Shape the future of FME with your ideas
Open ideas have been reviewed by our Product Management and are open for commenting and voting.
Currently in the FME 2018 for Mac, you have to lay out the transformers in order to use the grids or guides to make a nice layout. Then you have to add in a bookmark, if you want to change the layout or add in more transformers, you have to move the bookmark off the transformers.
Similar to disable transformer, option to make the transformer inactive so that features can pass through without the effect of transformers being applied on the features passing through them in disabled mode
I'm a professional fme user and i want to suggest an online job fair for fme users to develop their skills and experience and to be able to connect with fme partners all over the world, thanks Omar Abdelghany Omarabdelghany90@gmail.com
Our HDFS being HA, there is an active and a passive node. It would be interesting to be able to set both servers as NameNode host in the web connection.
IFC specifications are detailed in an EXPRESS file. A customer would like to be able to validate his IFC to ensure that it conforms to the schema in the EXPRESS file. The output and interface should be similar to the XMLValidator. It should allow the user to pick the EXPRESS files presently shipped with FME, or choose their own from a custom location.
***Note from Migration:*** Original Title was: Add Parallel processing level and Input is Ordered by Group parameters to Clipper
Similar to https://knowledge.safe.com/content/idea/19407/update-to-python-3x.htmlFME Cloud currently loads python v. 2.7.6, however quite a few use cases exist that require 2.7.9+ (or even better 3.x - but that appears to be on the roadmap). Our issue involves SNI limitations with the urllib library that are hampering our ability to connect to SSL enabled REST endpoints. It would be great to have more control over the interpreter in the cloud similar to what we expect in Desktop and Server versions. Thanks!
In the WebTiler, it would be nice if we can specify a parameter to skip creating tiles that only contain NoData. See https://knowledge.safe.com/questions/69315/how-to-filter-out-raster-tiles-that-contain-only-n.html
In the inspector when using table view it'd be great to be able to change the zoom, like you can in Excel. Typically there's lots of values and can be small to see. It'd also make presenting the DI easier.
Could be useful for readers as well.
In Google Maps Directions API it is possible to get the waypoints from one coordinate to another coordinate as polyline: https://developers.google.com/maps/documentation/directions/Request example:https://maps.googleapis.com/maps/api/directions/json?origin=46.9745777,15.6211898&destination=46.9752438,15.6144307Google returns a JSON (because I have requested a JSON) with several information - including waypoints in an encoded format (see line 63, 87 and 101):{ "geocoded_waypoints" : [ { "geocoder_status" : "OK", "place_id" : "EipLaXJjaGJhY2hlciBTdHIuIDk5LTEyMiwgODA4MSwgw5ZzdGVycmVpY2g", "types" : [ "street_address" ] }, { "geocoder_status" : "OK", "place_id" : "Ej1HdWdnaXR6IEJlcmd3ZWcgNTQsIDgwODEgUGlyY2hpbmcgYW0gVHJhdWJlbmJlcmcsIMOWc3RlcnJlaWNo", "types" : [ "street_address" ] } ], "routes" : [ { "bounds" : { "northeast" : { "lat" : 46.9774785, "lng" : 15.6212531 }, "southwest" : { "lat" : 46.9746073, "lng" : 15.6144114 } }, "copyrights" : "Kartendaten © 2018 Google", "legs" : [ { "distance" : { "text" : "0,8 km", "value" : 811 }, "duration" : { "text" : "1 Minute", "value" : 81 }, "end_address" : "Guggitz Bergweg 54, 8081 Pirching am Traubenberg, Österreich", "end_location" : { "lat" : 46.9752518, "lng" : 15.6144114 }, "start_address" : "Kirchbacher Str. 99-122, 8081, Österreich", "start_location" : { "lat" : 46.9746073, "lng" : 15.6212531 }, "steps" : [ { "distance" : { "text" : "0,4 km", "value" : 361 }, "duration" : { "text" : "1 Minute", "value" : 20 }, "end_location" : { "lat" : 46.9774785, "lng" : 15.61923 }, "html_instructions" : "Auf u003cbu003eKirchbacher Str.u003c/bu003e/u003cbu003eB73u003c/bu003e nach u003cbu003eNordwestenu003c/bu003e Richtung u003cbu003eGuggitz Bergwegu003c/bu003e starten", "polyline" : { "points" : "ivu}Gy_j~ASR]RuBh@{@Rg@PGBc@Ra@TUN_@V{AlAY^MPMRQ\\IR" }, "start_location" : { "lat" : 46.9746073, "lng" : 15.6212531 }, "travel_mode" : "DRIVING" }, { "distance" : { "text" : "0,4 km", "value" : 450 }, "duration" : { "text" : "1 Minute", "value" : 61 }, "end_location" : { "lat" : 46.9752518, "lng" : 15.6144114 }, "html_instructions" : "u003cbu003eLinksu003c/bu003e abbiegen auf u003cbu003eGuggitz Bergwegu003c/bu003e", "maneuver" : "turn-left", "polyline" : { "points" : "ghv}Gesi~ALh@X`CzAfHp@hBfA|BJPnAzBr@fAPT\\Z" }, "start_location" : { "lat" : 46.9774785, "lng" : 15.61923 }, "travel_mode" : "DRIVING" } ], "traffic_speed_entry" : [], "via_waypoint" : [] } ], "overview_polyline" : { "points" : "ivu}Gy_j~ASR]RqD|@o@TeAh@u@f@{AlAY^[d@[p@Lh@X`CzAfHp@hBrAnCnAzBr@fAn@p@" }, "summary" : "Kirchbacher Str./B73 und Guggitz Bergweg", "warnings" : [], "waypoint_order" : [] } ], "status" : "OK" }See the algorithm of Google to encode this, so it should be possible to decode the string when dooing the step in the other direction:https://developers.google.com/maps/documentation/utilities/polylinealgorithmOn this page is an example of a JavaScript code to decode the polyline string: https://stackoverflow.com/questions/40877840/google-maps-api-draw-a-route-using-points-of-a-polylineBut unfortunately it's not possible to include the JavaScript library in the JavaScriptCaller:https://maps.googleapis.com/maps/api/js?libraries=geometryBecause JavaScript seems not to be fully supported by the JavaScriptCaller, it would be great to do the decoding via GeometryExtractor or GeometryReplacer.There are already a lot of encodings (GeometryExtractor > Parameters > Geometry Encoding):https://docs.safe.com/fme/2017.1/html/FME_Desktop_Documentation/FME_Transformers/Transformers/geometryextractor.htmThe "Google Maps Polyline Decoder" should be added.
Since 2018.0 it is not possible to select multiple objects within a bookmark without moving the bookmark itself. Especially with bigger workspaces this can be really frustrating. It would be great if this can be possible again.
***Note from Migration:*** Original Title was: improve the generation of 3DTiles. Tiles generated with FME are less performant than tiles generated by AGI/Cesium.
Hi!Could you facilitate to select a list element from the back? For example, when accessing a particular list element in a Tester, I can input a number > 0 in the list element selection dialog. But why not from the back like -1 or -2.Sure I can use a ListElementCounter before, subtract -1 of the result in the following transformer of my choice and submit a _list{@Value(_element_count)} (what I have to type manually, as it's not possible in the list element selection to chose an attribute as value). Or I can use the ListIndexer before (where actually it is possible to choose the elements from the back), and create additional attributes, I don't really need. Why making life harder than necassary? Wouldn't it be better to do this things just in one step using only the list element selelction dialog?Of course this refers to all Transformers where I may need to input list elements as a parameter value.
During the FME WT (Seattle) we were shown that sensitive information/credentials could be removed from a workspace uploaded to FME Server. It would be helpful to have this sort of capability when sharing a .fmw or a .fmwt so credentials are not accidentally stored in an otherwise freely accessible workspace.
