Authoring

Hi there,Creating many workbenches over the years, and this is very rarely a problem.But it is very rarely a problem because FME automagically selects the postgres writer writes the data using the correct geometry type, not because I have been able to specify.However, here i have a workbench reading in spreadsheet data using a [vertex creator] on the X/Y columns, but when I write the data out to postgis/postgres i find that it arrives there in linestring format. At least, QGIS then sees the layers as linestring!What is super wierd, is that this workbench has been reused for several years, each year with the postgis writer being modified to create a new table for the years data. All previous years it has created point data, this year it has decided on polygon. This happens when i modify the workbench in FME 2022.1.3, or in the current FME 2024.2.3.So, obviously this is unsatisfactory, and I want to be able to tell FME to create that postgres.postgis table with the geometry type of point

Can someone please let me know how I can grab two values from a JSON snippet like below image? Using FME Desktop I am trying to grab two values “portalURL” and "id" from two fragments of "info" and  "events" and assigned them to two User Parameters “PURL” and “PID” Can you please let me know what Translator/former I can use to the task for me?

Hello! I have a problem regarding using a FMEFlowJobSubmitter within a workspace when running on fmeflow using only one engine. When testing the same setup on an old fmeserver i dont have this problem since a “child” engine is created and the procces does not halt. This also works ony my localhost fme flow. However when running on a new fmeflow setup i get the following error:346 2025-2-12 14:53:56 | FMEFlowJobSubmitter_5 (ServerFactory): Reason - 'Error communicating with FME Server host  FME-flow uses a “"FME Engine of a distributed FME Server deployment". Could this cause the issue or is there any other known problems that i can look into? /Fabian 

Hi there,Somehow I've managed to undock the Translation Log window from the bottom pane of FME Workbench and I've been unable to return it to it's original location.Currently, it is locked in the top-left of my screen and shows as a much more compressed version of itself (with no text) - see image.Can anyone enlighten me as to how to restore it?

hello! I am currently working on an API integration project. My task involves reading data stored in an ArcSDE geodatabase, selecting specific data, and posting it to a website. The website accepts data in JSON format, which should follow a specific structure:{  "type": "Feature",  "properties": {    "id": <id>,    "name": <name>  },  "geometry": {    "type": "<geometry_type>",    "coordinates": <coordinates>  }}I read data from ArcSDE geodatabase and the option that converts data into the format above is GeoJSON writer. however, this writer saves the GeoJSON file to a directory, which I do not want. Instead, I need to convert the data into the required JSON format using a transformer, Extract specific parts like geometry and property from the JSON format that can connect with an HTTP caller to POST the data. I'm not sure which transformer would be suitable for generating the desired output. I would greatly appreciate any advice or help! Thanks!

I want to send ~150 users a stock email. No data translation in this one, I just want to read a list and send the email. (also a attachment, with a file that matches mostly matches their email addy but one thing at at time). 

I am building a workspace to do data prep for an AI model training process. This will involve, in this order: Copying a number of image files of type A When copying completed, creating image label .txt files for these when all image copying and label creation complete, copying an exactly equal number of images of type B (from a larger pool than that of the A’s) creating the image label files for these Terminate process when counts match1 and 2 are working fine. I’ve added a counter after stage 2, and I want stage 3 to keep running until the count of type B images copied = count of type A images copied. How might I a) kick off 3. only when 2. is finished and b) stop 3. when count of B = count of A?  

I have been given an xml file which when opened in Notepad++ shows encoding “Windows (CR LF)” “UTF-8” I need to write a new file with the same encoding. My writer has these parameters It writes the file ok but in Notepad++ the file encoding shows “Unix (LF)” UTF-8-BOM”  What writer parameters should I use to get the same output?

I just learned that you can copy & paste Conditional Values from one transformer to another without having to open the "Parameter Condition Definition" windows each time. My question is: When pasting a Conditional Value, you have to select "Paste Conditional", if you select "Paste" it will paste a text equation that fails. I assume this behavior is a bug, can someone confirm this? This is better described in the second observation noted below.The screenshots below show how this works from AttributeCreator to AttributeManager. Observations I had which might be helpful to others are:When copying/pasting you need to select the text first (e.g. "3 Possible Values", not simply the Cell. (Right clicking the cell gives you options to Cut/Copy/Paste/Duplicate the row.) When pasting I noticed that you have to select "Paste Conditional", if you select "Paste" it will paste a text equation and refresh to appear like all is well (

Hello there,i want to mark some points which are already placed at each corner of the building polygons in a DGN(MicroStation) file,i want to mark points which exist within 10m buffer from the road, and also points which at the road facing side of the building not at the back side of the building polygons,when i tried using the buffer it covered all the points including the backside points, if buffer reduced some building points got missedseen the image, the intention is to mark only the blue circled pointslooking for some suggestion, ideas are most welcome thanks. 

