Shape the future of FME with your ideas
Open ideas have been reviewed by our Product Management and are open for commenting and voting.
It would be useful to be able to create a publication rss/georss feed in FME server and subscribe to it in your FME workspace to use the feed to perform another task like update your geodatabase.The rss/georss notification protocol is currently not available
Currently, the Logger only allows the severity value to be hard-coded or specified via a published parameter value. It would be helpful if the severity could be specified by an attribute value on input features.
Add the fourth dimension to the CoordinateRounder, so that it will be able to round not only XYZ but also M (measures, used in linear referencing).
***Note from Migration:*** Original Title was: Restore TM_Prioriy parameter support when submitting server job via web interface Imbedded TM_Prioriy workspace parameter prevent user from submitting server job via web interface.
I already mentioned this at the FME UC (to @NatalieAtSafe for instance), but I believe the forum could really do with some major improvements. Just dropping this here, so people can take notice :)1. Improved Upvote/Downvote system, Best Answers and Q&A OverviewAlthough technically, an "upvote" means that the post should actually be moved upwards, I believe that it makes a forum thread incredibly messy. Call me old-fashioned, but I much rather see a all messages in the thread in a chronological order. I have noticed by several reactions on this forum that other users sometimes also find it confusing (where did the message go?).I propose that the system behaves more like a Like/Dislike, but that the message stays where it is.If the message is turned into the Best Answer, I propose that a copy of that message appears at the top of the thread (just below the question), but that it will also appear at the original location within the thread, like they are doing here, for instance.2. Expose some message functionality as individual buttons Each message shows a little grey gear in the upper right corner, that will display a context menu if you click on it. I would like to see the most important functionality in there exposed as clear buttons. I'm mainly thinking of an "Edit" button and a "Mark as Answer" button, but also the "Reply" option should be a lot more visible. This makes it clearer to first time forum users and it stimulates them to mark a message as the Best Answer, which not only serves as a "thank you" to the person who answered the question, but it's also important for future users with the same problem to know that their question was already answered before.3. Flatten structure, ditch the comment option, add anchor link to original message for replies When you respond to a question, you can choose to post a comment or an answer. When you post a comment, it shows up as a hidden message underneath the question (or an "answer"). Not everyone notices this - it's too subtle. Furthermore, if a comment happens to provide the answer and the person who asked the question wants to mark it as the Best Answer, this is not possible. Instead, he/she has to mark the parent answer, which is a bit silly.I propose that the comment option is removed, so that it's no longer possible to create these "nested" posts. It makes the thread a lot cleaner and easier to read, imho. Replying on a message could then work like this:User clicks "Reply" on a message, types something and posts it.The message does not appear as a comment below the "parent" message, but as an answer at the bottom of the thread (chronological order).The message displays a link above it that says "Replying to @<username>" and links to the "parent" message using an anchor tag (for example). This would also allow people to share a link to a specific message and it causes less redundancy than using some kind of quoting system where the "parent" message gets included (because people are lazy and often quote the whole message ;)).4. Moderation, Question redirects and grouping of Ideas Safe should maybe hire someone who constantly moderates the forum, but it would be better if they assign this task to a few forum regulars (I noticed that @takashi for instance already has the "power" to mark something as "Best Answer"), who should have the ability to:Add tags/topics to messages if the user forgot that so it's easier to find it in the future.Remove offensive posts or spam (which is rare here, fortunately!).Close and redirect a question to another one if it has been answered before (there is a lot of duplicate stuff on this forum).Group ideas together if they are (partially) similar under a <new title>. All votes will then be accumulated (duplicate users filtered out!), so that we can truly see which ideas matter the most. The ideas can remain within their own threads, but at the top of each idea, you will get a link to display the other similar ones. If a user votes for the idea, the linked ideas (i.e. the group) will also receive that vote. In the Ideas overview, we will see these grouped ideas under their <new title>. Clicking on that will expand all linked ones.5. Introduce emoticons! ;)Of course, it's all a matter of personal preference, but I hope that some of the stuff I said makes sense :)
When you publish a script to FME Server with a choice (multiple) parameter, the FME Server interface only shows the first 30 choices. You can still add a new choice by typing it in, but you have to know what you are looking for. It would be better to have the choice (multiple) select box have a larger number.
FME 2016 can read a fmw file but it can't write a workspace.A new writer, please !
It would be nice if the RasterConvolver could have an additional parameters. (In order of importance)Apply kernel:at all valuesonly at specific valuesdon't apply at specific valuesCompute Value:use all values in computationuse only specified valuesignore specified valuesReplace or Append BandCurrently the RasterConvolver replaces the existing band with the results of the convolution. On occasion it would be nice if the results could be appended to the original raster instead of having to send the original raster and the output of the convolver to a RasterBandCombiner.Potential use cases include:eliminating a specific class from a classified raster with a majority filter.filling in nodata values with the average of the surrounding values, but leaving the data pixels untouched.
Redshift is a powerful columnar-based Data Warehouse but requires some tuning out of the gate to enable its full potential. One of the most important components is the selection of the Sort and Distribution keys. Currently, the Redshift writer allows for table creation (IF NOT EXISTS) but does not expand upon the traditional SQL statement for the creation of Sort and Distribution Keys. Requesting an extra option on the REDSHIFT WRITER to allow for the selection of Sort and Distribution keys if the writer is creating the table. Redshift Sort Keys: https://docs.aws.amazon.com/redshift/latest/dg/c_best-practices-sort-key.html Redshift Dist Keys: https://docs.aws.amazon.com/redshift/latest/dg/c_best-practices-best-dist-key.html
The FeatureWriter now has optional FeatureType output ports, it would be nice if the features output through these ports also included the _dataset attribute that is available on the summary port.