Using FME Cloud to run a lot of short jobs can be very expensive for the amount time the server is run for. For instance, if you need to run a translation that takes 5 minutes to run every day, you'd be billed for a lot of server time that isn't used. AWS recently allowed billing on newer linux in intervals less than an hour. It would make some workflows very affordable on FME Cloud if billing was less than an hour. Based on this knowledge base question: https://knowledge.safe.com/questions/37275/fme-cloud-pricing.html
I use the LineCombiner a lot, a nice option to have is able to set start of the line consist of a certain attribute we can select from the list of available attributes =; like; <>; < or > etc. I had to workaround by using Snipper to shorten the beginning of the line so that it's not connecting to the prior line, then use Snapper to snap to prior line without connecting to it.
Quite often we test spatial location by setting one coordinate equals the other but forget to make sure the number of decimal places match since source can come from different databases. It would be nice to have a global variable in the workbench that all the numbers are automatically round to nearest xx decimal.
Hi there, is it a good idea to keep the formatting of the SELECT statement in the MSSQL reader? For example, once I created the adjusted query, like this one:After pressing the OK button and reopening the query field, it looks totally scrambled up like this:An sich this would be acceptable for simple queries like the one above, but if one uses complex queries that are under construction and therefore change quite often, this becomes VERY annoying. Using SQLCreator I loose the processing order of SQL statements, so not really an alternative. And of course one could use a textfile reader and have the contents fired by an SQLExecuter, but hey, what for do we have the MSSQL reader then?
Currently the VertexCreator in Insert Point at Index Mode will reject point features as invalid. These features would benefit from Insert point at index 0 in cases where the orientation of the line matters.
Would be nice to have an option to preserve a schedule even if you set an end (expiry) date instead of deleting it.
***Note from Migration:*** Original Title was: Allow users to pick measures from one dataset in LineOnLineOverlayer when multiple measures exist Currently when multiple sets of data enter the LineOnLineOverlayer transformer, the first set, if has measures, will be used for the segments in the output. If the first set does not have measures, the segments in output will not have measures. The order of the data going in is by the reader and then alphabetically if all the datasets are from the same reader. So if you want measures from a specific dataset, you will have to add an additional reader to host that dataset and make sure it is read in first.My idea proposal is for a new parameter on the transformer to allow the users to pick the measures from one of the datasets.
It is very beneficial to be able to store all custom FME work between versions and installations. FME has "FME Shared Folders". However, while this works for connections, transformers et.al., it apparently doesn't include custom formats developed with e.g. Python. My suggestion is to add this possibility to the "FME Shared Folder" functionality. Basically it would just require FME to scan a few additional folders during startup, so it's not that big change, I think. Cheers.
Right now, there is no ideal configuration to store a Custom Transformer exactly one time and use it within FME Desktop and FME Server as a Linked Custom Transformer. There are several workarounds:1. Create Shared FME Folder --> use the folder within the Workbench --> upload Custom Transformers to FME ServerThis works, but in fact it duplicates the Custom Transformer on each FME Server (Development, Staging, Production). When creating a new version of the Custom Transformer, it has to be uploaded to all FME Servers again.2. Create network share and reference it as a Custom Resource within FME ServerUnfortunately, the FME Server Engine doesn't scan Custom Resources, so a Workspace with a Linked Custom Transformer from the network share will crash in FME Server.3. Share the Engine Resource Folder of FME ServerThis would also work, but I don't like the idea of exposing some server internal folder by sharing it to the network. Furthermore, it doesn't solve the problem of sharing a Custom Transformer across different FME Servers (Development, Staging, Production), because each instance would have its own Engine Resource Folder which would lead to multiple Shared FME Folders within Workbench.4. Use Embedded Custom TransformersNot really an option because it is impossible to use exactly the same Custom Transformer per definition.Thus, FME Server and Desktop should be able to reference any arbitrary network folder as a source for Custom Transformers.
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.
