Shape the future of FME with your ideas
Open ideas have been reviewed by our Product Management and are open for commenting and voting.
The FME Server datadownload and jobsubmitter services have a directive opt_requesteremail where you can specify the email address to use in the success/failure notifications.This directive in not available via the REST API v2 or v3.
Please update GTrans in FME. I frequently transform coordinates between systems and have recently learned that there is a new version of GTrans. It would really help my colleagues and I if the new version was added as a transformer to FME (or the old version would be updated). As of now I have to utilize a pythonCaller and have a bunch of dependencies to databases and trasformation libraries which makes sharing and versioning my work difficult. GTrans is a fantastic resource for coordinate transformation and the new version supports transformations to MGRS directly from Sweref99TM as well as GeoRef which we had to calculate by hand earlier. This would be an amazing addition to an otherwise great software.
Add some photogrammetry tools to FME to enable the creation of 3D models from imagery
When clicking Resubmit on a job that is cancelled or finished - it would be nice to have an option to automatically go to the job-status page for that new job - ie https://XXX.fmecloud.com/fmeserver/#/job/33/summary - Instead of returning to the previous job (32)
It would be nice to have an External Action in the FME Server Automations for uploading files to network folders, similar to the Dropbox Directory or FTP/SFTP.Right now you can easily do this with an FME Workspace, but still it feels more intuitive to do it directly with an external action.
On FME Flow, it can be very difficult to track down what Web Connection or Database connection is being used in a workflow without downloading it. When a web and database connection is used, we should log its name.
A few improvements for logging in FME Server:Option to enable auto refresh of logs - this could use websocket instead of HTTP?Auto refresh of the job status (running, cancelled, completed), time finished, features writtenProvide pagination buttons at the top (in addition to the bottom of the screenFiltering & view to save per user between jobs (e.g. descending/ascending, timestamps, warnings only etc)
Has anyone considered having an option to add details or comments of a workspace at the file level? In Windows Explorer, there are options (when using View --> Details) around different attributes you can view (for images, it's useful to view 'Dimensions' or 'Bit Depth'), but for fmw files, it's be good to be able to edit the 'Comments' attribute here to add detail of the workspace, without the need to open up the workspace itself?We can obviously name the file accordingly, but in some cases, the name would be very long if trying to describe some complex workspaces.
I have a job scheduled to run every x minutes. If the engines get backed up, it could get queued multiple times and start running concurrently when engines are free. I want to avoid this. I am currently accomplishing this by querying the FME Server REST API and using a StringSearcher to see if a job of the same name is already running. If it is, I use a terminator to stop the job. It seems like a GUI option on FME Server would be more elegant. Maybe a checkbox that allows/disallows concurrent runs?
Or ‘list slicing’ of course. For example this list with element positions also as values for illustration:List[0,1,2,3,4]A following AttributeManager wants to extract:Attribute1 = List{0} so would retrieve the value of ‘0’Attribute2 = List{1} would retrieve ‘1’Attribute3 = List{2:} or {List{2:5} or List{2:-1} would retrieve ‘2,3,4’Doing this without python would save time and also those without python skills would benefit. I am aware of the ListSlicer and @takashi could correct me if I’m wrong but it doesn’t quite seem to achieve this if the element range is open-ended or goes from e.g. element 2 to the end of the list. Maybe it’s just an enhancement to that transformer?A ListIndexRetriever type transformer might be a good way to do this, or bake it into attribute transformers in addition to the current single-element functionality.
