Open ideas have been reviewed by our Product Management and are open for commenting and voting.
It would be a great option if it would be able to store a feature cache after a run and to reload (import) it any moment. Additional it would be great if this option can be extended to the WorkspaceRunner transformer. This allows to look back into a run to see intermediate results.
***Note from Migration:*** Original Title was: TestFilter - Add Text Editor capability to write the Test Clause as a SQL-like statement The test condition in the TestFilter can be further improved for user friendly experience if the user has the ability to build the Test Clause than the current way to choose the attributes and conditional clause.Something like the shown snapshot
when reading autodesk dwg files there are several potential sources (see below) for transformation/projection definition. Please make these more defined and possibly allow selection of the source. Also if this were a format attribute that allow validation at read. AutoCAD Map3d CS definition AutoCAD transformation Esri assigned transformation prj fileI have found files with 2 differing projections assignments and fme has a non-setting default assignment to the esri assigned projection. This is not documented and not even something that can be changed. without knowing what it is doing and over-riding it.
Right now the Feature Type Name can just be added as a text or from a Parameter. I would like to be able to choose this from an attribute as for "any" other format.
Not knowing how many rows I will be writing out, using an Excel template that contains contents below my destination table, when writing out using the Excel Writer, I need the table to expand/add rows so that the contents below is written over. @sammy provides a good example in this post in the FME Forums.
***Note from Migration:*** Original Title was: Add 'Accumulation Mode' parameter to StatisticsCalculator to allow attribute accumulation at Summary port I hope the title already stands mostly on itself. In general I think it would be nice if it was possible to configure which attributes exit the Summary port of a StatisticsCalculator (besides the group_by attribute(s) and the statistical attributes). I would think something like an 'Accumulation Mode' parameter, with options 'Use Attributes From One Feature', 'Merge Incoming Attributes' and 'Drop Incoming Attributes' would be a really nice to have as build in functionality in the StatisticsCalculator transformer to have control over the attributes that exit the Summary port. Just like this parameter/functionality is already incorporated in most other transformers where attributes are accumulated. E.g. most noteworthy being the Aggregator transformer, but e.g. also in the Hull- and BoundingBoxAccumulator, the AreaOnArea- LineOnLine- or LineOnAreaOverlayer, (Surface)Dissolver, RasterMosaicker, etc.Of course there is an easy work around by using e.g. a DuplicateFilter on the attributes that were used at the group by condition of the StatisticsCalculator, and merging these unique features from the DuplicateFilter with the Summary port of the StatisticsCalculator, but still :)
***Note from Migration:*** Original Title was: Markdown table support in the overview/help for Workspaces and Custom Transformers Make it possible for users to create tables using Markdown in the overview / help sections of Workspaces and Custom Transformers.
The dissolver only merges features base on common values. It would be really cool if you could specify range values.
Route 3D objects to this video gaming format for end users to walk through and explore
One of the main advantages of fme desktop is that you don't have to write any code to perform the most complex tasks. This is a visual editor.But when it comes to adding custom crs you have to deal with this:Sometimes it makes you feel like cryingIt would be great to have a graphical coordinate system editor and the ability to import/export coordinate system in various notations like proj, wkt and so on.
Developing a reader/writer for Blender software .blend format. www.blender.org
We need a way to extract the properties of a coordinate system, such as the units. There is at least one custom transformer I found on HUB (CoordinateSystemMapUnitExtractor) but I think the capability to extract properties and use this data in the workspace logic should be available as a simple to use transformer, right OOTB. You can find more discussion on this here (specifically to the Units property, but it should apply to others).
