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 XML reader now has a tree view for picking the Feature Paths Configuration Elements to Match, Elements to Exclude.It would be nice if the same tree view was available in the Sub-Elements To Ignore parameter in the XML Flatten Options
It would be great to have these two installed in FME Cloud so that we can produce heatmaps and follow the following guide. https://knowledge.safe.com/articles/68509/how-to-create-a-heatmap-using-the-rcaller.html
As an amateur genealogist for over 45 years, I currently use Family Tree Maker to record my data. I have used others over the years and exported the data out as a GEDCOM to send to others with different software. That seems to be the universal language of the various genealogy programs. I've often thought that it would be helpful to be able to transfer the data to other formats, even just Excel, and have more control over what parts of the data I use so that I can manipulate it in other ways (for example using the location coordinates to map in Arc).
If you start a workspace you get a prompt for parameters You realize that you don't want this anymore. You hit Ok and the worspace starts. But...then you want to turn off the "Run with prompt" but it's greyed out. I think it's better to allow the user to adjust this setting while a workspace is running. I will forget to turn it off and will run into an extra prompt again. Déjà vu
It is often helpful to understand how long a job has run for. Whilst this can be derived from the start and end times, it would be helpful to have this information more visible, particularly for sorting so that long-running jobs could be more easily identified.
I have had a number of clients ask me why the FME Server user name is case sensitive.I did not even realize this before, but they were right.Can an option (configuration) be added to make the user name not case sensitive?Most systems have user names that are not case sensitive.
Send Job Completion Notification to FMEServer Web-UI When User is logged on.Maybe a notification panel at BottomRight corner and open job log upon clicking .
It would be nice if you could drag a file on to a FeatureReader as a way to populate the parameters. It would eliminate the need to have to copy and paste paths or navigate to a file when you already have it open in a file browser in front of you.
It would be handy to be able to clone a Git repository using a transformer. I'd be using this to clone a Git repository onto S3, so the contents could be downloaded more easily. Git pull would also be useful.
I'ts nice to have an option to collapse all or expand all bookmarks at once.
It would be very useful to be able to set a "Translation Timeout" as a workspace translation parameter, like this: If the total runtime of a translation exceeds this timeout configuration, then the translation immediately terminates with a timeout error. To my knowledge, currently we can only control job expiration times for jobs submitted to FME Server, as an advanced parameter of the FMEServerJobSubmitter transformer. That's great for many use cases, but it doesn't help with controlling timeout issues for job chains run locally with FME Desktop, or for hybrid desktop/server workflows. I feel like a more object-oriented approach would involve allowing the workspace author to set a default timeout from within the workspace, regardless of whether it will be running on desktop or server.This translation timeout parameter could then be linked to a published or private parameter, depending on the use case. As a published parameter, the workspace could then be cleanly integrated into a controlled complex job chain.Perhaps there are some under-the-hood complexities that I am unaware of that would prevent this type of functionality?
At present if you add a Transformer on a connection line it removes all the connection vertices you've previously added. You then have to go an recreate each vertex.It would save a lot of time if the connection vertices were maintained after a Transformer has been inserted.
There should be a way to have a dialog box appear in the workspace to notify the user about the output of the transformer. There is the logger transformer, but the message supplied only appears in the translation log. It would be helpful if you could run the workflow and get feedback that could then lead to alternative workflows depending on the output
***Note from Migration:*** Original Title was: Annotations on a connection should not be deleted when the connection line is interrupted. Currently if you have an annotation on a connection line and add a new transformer along the connection line, then the annotation is deleted.Ideally the annotation should be attached to one of the new connection lines, depending on whether the transformer is inserted to the left or right of the annotation tie point.To keep the annotation now, you have to detach it, insert you transformer and then reattach it, but when you have an eloborate workspace with dozens or hundreds of connections, it can be difficult to find and select the right connection in the drop down list.