Esri has new functionality called "Attribute Rules" which store trigger-like behavior in the XML of a feature class with customizable functions. For example, a new asset ID number can be assigned automatically while editing features. As it's stored in the database not at the project level, it is portable and the functionality also persists over published editable web services. Unfortunately, there is no backwards compatibility and the feature classes are no longer viewable in ArcMap or ArcCatalog when these rules are created. It also appears that the FME geodatabase reader is unable to consume these feature classes when the new functionality enabled.Documentation:https://pro.arcgis.com/en/pro-app/help/data/geodatabases/overview/an-overview-of-attribute-rules.htmEsri Video on Attribute Ruleshttps://www.youtube.com/watch?v=-2W1H0xr90oThe rules seem to be stored in the sde.gdb_items table of the database and more specifically in the "definition" XML type field of the feature class row. I created an attribute rule and sequence following this tutorial that adds an asset ID based on an auto-incrementing sequence I named curbseq_agspro to new features. The following XML was appended to the feature class definition XML after using the esri geoprocessing tools:<AttributeRules xsi:type='typens:ArrayOfAttributeRule'><AttributeRule xsi:type='typens:AttributeRule'><ID>3</ID><Name>AssetMgmtLinkID_Txt</Name><Type>esriARTCalculation</Type><EvaluationOrder>1</EvaluationOrder><FieldName>astmgmtlinkid</FieldName><SubtypeCode>-1</SubtypeCode><Description>Update the Asset Mgmt Link ID field with the next sequence value from curbseq_agspro.</Description><ErrorNumber>-1</ErrorNumber><ErrorMessage></ErrorMessage><UserEditable>true</UserEditable><IsEnabled>true</IsEnabled><ReferencesExternalService>false</ReferencesExternalService><ExcludeFromClientEvaluation>false</ExcludeFromClientEvaluation><ScriptExpression>var assetfield = $feature.astmgmtlinkid; IIF(IsEmpty(assetfield), Text(NextSequenceValue(&quot;curbseq_agspro&quot;)), assetfield);</ScriptExpression><TriggeringEvents xsi:type='typens:ArrayOfString'><String>esriARTEInsert</String><String>esriARTEUpdate</String></TriggeringEvents><CheckParameters xsi:type='typens:PropertySet'><PropertyArray xsi:type='typens:ArrayOfPropertySetProperty'></PropertyArray></CheckParameters><Category>-1</Category><Severity>-1</Severity><Tags>Asset ID</Tags><Batch>false</Batch><RequiredGeodatabaseClientVersion>12.3</RequiredGeodatabaseClientVersion><CreationTime>2020-01-06T21:34:00</CreationTime></AttributeRule></AttributeRules>It would be great if FME could read / write features with this functionality as it's a highly useful ability now to create these rules in an Esri supported manner rather than customized database triggers and I foresee their use expanding among customers.Here is a .gif animation showing the rule in action on ArcGIS Online. Notice the assetid_attrule field is automatically updated with a new value (after I pan/zoom to refresh the data).
As a user who may have had jobs that have failed to run in FME Server, I want to be able to see a consolidated list of all the failed jobs that have not been re-run successfully. Current implementation to discover failed jobs in fme serverIn order to discover which jobs have currently failed, I need go to the "completed" tab in "Jobs", I then need to add a filter for "failed jobs" then manually re-run each job one by one. After I re-run these jobs, they still appear in the filtered list which makes it hard for a user to under stand which jobs potentially need to be run again and which ones have been already re-run successfully. What will this feature implementation help solve?This will potentially minimize data loss if users understand which jobs have failed to process and may need manual intervention or simply re-running the workbench manually or in bulk. Where would this failed job queue exist?
Hidden connections are a great feature of FME, but especially when roaming about a Workspace for the very first time, I really miss a button to show all hidden connections.
Often processes will need to go through a Dev/Test/Prod workflow where referenced resources/data/urls change depending on the environment.When only one reference needs to be changed, it is straight forward to setup a Choice Parameter that allows the environment to be changedWhat would be great is if there could be multiple values stored against each choice.these can be referenced in a similar way to either lists or JSON objectsIf Test is selected, then $(PARAMETER[0]) = testapi.com$(PARAMETER[1]) = test.sdeor$(PARAMETER["api"]) = testapi.com$(PARAMETER["sde"]) = test.sde
Likely dependent on ESRI and the AGOL API, and not under Safe's control, but in some cases for example when migrating a hosted feature layer with multiple sub-layers I have wanted to control the layer IDs using unique layer ID assignment - similar to how it is done within ArcGIS Pro. Note this is referring to the layer number e.g. s.arcgis.com/XTtANUDT8Va4DLwI/ArcGIS/rest/services/nz_regional_councils/FeatureServer/0 and not to an ID within an attribute, in case of confusion.If it's not possible it's not possible, but it would be useful for sure. Have looked at the API and can't see a call to set this.
