Shape the future of FME with your ideas
Open ideas have been reviewed by our Product Management and are open for commenting and voting.
When publishing a workspace to Server, you have the option of pushing up your database connections. My idea is to be able to download these to be added to your Workbench database connections. This could be helpful to editors who do not have the connections settings, and would also help rebuild your desktop environment (i.e. if you have a new computer or your system crashes.)I suppose this could be a security issue, but my assumption is that if I gave you sufficient privileges to download/publish then you should also be granted access to the db connections. It may also help that the password characters are not displayed on the connection properties dialogs. I also assume these are not stored in plain text.
I usually don’t like hiding code behind custom transformers or keeping bookmarks collapsed, so I avoid that approach. At the same time, I often need to turn off feature caching for specific parts of my workbench. Right now the only way to do that is to collapse the bookmark, which messes up my layout.It would be great to have an option on each bookmark to toggle feature caching on or off without having to collapse it. That way you can see your code and layout exactly as you want, while still controlling caching behavior where you need it.
When accessing the FME Flow WebUI, the initial focus should be set to the username field again.This used to be the case.Currently, it takes four tabs to get there. This isn't particularly user-friendly.
It would be really helpful if the FME Flow RUN WORKSPACE UI could display the workspace start time directly during runtime.I think switching to the log for this is a possibility but unnecessary.
Currently, searching for a location in a workspace app’s geometry parameter window appends the search bar contents to Nominatim’s URL. Example 1:Nominatim URL: https://nominatim.openstreetmap.org/ Search entered in workspace app: “vancouver” Appended Content: “search.php?q=vancouver&format=jsonv2” Resulting GET request: https://nominatim.openstreetmap.org/search.php?q=vancouver&format=jsonv2 What gets appended does not allow users or workspace app authors to include additional parameters limiting users’ search results in the Nominatim URL. Adding parameters that are compatible with the q parameter such as countrycodes, limit, and layer to the Nominatim URL breaks the resulting GET request. Example 2:Nominatim URL with parameters added: https://nominatim.openstreetmap.org/search.php?countrycodes=ca&limit=1 Search entered in workspace app: “vancouver” Appended Content: “search.php?q=vancouver&format=jsonv2” Resulting GET request (invalid): https://nominatim.openstreetmap.org/search.php?countrycodes=ca&limit=1search.php?q=vancouver&format=jsonv2 FME Flow should check if the Nominatim URL already includes parameters designed to limit users’ search results before deciding what to append so the final result is always a valid GET request such as: https://nominatim.openstreetmap.org/search.php?countrycodes=ca&limit=1&q=vancouver&format=jsonv2 The current workaround is to add the additional parameters in the search bar. Example: “vancouver&countrycodes=ca&limit=1” This is cumbersome for users. Posted from:
Currently, when building a workspace app that includes a geometry parameter, the parameter is collapsed by default. This means users have to know to expand it before they can provide any input.It would be great to have the option to set the geometry parameter to display inline (expanded) by default. This would make the input more intuitive and improve the overall user experience.
If I want to find intersection points between two feature classes but I do not want to include the points where lines within the same feature class intersect.As in these two examples: A check box option to be able to exclude intersections within the same feature type would be one solution. Or, having two input ports and having an option to exclude intersections within the same input would be another solution. As stated in one of the above cited threads it would be similar to the opposite results of using the “Group By” function on the fme_feature_type.It seems unnecessarily complicated to have to add topologyBuilder just to be able to do this task as seems to be the general solution offered in both threads above. The resulting functionality would be similar to the Intersect geoprocessing tool in ArcGIS Pro.
I know you can expose list attributes by right clicking on them on the results on transformer before you need them however they can easily disappear and you have no visibility that they were exposed at one point. Also it takes me ages to remember where it is. What I would like is a transformer that look similar to attribute renamer but gives the option to expose the list elements. Maybe call it ListAttributeRenamer where you select the list (in the yellow area) and it exposes all of the elements as attributes. Bonus point if it could expose all of the list elements and defaults to give them a name plus a sequence like 1,2,3 or A, B,C. This would mean you don’t have to guess at the number of list values, it would only push out an attribute if the list has one and if there is a new one it is visible in the transformer instead of buried with you guessing if it exists or not. You could then update the ones you want to a name you want. This would make it so much more intuitive and useful.
Could there be an option so that if no features make it to the FeatureWriter, there is still a way to know that the workspace has run but that there were no outputs? I have found a workaround by generating a count_id from the transformer before the writer and another count_id directly from a creator. From the creator, the count will always = 1. From the transformer, the count varies depending on how many features pass through. These are input to a FeatureJoiner which joins on both count_id’s and if there is any output from the UnjoinedLeft port then this feeds an AttributeFileWriter to inform the user there was no output.