You can currently set a preferred python interpreter in the FME Options but there is no way to set it as default across a range of installs.This would be excellent to add to the silent command line installer so that enterprise customers can control this setting easily.
We have some memory hogging workspaces and would like to set memory limits by engine so that we can ensure other lighter tasks run smoothly.
It is sometimes necessary to notify external applications of the server name (host name) on which an FME Server job has been submitted. This can be useful, for example, for the external application to monitor the status of a submitted job to determine when the job has completed. The FME Server host name for a submitted job is not available anywhere via workbench. (The FME Server host name is available for the currently running job but this is not necessarily the same as the Server a job has been submitted to.) It would be great if the FME Server host name could be made available on all 3 output ports of the FMEServerJobSubmitter (Submitted, Summary, <Rejected>). See also this question for further details: https://knowledge.safe.com/questions/73769/exposing-web-connection-details-in-workspace.html
The ability to select multiple objects, drag one of their input or output ports to a corresponding output/input port from another object and have them all connect.http:// imgur.com/S1LpfQU
It would be nice to have a warning dialog appear when saving and/or running a workspace, if "Redirect to Inspector" is selected.
When I run a job or create a schedule in FME Server, the workspace parameters are populated as they were when the workspace was saved in Desktop. When I want to change such a 'default parameter' (eg. path changes, email changes, carre centre openings/closes, etc) the update requires the cycle of download the workspace - change it - re-upload. The same with private parameters (that I set to private because I don't want users to change it). I also use these default values to make the workspace URLs short, preferably no URL parameters at all (so users don't need to bother with things, just click the link)The overriding/management of these 'default' parameters (all private, public and FME Server related ones) in FME Server would be really handy.A nice extra would be to set different defaults per service or even per security groups. Eg. download service fail notifications could come to me and streaming service fails to web team, too.
I'm trying to do the "right thing" and use the FeatureJoiner instead of the FeatureMerger where possible, however, there is a pattern which I commonly use which is not supported by the new FeatureJoiner.In the FeatureMerger I can specify hard-coded values for the joins, so I quite often join on "1" = "1" (the actual values don't matter, as long as they are the same). I do this to validate that one path of my workspace has produced a result, before continuing with the rest of the translation, or if I want to join all attributes from features in one translation path with all features from another path.The new FeatureJoiner requries me to specify an attribute for my join values so, in order to acheive the above, I would need to use an AttributeCreator to supply a dummy value to each translation path, before using the FeatureJoiner. This is possible but is an extra step that I don't need to do with the FeatureMerger.Please allow the FeatureJoiner join fields to specify a value, rather than requiring an attribute.
I think it would be nice to have the ability to have a bookmark set to colourer-filled but to also the ability to disable selecting the bookmark when clicking in the body of the bookmark. (Currently when the bookmark is not-filled, it has this functionality).
I find it useful when the Textfile is several MB and automatically shipped via email. As you already know, if you use a simple csv writer or text writer, you should be able to zip the attribute files using the Zip icon button to the right of the dataset parameter. Otherwise in some scenarios you need the attributeFileWriter, instead of the text or csv Writers.
Safe has been standardizing a lot of transformers in the last few years. Think about the attribute handling, a single RegEx engine, adding Rejected-ports.What I am missing is a standardization of the group transformers. I would like to see the following elements on all group-transformers (if possible):- Group By (but that defines the group transformer)- Suppliers / Base / Clippers / Etc first- Parallel processing- Input Is Ordered
How about enhancing the output from WorkspaceRunner, e.g. by adding the workspace path and other relevant run information (time executed, number of features read/written), so it is easier to create a meaningsful output e.g. by email ?When I have a complicated translation scenario, I utilize a "master" workspace, which spawns a number of "slave" workspaces. I would like to capture the ones that fail, and send myself an email with a list of which workspaces fail. As the workspace path is usually hardcoded in each WorkspaceRunner, the output is completely anonymous.Cheers.
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.