When running a workspace in either FME Server or FME Cloud it would be a great secure feature if FME_SERVER_REQUEST_HEADERS contained the real request headers sent to the server from the client, not the localhost.I hope they choose to include the fmetoken, since then it would make certain streaming services much safer.
As a developer I like to use a LogBased FMEServerJobSubmitter to submit the same job and parameters that can be found in the given logfile. This requires extracting the request from the logfile.
When publishing a workspace as a streaming service it would make sence to have an option to add a directory as a CDN in the current namespace.so if your directory contains ./a.js an ./b/b.css you can refer to them with the same path in your html file after publish.
Certain transformers (StatisticsCalculator, FeatureWriter) drop attributes from incoming features.The decision to keep or remove attributes should, in an ideal world, be down to the workspace designer and not down to individual transformers.Can this please be remedied - even if just by an 'attributes to expose' option on the attribute-killing transformers.
It would be nice to use FME to synchronize contents of directories, where said directories could be any combination of Local (`D:datastuff`), Share (`\\serverdata hings`), and S/FTP (`sftp://example.ca/opendata/lizardness`).Currently we're doing this by using Windows Scheduled Tasks and a variety of scripts utilizing some combination of python, WinSCP, FreeFileSync, and robocopy (depending on what year said script was written). Each of these programs and scripts work well enough for their specific task but I'd like to manage them from the same place we manage our other ETL tasks: FME Server.We could add a sub-process transformer using PythonCaller to any of the above in a workbench but there's no insight into whatever messages the process might emit or ability to respond appropriately to situation X.Alternatively we could recursively use FTPCaller/FileCopy/... for each level of the tree, enumerate contents, and build the parameters from those, but that's a lot of small incremental busy work with meagre gains in each step.Note: I said 'synchronize', but we're only interested in mirroring, making one tree an exact replica of another. A natural extention would be to seek bidirectional synchronization, but that's a much harder problem and I think it's fine/better to add that as a separate idea.A version of this desirable feature was talked about in 2017*, with indication there's an associated Idea, but I couldn't find it, thus this new one.Use Directory and File Pathnames on ftp and sftp?
The workspace runner should add the pid as an attribute to submitted jobs.When submitting jobs in parallel this pid could be used to monitor the submitted jobs to check for their completion. For FME Server you can use the job id if needed in order to monitor things. But for FME Desktop you can never really be sure when a submitted job has completed
Ability to define time when setting up an automation to run, currently it seems when setting up a time slot it takes the current time when I add the time to an automation to trigger the run of the workpace, I would like to explicity set a time like I can do with Windows Scheduler.
In Manage Schedules, there is a column named "Total Runs"; this tells you how many times a Schedule has been run; It would be nice if a FME Flow Setting were created to have a particular Schedule "auto-refresh" the Total Runs at a specific interval. Currently, to view the total runs a user needs to refresh the page or step out of and back into Manage Schedules.
Hello,the company opendatasoft offers the possibility of integrating a geofilter into their api. This comes in very handy when using the User Parameter "Geography". However the dataformat is very specific and doesn't follow any of the existing formats proposed in the GeometryExtractor.The form is as follows : (lat1,lon1),(lat2,lon2),(lat3,lon3)Therefore it would be very helpful to integrate it, in order to fascilitate the usage. Feel free to contact me for further information or to discuss questions.Thank you very much it advance !
***Note from Migration:*** Original Title was: Line in Polygon Lookup: Add Attributes to the Lines from Nearby Polygons (via a buffer, etc.) I am in the 9-1-1 addressing world and maintain a roads dataset along with the different 9-1-1 polygons. My problem is that current tools will give the road the attributes for the polygon that they physically sit in, instead of the one that's a few feet away (see example below). For 9-1-1, these roads must get the attributes of the nearby polygon, not the one that the line sits in.In the example below, the highlighted segment needs to say Seagoville to its left and Dallas to its right, but since it falls entirely within the Seagoville polygon, it will currently assign Seagoville to both sides.
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.
