Shape the future of FME with your ideas
Open ideas have been reviewed by our Product Management and are open for commenting and voting.
This is purely just an organisational/visually pleasing feature and has nothing to do with the function of FME, but it would be cool to have an option when you right click on a item on the canvas, and under the “size” drop down, have 2 options call “expand to grid”, and the other as “shrink to grid”. The expand option would snap the corners to the outer boundaries of the intersecting grid polygons, and the shrink would do the same but to the shorter (inner) boundary. I think it would also make the grid easier to use in terms of spacing items evenly on the interface. This is the defaults I get when adding items to the canvas: Items just slightly off filling up the grid polygons. I usually have to manually expand these out to make organising the workspace easier.Keen to know if this has been suggested before or contradicts logic of sizing transformers? Michael
It would be great if loading spinner is displayed while streaming is in process. Current way of displaying two sentences is not as intuitive as spinner like in this examplehttps://getbootstrap.com/docs/5.3/components/spinners/
Please enhance the Rejected port logging. For example, with SQLExecutor and DatabaseQuerier, right now we only see one line from the error logging: “An error has occurred. Check the logfile above for details.”More detailed, actionable errors (as requested for FeatureReader) would remove the need for complex in-process log streaming with dynamic queries.
I love the redirect, but please move this to user preferences instead of being saved with the workspaces. It is way too easy to forget to turn redirect off before saving a workspace and sending it to production. Also, it doesn’t appear that the Redirect option is discoverable by using the FMW Workspace reader. We have 100s of tools, and would be convenient to have ability to search for workspaces that have Redirect turned on.
Since FME 2025 removed the dedicated canvas toolbar zoom buttons which zoomed a reliable 10%, it’s time to fix the only button which we have now.The zoom in/out tool current zooms in/out by 10% relative to the current zoom level. It multiples by 90% to zoom in, and multiplies by 110% to zoom out. This should be changed to a x 90% to zoom in and a divide by 90% to zoom out.Current behaviour → 90%, zoom in goes to 81%, zoom out goes to 89%.Desired behaviour → 90%, zoom in goes to 81%, zoom out goes to 90%.If you zoom in and then out again, it should return to the same zoom state.There are bad zoom levels which cause issues in text rendering, which I’d like to avoid. Multiple zooms in/out will drift the zoom level from the starting point onto a poor number where the text is misshapen or extra bold.
It looks like that either the Postgres Reader and Writer doesn't support multiple column / field constraints. See discussion:https://knowledge.safe.com/questions/78765/multiple-column-key-constraint-is-wrong-interprete.html?childToView=78768#answer-78768If you ask me it's a shortcomming that should be fixed!
Make ProjectWiseWSGConnector able to retrieve and download specific file version and attributes
AttributeExploder: By default the AttributeExploder adds fme_attributes to exploded features. Sometimes you would like this, but in many cases they are unnecessary and must be filtered out. Is it possible to do one of the following options:1. Add an option to choose (or not to choose) whether fme_attributes should be added in an Explode.2. Is it possible to integrate a AttributeExposer to the AttributeExploder in which you can set which (un)exposed attributes should be added and if none are selected none will be exploded.
It would be great if there was a list of all transformers (and their version) and in which version of FME each one was released
Hi there,Depending upon their source Coordinate Systems in FME may or may not have EPSG codes matched to them.For example if our user picks NZGD2K.MountEden from the list I can use the CoordinateSystemDescriptionConverter transformer to extract the EPSG code and use it in the workspace. However if they choose MOUNT_EDEN_2000 then there is no EPSG code associated with it and the CoordinateSystemDescriptionConverter returns nothing.Would it be possible to sync the various coordinate systems so that all available information is available to FME no mater which CS version they choose? I am not able to limit the end users choice in the Server app as I need to cater for all possible coordinate systems.I understand the different coordinate systems have different "metadata" with them, depending upon their source, however this can lead to an inconsistent experience for the end user.Regards,Marc
When adding components to an FME Server project sometimes it can be that I want to add just specific files - e.g., 3rd party python dependencies which are in the root python3.XX folder. 9 times out of 10 its fine to have just the folder, however, if I add the python311 folder to my project because I was just one file from the folder then I end up with all of the other stuff in the folder. Then if I choose to delete the project (and all of it’s contents) then all of the files in that python311 folder are gone too. The same is true if I have just a file sitting in the $(FME_SHAREDRESOURCE_DATA) folder. The fix here is easy enough - just put it in a folder, however, that then requires changing the workspace which may use that file which might not always be that easy or my have knock on effects.
