to read in to my workbench. My plan was to use a FeatureReader to subsequently read these files in but I am struggling to get this to work.There are multiple worksheets in each .xlsx and I only need the contents of one - in all cases the applicable worksheet has the same name and an identical schema, the column headers are all on the same row as well.I haven’t been able to work out how to configure the FeatureReader as when specifying the dataset from an attribute the parameters to specify ‘Sheets to Read’ and ‘Field Names Row’ are not available.I would be very grateful for any advice - thanks in advance.

We are preparing to upgrade to FME Form 2025.  I am running a job in the 2024 and 2025 version to compare the output.  I have found that in 2024 the FeatureReader didn’t reject any records, but now in 2025 some records are being rejected.  FME(R) 2024.1.3.0 (20240927 - Build 24627 - WIN64)FME(R) 2025.1.3.3 (20251104 - Build 25648 - WIN64)The rejected records say:fme_rejection_message    buffer    Coordinate system does not match first feature's coordinate systemfme_rejection_code    buffer    INVALID_GEOMETRY_COORDINATE_SYSTEM Any thoughts on why the records fail in 2025 but not 2024?I will continue working backwards to see what is unique about these records.

I have an existing process that processes versioned views from an ESRI SQL Server Geodatabase to another SQL Server Database creating SQL Spatial tables with a geometry column that I have reprojected to WGS84 Lat Long. Can I add a second geometry field to the SQL spatial tables and populate that column with Web Mercator geometry? I am trying to figure out a single workflow that would calculate both columns in the same table. Or am I restricted to calculating one geometry column at a time in separate processes?  

Hello, I am struggling with using the ArcGIS Attachment Connector to upload multiple attachments in an ArcGIS Online Feature ServiceI am able to add one attachment but for some reason it will not add multiple photos. This is for a Survey123 that I set up an automation in FME Flow with webhook to add certain features to a hosted feature layer in ArcGIS Online. It runs fine, adds the data, but the moment there is more than 1 photo that needs to be attached, it only uploads 1 photo. Any advice would be greatly appreciated. Thank you!Edit: This works perfectly fine on the workbench (it will add all attachments) but when I try to run it on flow, it will only add 1 attachment. I do receive the following error message:  158 2026-1-2 10:50:31 | Message Type: fme::internal::_v0::py::Exception 159 2026-1-2 10:50:31 | Python Exception <TypeError>: Unknown output tag 'Output' passed into pyoutput() 160 2026-1-2 10:50:31 | Error encountered while calling method `input' 161 2026-1-2

Why does XML header of the FME generated file displays the XML header as <?xml version="1.0" encoding="UTF-8" > instead of <?xml version="1.0" encoding="UTF-8" standalone="yes"?>

Hi everyone,I have a geometry problem in FME and I’m not sure about the best workflow.In the image below, the red point belongs to a feature that is connected to a black point (the black point is the reference).What I need to do is: Take the line highlighted in green. Starting from the black point, find a position 10 cm along the line. Move (or recreate) the red point exactly at that location — meaning: the new point must lie on the line, and must be 10 cm away from the black point. I know the relationship between the red point and the black point from an ID attribute, but my problem is:How do I calculate this “point on line at 10 cm distance” inside FME?  

Hi, I updated a workspace from FME 2020 to FME 2025, and I have an error in 2025 that I haven’t in 2020, with the same input data.On the Excel writer, I get an error on a specific string regarding the validation as UTF-8.ERROR : XLSXW writer: The encoded string is not a valid UTF-8 string: 'S\x00u\x00r\x00 \x00c\x00h\x00a…….'Any idea why and how I can check the strings before the writer so it doesn’t stop the workspace with an error ?

Hello, thereI would like to prepend one dataset with other datasets and return other datasets as output. Please refer to the screenshot below, this should be done with only one workbench. Highly appreciate your thoughts.  

FME Form 2025.1SQL Query in SQLExecutor: SELECT * FROM table_name WHERE ID BETWEEN $(polystart) AND $(polyend);Oracle tableWorks fine in DBeaver with actual ID numbersFails in FME with actual ID numbersFails in FME with User Parameters for the numbersError Message:Execution of statement `SELECT * FROM table_name WHERE ID BETWEEN 35480 AND 35484;' did not succeed; error was `ORA-00933: SQL command not properly ended'.Does FME just hate a BETWEEN statement?Going mad trying to figure out where I am going wrong 😂