I’ve a Microstation DGN file where Building as polygons and each corner of the building polygons have a point, those points have some attributes in the form of MicroStation TAG, i want to check the TAG attribute of the points.Each point will have many attributes like building no, street name, etcif any point in the Building with TAG value ES=1 then the trailing point and successor point must be either ES=1 or ES=6in the below image red point are ES=1 and the orange points are ES=6i want to check the validate this caseany ideas most welcomethanks in advance 

It’s about this error message "No Module named GeoEncoder", how can I solve this error?  

I’m trying to identify the vertex/vertices in a 3D dataset which are at the zmin.The task is very easy but I’m trying to figure out the fastest/cleanest way to do this. The dataset could be quite big and the geometry could be any type (not point cloud or raster).  

Hi,I need your help.I would like to read the data from an excel file that change every week.  Moreover the excelfile have a different name every week.I would like to try to automate the process and as an input i would like to have the directory instead of the file. How can I do that? Thank you for your answer.Pauline 

How to modify Adjacent feature to find a value of two next cells?Thanks 

Hi all,I just wanted to get an idea on how others are working with FME when it comes to inspection.I have FME configured to always open up the FME Data Inspector rather that the Visual Preview in Workbench. Mostly I do this because I like to have a larger window and I’m used to it from the pre visual inspection days.Am I missing out? Any tips on using the Visual Preview? I’m interested to hear from other on how they handle inspection 

Hello , I would like to read all versions name without selecting them in feature reader arcsde esri i need to let feature reader to show me the data from all versions from users ,to know which data that users are currently  working on  .could someone advice me ?thanks in advance 

We have hundreds of mappings we execute via command line.  We have implemented standards regarding the use of source, destination, and logging user parms.We started our 2024 migration and found a bug while trying to pass in our --LOG_FILE parm from the cmnd line (fme ...fmw --LOG_FILE xxx).The workspace will not execute and throws this error:ERROR : Unable to open log file `X:\...\FME\X_\...\Log\mapping_name.log'ERROR : The preceding error has occurred during FME logfile setupFor some reason, in 2024 it is always prefixing the log location with the location of the fme workspace.The only workaround I found to get the log file to write in our desired location was to eliminate the --LOG_FILE parm and use LOG_FILENAME instead.I have opened a ticket with Safe and have not heard anything back.  Having to modify hundreds of workspace executions is a show stopper. 

I have a unique problem. I have a set of polygons which take the shape of numbers. In the example below, the number 1875 has been extracted from a raster. As it sits, the 4 numbers are separated but could easily be aggregated. I’m looking for a way to convert this value ‘1875’ to an integer attribute.I can’t think of a way apart from some kind of AI solution, but thought I’d ask the collective to see if anyone has any thoughts?  

I want to convert a tiff to TINrelief in citygml. Is there a fmw workbench file or step by step guide for this? I’M a complete begineer to fme and citygml format.

I have an AGOL text field with a limit of 1000, but the reader appears to be cutting off the fields at 255 characters. It appears to be happening in the reader directly, not later on in the translations. Any workarounds on this other than just adding an additional text field, or is this a known issue of the AGOL Connector?

Hi,I have a parameter that I have set in an FME Server/Flow app that is conditional on another parameter. I.e. The first is set by the user in an app. If there is a certain value, then it would require you two set a secondary variable.I used to have this set as a simple ‘dropdown’ choice, which then would require a simple text input (if needed). I tried to change the original parameter to a ‘choice list’ instead, to make it easier to follow, but this is causing me issues.When I preview how this looks in FME Form it works exactly how I want to. Though when I try to on Flow it doesn’t show/require the secondary parameter. Is this a bug? Or something I can fix? I am constrained to version 2023.2.1.0 (23774) for at least the foreseeable futureAny advice would be appreciated, I can’t seem to get it to work no matter what I try

Hi,when writing using a dynamic Excel Writer (in my case there are no columns defined at design time - see attached screenshot), the font settings get lost in the output file and also in the Feature Type Writer itselfs when I double-check it.Has anyone solved this issue?Thanks! 

 Hello Forum, I’ve set up a simple workspace to convert a csv with eastings and northings (EPSG:27700) to a csv with long lat (EPSG:4326). This is working OK. The workspace is something I’m going to use quite often, so I’m keen to automate it as much as possible. Is it possible to automatically make the ‘CSV File Name’ value in in the FeatureWriter (Parameters tab) the same as the input csv name. For example, if my input csv is called ‘Properties030225.csv’, then when this reaches the FeatureWriter the default ‘CSV File Name’ value in in the FeatureWriter (Parameters tab) is ‘Reprojected’. Instead, I would want that default csv file name to be the name of the input .csv. In this case, Properties030225.csv. Yes, of course, I could type in Properties030225.csv in, but I don’t want to do that every time. Instead, I want it to automatically pick up on the input file name. How can I do this? Thanks,Stuart 

HiAttached is workspace what wrong with it, as I can't write the output in to feature class inside esri gdb.