Innovyze, now part of the Autodesk Group, have a number of hydraulic software packages which are used world wide. The InfoAsset and ICM suites are used by the majority of UK water companies to manage asset survey data and model their clean and wastewater networks. The software utilises a proprietary database format, across the suit of software. The spatial data held within the database can we exported from within the application to a variety of standard data formats shp, tab, csv etc although there would be a significant benefit if data could be manipulated without the using the software directly
Hi everyone,I’d like to share an idea that could significantly streamline our workflows in FME Form, particularly when working with a lot of fields with the AttributeManager.Imagine this: an AI-powered enhancement that automatically detects and intelligently renames attribute fields based on your past workbenches or predefined naming preferences.🔍 Here’s how it could work:The AI would analyze field names like horizon_canopy_trees_percentage and suggest a cleaner, more intuitive name like CANOPY_PERCENTAGE. It could learn from your previous projects to suggest consistent naming conventions. You could set parameters such as: Use of capital letters or snake_case Abbreviation rules Context-aware naming (e.g., recognizing that “canopy” and “percentage” are key descriptors) Share an Excel file with fields name that the AI could associate with the fields in the AttributeManager.💡 Why this matters:Reduces manual renaming effort Promotes consistency across projects Makes workspaces easier to read and maintain Saves time, especially in large or complex datasetsI believe this could be a powerful addition to FME’s smart automation capabilities. I’d love to hear your thoughts, feedback, or any ideas to expand on this concept!
Working with readers that are either single merged feature types or datasets that do not expose attributes, e.g. CAD, XML, JSON or OSM PBF, as an author, I’d like to have as an option to be able to see attributes that are imported in the AttributeExposer to show the number of features that have values..Adding an option to show null, missing and empty values would be a nice feature too. When I import from cache, it would be nice to see the attribute value stats when Importing attributes. There’s room for columns to show the count for all.With this feature, the author can pick attributes that have values, keeping the destination cleaner. AI Assist recommends a Tester transformer plus counter to do this.The community hub transformer AttributeValueCounter was a good option to use for this but it does not work in FME ENGINE 2025.0.x. Just some random ideas, when Feature Cache is enabled for a workspace, during a translation run a background process can calculate the stats on the source feature types initially. Perhaps add that to the reading of the schema when a reader is configured to be added to the canvas.Get the unexposed attribute stats immediately optionally. Maybe just have a checkbox on the Import from Feature Cache or Import from Dataset dialog to allow the option.
In the Data Inspector and the Visual Preview pane, I would like to suggest a function that uses an LLM to automatically assign styling to the shown data layers (best effort), based on the name of the layers.This function should allow quick symbolization and automatically suggests fitting colors and symbols for vector layers. It should not replace the existing style dialog.The result should be predictable and reproducible layer styling between workspace executions on the same, or similar, datasets.
Today we did an upgrade to FME Flow 2025.1. Unfortunately we noticed it is not possible to have a clean start (completely back to default) of FME Flow, without doing a new installation. Of course there is an option of overwrite items via a 'clean' back up file (.fsconfig). However, folders already present in FME Flow will not be overwritten if they are not present in the clean back up file. See also this link.It would be nice to have something like a 'clean up button' to restore the FME Flow default without a complete new installation. Is that possible?