Would be nice to provide better info for the {StatusMessage} variable when jobs fail (instead of the generic: "A fatal error has occurred. Check the logfile above for details.") A start would be to include all the lines in the log that contain ERROR: instead of just the last generic line. Partially related to https://knowledge.safe.com/content/idea/44230/provide-better-error-messages.html
The default Knowledge Center setup is "All", which apparently only shows "Articles".I seem to remember when "All" was "Questions" + "Ideas" only.As it is now, "All" is completely useless. Please consider changing it back to Questions+Ideas.And articles look like they are being timestamp-updated daily. This is not beneficial.Cheers.
According to the site "GraphML is a comprehensive and easy-to-use file format for graphs" http://graphml.graphdrawing.org/ A reader for this open XML format would create FME features from these documents including, graphs, nodes, edges, labels with associated styles. Note that some preliminary work has been done using an XML reader in a workspace.
Hi,FME already supports (b3dm) batched 3dtiles format, however does not supports i3dm (instanced 3dtiles). As you know, b3dm tiles are most used format to represent buildings, city objects in CesiumJS environment. I3dm is also another 3dtiles format, mostly suggested to representing 3d models sharing same geometries like trees, billboards, city lights and etc. FME have lots of essential and powerful transformers to create instance 3D models and calculate coordinates of these models (for exp. RandomPointGenerator and PointOnRaster ValueExtractor are very useful to distribute these models on to map). I think it will be a great feature to support i3dm 3dtiles with ability of combining multiple tilesets into one, like in this sample: https://github.com/AnalyticalGraphicsInc/3d-tiles-samples/tree/master/tilesets/TilesetWithTreeBillboards
How to run FME jobs on server to update data by click button on portal for ArcGIS. May be something like widget or script? I am using python on local folder. We need to move this job to server that started by button on portal for arcgis map
The WFS reader currently does not give the option for parameters to be used inside the feature type setup, e.g. for a search envelope. These can be set in the navigator window after adding the feature type, but it would be useful to set them in the initial dialog.https://fmecommunity-ideas.s3-us-west-2.amazonaws.com/2018-09-07-13-47-11-multi-none-mahnamahna-arcgis-t.png">
About possibility in FME Workbench to write in Microstation DGN XFM format, linear elements (polylines) with attributes for a dataset that have lines with elements but other format not supported by Microstation. I need to be able to see in Microstation Bentley Map the polylines and their attributes. Currently, I have the lines and their attributes in MS access format for Geomedia and ESRI and would be great to be able to export them in DGN with XFM attributes to view in Bentley Map.
Writing DGN8XFM, how to write polyline elements with XFM attributes?There is a Reader from DGN8XFM files, dgn with attributes linked to elements, but no writer to be able to obtain a DGN file with attributes to graphical elements? How to create a FME schema that writes DGN file with XFM attributes?
It would be nice if FME can read DJI drone Phantom log files.
It seems weird, that ImageFetcher has an option to auto-detect an image MIME type, but not a way to (optionally) return the found MIME type into a parameter.Please add this feature.
FME Data inspector have a function of Filter Features,But it seems to work only on Feature View,and not on Table View.Most of the time, filtering the features, but also hope that the Table View can keep synchronized,If this function can work both on the Feature View and Table View,It may be more convenient.
We often need to use FME to process DWG 3D model files that are created using AutoCAD vertical products such as AutoCAD Civil3D, AEC, MEP or add-on products such as Bentley AutoPlant, Intergraph/Hexagon CADworx. Each of these products use AutoCAD Object Enablers (OEs) to create graphic objects. For FME to read the model geometry, someone must first save the model to a separate DWG file with proxy graphics. This means an extra step, more file management, and typically results in very large files. It would be great for FME to support the freely available OEs so those steps wouldn't be needed.
The ability to scrape attachment attributes without actually reading/downloading all of the attachments into a Workspace would be phenomenal. There is a huge performance hit with a large dataset to "Include Attachments" so that you can get to the attachment attributes.Use Case:A User simply wants to review their dataset to determine whether or not attachments are present for every feature in a dataset of 5,000+ features, so that they can narrow in on the missing and ensure that attachments are made.
***Note from Migration:*** Original Title was: PointCloudCoercer: Add an option not to convert color components to "fme_color" automatically The PointCloudCoercer (Output Geometry: Individual Points) automatically converts color components - color_red, color_green, and color_blue to "fme_color", if you have set parameters to preserve the color components as attributes.In some cases it might be convenient, but I found that it would just make worse the performance if you want to extract the color components individually. In my case, the workspace has to save every point including components into a MongoDB collection. Firstly, I split "fme_color" generated by the PointCloudCoercer into three parts and restored original values, then wrote the points into DB. It took a long time.Next, I renamed the color components with the PointCloudComponentRenamer beforehand not to generate undesired "fme_color". Result, writing speed with the second method was faster than the first method, 2+ times.It would be better if the PointCloudCoercer would have an option to extract color components as is, without converting them to "fme_color" automatically.
When you need to edit a named connection, you need to go in Tools / FME Options / Database connections or Web connections.It would be great if we could edit connections directly from other places like readers and writers (basically a shortcut to Tools / FME Options / Database connections or Web connections) :And the same to test the connection in place.
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.