Hi, using other ETL and data processing tools I have noticed diagnostic tools that collect and present in a clear way the processes being performed. I know that there is a log for information and debugging the script, but it would be nice if there was an alternative tool built into FME Form for presenting processes (e.g. elapsed time, function, process, RAM, CPU, etc.). Especially comparing the times of tool operation would give a general idea of the length of the processes (the log is not perfect in this respect)
On complex workbenches with multiple bookmarks, it can be difficult to read the bookmark titles. I think it would be good to be able to adjust the bookmark title text size and colour, to enhance readability and improve presentation. I appreciate this would potentially have complications for bookmark spacing, etc.
Since FME Form 2024.2, the “Generate Workspace” button on the “Start” page ist missing. Please add it again.
In 2024 and previous we had Get StartedBut alas in 2025 we only see The Generate workspace is key for new users in quickly learning how to form a translation. I understand we can get to this from the File menu, however please bring it back to the start tab
Allow exporting/importing of toolbar configuration. When upgrading to a new version or working from a different workspace, it takes time to reconfigure the toolbar manually.
under View menu, you can't turn back on Feature Information window should it be closed by a useraltering the windows view and lack of Feature Information window?The only way to show this window, is launch the hard to find “Info” button in visual preview show hide Feature Information window
In FME Flow Hosted, forgetting to manually pause an instance after use can lead to unnecessary cloud costs. While it's currently possible to create a daily scheduled pause, this method can be frustrating — for example, if you're actively working on the instance at the scheduled pause time, the instance will shut down anyway, interrupting your work and requiring a restart.1. Auto-Pause Based on InactivityIntroduce an option when creating (or editing) an FME Flow Hosted instance to auto-pause after a user-defined period of inactivity, such as 30 or 60 minutes. Inactivity could be defined as: no jobs running, no automations triggered, and no user interaction in the Web UI or API. The timeout should be configurable by the user. Benefit: This eliminates the need to rely solely on fixed scheduled pauses and helps avoid instances being left running unnecessarily — especially overnight or during weekends. Why This Matters: Prevents wasted cloud spend due to forgotten instances. Reduces frustration caused by hard-scheduled pauses that stop active work. Encourages smarter and more sustainable resource usage.
Instances are sometimes brought up for quick development or testing reasons. But may be forgotten about after the fact, leaving an active instance running, burning through credits.An alert that detects when an instance changes to a running status, then sending an email notification every x hours saying this instance is still running. Or a alternative could be a schedule which pauses the instance at the end of the day as a one-off.
When using FME Server the standard approach is to provide the service account access to the relevant files and folders used by the workspaces. This would technically allow other authors to publish and run workspaces on files they shouldn’t be able to access if they know where the service account has access. This idea is to propose that for any reader/writer that accesses files that set of credentials can be supplied to be used for that read or write operation similar to the way web connections are handled.
As the dash (-) is not allowed in ArcGIS Portal File names, it would be great if FME could update that on the fly if an invalid filename is passed in, similar how characters like { } are changed to _ for attributes.
Hi FME team,Referring to the post Read AEC Objects (Geometries and Attributes) in FME | Community we’re also very interested in reading AutoCAD AecDbSpace objects. It would be great if it’s implement in the next version. Thanks,cart
It would be nice if you could select a Web Connection to connect to ArcGIS Portal and have the Portal URL in the reader/writer populate from that.
We run many Parent-Child jobs where the amount of Child workbenches can be in the hundreds with different User Parameter values for each. Therefore it can be very time-consuming to search each log file to find the right one.What would be very handy is a search box in Completed Jobs that searches the User Parameters for jobs with that value.
Hi there,Similar to this idea: I would like to be able to filter FME Flow completed jobs more effectively.Things I would like to be able to do:Exclusion filtering on all filter categories Apply multiple filters at once Be able to select multiple items in a single filter - i.e. show all jobs from repository1 and repository2 Be able to permanently hide certain workspaces or repositories from the job queue and then unhide them if required. For example: we have a job that runs every 5 minutes to query an api and update a sharepoint page. It is simple and low risk if it fails, but it clutters up the completed job list. I would like to be able to hide it. Thanks,Marc
I recently discovered that when using a TestFilter in FME, Null values are treated as numbers, seemingly to behave like negative infinityIf someone is not aware of this (such as myself) this leads to incorrect output data. My expectation was that Null should be routed to the 'Unfiltered' port, rather than being processed as a valid numerical value. It is not even treated as zero, as it is in other applications/programming languages, which would at least have some logic to it. I am of the opinion that this issue has impacted MANY of your users who were unaware of this.If there are default options which can be set somewhere to change this functionality, the default should be to rout the null values to the unfiltered port.try it out for yourself:
I would be great if the PolylineAnalyzer had a group by option as I find it currently only works if a polyline / polygon does not share a boundary with another feature.
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.