Reading a OGC API - Features (WFS 3.0) currently always returns data in WGS84 no matter what storage coordinate systems are available.It would be nice if you could choose to return data in the default storage coordinate system. Or failing that, the crs parameter should be available in the collection query parameter drop down
Users in my organization use SSO through SAML in Microsoft Entra to log into our FME Flow server. Disabling the fmeserver.saml.authentication.force setting (as documented under the FME Flow Settings section of SAML Configuration) allows login to be a single click experience on our FME Flow server. This setting allows the web browser to provide the login credential for the user’s Windows session on their computer to authenticate through the SAML connection. There should be a similar configuration option (or default configuration change) in FME Workbench in FME Form to allow us to let users have a single click login experience when they connect FME Workbench to our FME Flow server. I suspect that FME Workbench uses the WebView2 control to present a browser window to follow the authentication process when the user connects to FME Flow from FME Workbench through a SAML connection. Microsoft has documented a setting that can be enabled on the WebView2 control to allow passthrough of the Windows account credentials when authenticating through SAML called AllowSingleSignOnUsingOSPrimaryAccount (as documented at CoreWebView2EnvironmentOptions.AllowSingleSignOnUsingOSPrimaryAccount Property (Microsoft.Web.WebView2.Core) | Microsoft Learn). With our recent upgrade to FME Form/Flow 2025.1, we want to be able to share our web connections hosted on our FME Flow instance to all users of FME Form. This would be simplified if the SAML login process in FME Flow were a single click (because FME Form is forcing reauthenticating the SAML connection to FME Flow every time we open FME Form). Please allow us to configure (or change the default configuration for) FME Workbench to allow authentication through a SAML connection to our FME Flow server to be a single click experience like how we have FME Flow configured.
Most of my workflows have a lot of scripted parameters with various amounts of boilerplate code just to deal with handling defaults and fallbacks.The parameters with uppercase names are all scriptedThis is all just to handle both cases in the simple scenario where “Recent bound (current time if empty)” is, as the name implies, optionnal and defaulting to the moment where the script was launched if left empty. If I was dealing with attributes, it would literally just be this: I believe we should have a type of scripted parameter that allows using the same facilities and functions for parameters as we do with attributes, especially conditionnal values. This would make using default values and dropdown-presets, as well as format uniformization a lot easier to encode in the parameters without having to include (which often means duplicate) that logic in the transformers, or having to use Python scripts (which do not generate log files if they raise any errors) for simple operations like this.
At the moment the VoronoiDiagrammer only calculates distance "as the crow flies" this may not produce the best results. By taking land boundaries into account the transformer would produce much better results for land-based travel.
Often when I am analysing data with the statistic calculator I only want the summary statistics, but have to wait quite some time for the complete output to run before I can see the results.It would be useful to be able to turn off the Complete port, similar to the Cumulative port
When testing/debugging unexpected behaviour on FME Flow one of the first steps is to use FME Form to run it locally and take advantage of the sweet sweet feature caching.In some cases there can be quite a few parameters which need to be copied which is a real pain. A feature which would save a lot of time is to be able to auto populate the local settings of a workspace based on a FME Flow job.
Please bring back FME wall of fame as it is very useful to demo skillset
Database and Web Connections should be shared via FME Flow in the same way as Deployment Parameters.Doing so would not only be more intuitive, but it would provide more granular security since each user or role can be granted access to only the connections appropriate for that user or role.The process described below is unintuitive because is not consistent with Deployment Parameters, and it would be tedious to set up multiple “Shared Database” of connections each with different permissions.https://support.safe.com/hc/en-us/articles/25407698305037-Sharing-Database-and-Web-Connections-in-FME-Form
There are many transformers that have a tolerance parameter (ex Clipper, Dissolver). While they default to automatic calculated tolerance, for features in lat/long, that generally works out to something like 8.800000000000009e-14. It would be nice if we could set the tolerance on the workspace, and any transformer that does not have an explicit tolerance set would then use the workspace tolerance and not something sub-atomic.
In-depth courses for all FME skill levels.
Explore our extensive collection of how-to guides and tutorials.
Log an issue with us. Our support team is ready to help.