Named Connections for databases and webservices is really useful.It allows authors of FME Workspaces to separate sensitive information into "environment configuration" that in the case of publishing to FME Server also becomes manageble in terms of authorization. It's often the case that the logic of an FME Workspace needs to be spread to a wider group than the sensitive information.By physically separating these, there's less risk of the sensitive information being saved where it should not. However, there are times when Named Connections cannot be used because they are to specialized towards specific applications.This is where this concept would have its role. Named Values should work in the same manner as Named Connections but be very generic on the kind of value that is managed.Probably, the managed value should in fact only be a string value. Access to Named Value would probably have to be implemented as some kind of macro substitution or function call. Some keys to being useful as a concept:It should be available to use anywhere in a WorkspaceIt should be available to use in all stages of the execution of a Workspace (before, parsetime, runtime, after...)It should be documented clearly how not to use it, e.g. assign as an attribute to a FME Feature.In my opinion related ideas:https://community.safe.com/s/idea/0874Q000000TlQqQAK/detailhttps://community.safe.com/s/idea/0874Q000000TkjuQAC/detailhttps://community.safe.com/s/idea/0874Q000000TlGpQAK/detailhttps://community.safe.com/s/idea/0874Q000000TlBAQA0/detailhttps://community.safe.com/s/idea/0874Q000000Tkv6QAC/detailhttps://community.safe.com/s/idea/0874Q000000Tl9DQAS/detail
Currently if you set up a Feature Writer to write to a Text File the text_line_data attribute is automatically populated and not editable. However, this means that is not highlighted green or red depending on whether it exists or not. The FeatureWriter should alert in some way if this attribute does not exist, as it's fairly important for the Text File writer :-)There's also no warning in the translation log to say the attribute does not exist.
Same option like the ListConcatenator (Drop Empty and Null Elements) for the Aggregator. So the output strings, when using many columns, is clean in one transformation step.
The possibility to create a webhook through the FME Server interface is great, but when choosing the parameters I miss the Advanced parameters which you had in the Run Workspace form: You can always add the Job Directives tm_ttc, tm_ttl and so on (https://docs.safe.com/fme/html/FME_Server_Documentation/ReferenceManual/Job_Directives.htm) to the webhook manually, but it's not so obvious and I don't think every user knows this directives
When using the Salesforce Reader and Writer FME leverages an outdated API version to get the metadata it needs to talk with Salesforce. This means that any newer objects, for example most of the FSL Objects, will not be available to map to as tables. This is a very big blocker for us due to FSL being the one place in SF that we use geographic data. We are trying to update our ServiceTerritory object that is in API v38 or later. FME Salesforce Reader Writer is running v37.
Currently, when a Google Sheet reader is set up, you have to specify the spreadsheet URL. Unfortunately, the reader can't differentiate whether the URL is of the whole spreadsheet, or a spreadsheet 'view'. The reader ingests every column and row automatically. GSheet users filter the rows/column via created 'view' to narrow down a subset of data. Hence, it'd be great for the GSheet reader to be able to use a 'view', so the features/records are clean and organized when they are loaded
It would be great if the PythonCaller could get access to FME named connections, even in an obfuscated manner, for use in connecting to web sites and databases. For example, making a web connection available in the PythonCaller as a requests.auth object for use by the "requests" module for HTTP requests: http://docs.python-requests.org/en/master/user/authentication/
The issue is a security one related to workspaces accessing network drives, particularly when run as jobs on FME Server. Rather than having a service account run a separate engine with access to specific network resources, why not allow named connections for network resources in workspaces for more fine-grained control of the security of these resources.
***Note from Migration:*** Original Title was: Add Header list parameters for FeatureReader request like can be do it with HTTPCaller For security reason I need to sent to ESRI header attributes (Referer) to identify the request as FME request and allow access to ArcGis services. What I need is easily available in the HTTP Caller Transformer with the Headers List. Can add this design utility to FeatureReader for next FME release.
I'm very new to FME and only completing the desktop training. Maybe it's been suggested before, but to avoid connections overlapping, could curved, squared and straight connections be incorporated together on the same workspace and this would mitigate against overlapping or at least reduce its occurrence?.
While the csv reader can have a delimiter specified, it would be great if a setting was introduced that would allow the CSV Reader to guess the delimiter of the supplied file (just in the same way generic reader can guess a format). This way the similar files using different delimiters could be read without the CSV delimiter needing to be pasted by early processing to guess the delimiter.
There is no possibility to activate a topic on cancellation. This can happen if the 'queue or job expiry time' exceeds their time or someone cancelled the job by accident. Jobs can exceed due to occupied engines, expired licences, bigger than normal dynamic input datasets (or even bad scripting).
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.
