Open ideas have been reviewed by our Product Management and are open for commenting and voting.
Currently, the OBJ writer names each texture noname[n].jpg. This can be a problem when importing multiple OBJs into a model or data store when texture files have the same name. It seems that changing the Appearance Name with AppearanceSetter has no effect on OBJ texture names. My suggestion: iBy default, the OBJ writer should save textures for an OBJ as Basename[n].jpg. Or at least honor the AppearanceName. Currently, achieving unique texture names requires a lot of awkward hacking. This fix would make for improved file hygiene and much simpler workflow. Thank you.
***Note from Migration:*** Original Title was: Write out to separate levels for tagsets and geometry in the Microstation DGN Writer It would be really useful if there was the ability to separate out the levels for geometry and its associated tagset in the Microstation DGN Writer. This is required by my organisation so that the file can pass our internal CAD Quality Assurance tests. At the moment the <tag name>.level format attribute is only supported by the Microstation DGN reader and not the writer.
The crux of the idea is outlined in this question: https://community.safe.com/s/question/0D54Q00008VsdhhSAB/gltf-output-is-transparent-when-it-shouldnt-really-beBut the idea is this:When writing out texture materials to GLTF (and 3DTiles) which contain and alpha channel (e.g., a png), change the default alphamode from BLEND to MASK.This would solve many rendering issues which pop up from BLEND rendering. This is not a one size fits all as BLEND could indeed be what is desiered so as follow up suggestion:Add an option on the gltf reader to select the default alphamode - This is what is used when exporting from Blender (https://docs.blender.org/manual/en/2.80/addons/io_scene_gltf2.html).A step further still would be to allow a geometry trait whitch could dictate which alphamode is used (e.g., gltf_alphamode). At present this is a pretty big issue for us
For example after creating a temporary path I want to download my item to _dirpath and can't pick the attribute, I have to use @value(_dirpath).
We're increasing usage of HTTPCaller to interact with on-premise enterprise system APIs (mainly Cityworks). In reviewing some of the job logs for Automations that contain HTTPCaller transformers, I notice repeated messages like this:HTTP transfer summary - status code: 200, download size: '1061 bytes', DNS lookup time: '2.5e-5 seconds', total transfer time: '0.135342 seconds'While that lookup time is marginal (0.000025 seconds), in a workspace where 10K or more features pass through, it adds a bit of overhead and seems to make redundant DNS lookups (I can't think of a situation at the moment where the HTTPCaller will hit different servers in the same flow of data). Would it be possible to cache the DNS lookup when the first feature passes through and just use the cached value for subsequent features? Or if for some reason there would be switching of servers, cache the unique servers DNS results and use them as appropriate?
Currently we can only read/write feature service from AGOL in fme. As we deal with lot of raster data avilable as Raster service in AGOL. Need a reader for Raster service from AGOL
Ability to read from http://www.maptiler.com/ with a WMTS Reader. Consider a WMTS Reader for this service.
Not just a reader for maptiler (like https://knowledge.safe.com/idea/49606/maptiler-wmts-reader.html) but a wmts reader for general use. i.e. PDOK or your own WMTS.
Clipping Meshes now involves Deaggregating, Surfacedraper, TinGenerator, Clipper, Triangulator and takes quite a lot of time on large areas. Would it be possible to implement this faster within a MeshClipper that takes a Mesh as input and gives the same clipped Mesh as output? Would be great for dealing with large terrain, even though they are tiled.
A lot of users of our 3D data would like to export textured multipatches in ESRI FGDB from our 3D platform. FME Cloud doesn't support that at this moment.
Basically I want to pivot my table so that the Value of the attribute to analyze is the value of the new columns. To make sure I have one attribute per column, I first created a list of the actions based on the same group by as the pivoter with the addition of the column attribute and then concatenated that in one attribute. Now I want to see my group by attributes in the rows, with the column attribute as columns and the values (text) of the Actions as values for their columns.
A quick option to flip the order of requestor and supplier on a FeatureMerger. I often connect and set up the join attributes before realizing my supplier should have been the requestor, or the other way around. It's a bit cumbersome having to both reconnect the incoming lines to the ports + changing the join attributes.
We've heard several people interested in the support for 3D Studio Max file format, i.e. .max files. Please vote for this idea, if you are interested in this support as well. We would really appreciate some more detail on how you'd intend to use this format in FME environment. Do you need Reader or Writing or both?
Can we add a storage tier selection flag on the actions (tier on write) in the ABSC? It would be incredibly handy to set bulk actions on the upload or a modification of items or subset of items in a container path using a 'List' action in one followed by 'Set' action in another AzureBlobStroageConnector. {"Hot", "Cool", "Archive"}I believe the AzCopy action is :$blob = Get-AzureStorageBlob -Container "<enter container name>" -Context $<enter context>$blob.icloudblob.setstandardblobtier("Archive")
It would be useful if there was some way of “closing” the receiver-family of transformers without causing the translation to fail. WebSocketCloser?Currently we can close down by using the Terminator, but that causes a failure..The WebSocketCloser would make it possible to carry on in a more controlled fashion.When integrating a translation in a complex workflow, I often find the success/failure being the most valueable information. Having to use a Terminator in the "normal case" takes away a freedom of keeping the meaning of success/failure distinct.
Like a Tester, but you can pick an area. It would be nice, if a small window with the cached features shows up and you drag an area over the features or a basemap where you can choose your area you want.
The PointCloudFilter let's you build expressions from attributes. ex. @Component(color_red)==@Value(_Red) && @Component(color_green) == @Value(_Green) &&@Component(color_blue)==@Value(_Blue)It let's you supply a userParameter with an expressionex @Component(color_red)== 178It let's you supply an attribute containing a numeric value (ex 178)But it does not allow you to supply an attribute containing an expression |ERROR |Expression Evaluator: Attribute '_ptFilter' has non-numeric value'@Component(color_red)== 178 && @Component(color_green) == 178 && @Component(color_blue)==0'This makes it very difficult to build dynamic filters where the number of clauses is variable.
It would be very helpful to improve the way selected transformers a.o. elements in workbench are highlighted.In newer versions of FME it is very hard in Workbench to detect the highlighted transformer. Current behaviour would be better described as "low-lighting" instead of "high-lighting".Even in a zoomed in state I have a hard time to see the "light-blue" coloring behind the blue transformer. In a larger zoomed out extent with a larger number of transformers it is next to impossible.I remember old times like in FME 2012 when I could see any marked tranformer on first glance even among several hundred. The marking back then was prominent and user friendly. Now it is like a hide and seek game.Some ideas an how to improve the situation:use different color, e.g. redmake marking more prominent with corner squares like in older FME versions...
A nice feature for the Transformer Package safe.emailer would be to have a 'response to' field. Because the sender addresses do not always have to be the reply-addresses.Can you please add this in the transformer. :-)
Usually the "Snap to Grid/Guides" function is helpful to arrange objects orderly on the Canvas, but sometimes it's also annoying when move an object subtly. Although you can disable snapping through the "Grid and Guides" dialog, it might be convenient if the "Enable/Disable Snap to Grid/Guides" command would be added to the right-click menu on the Canvas (and/or a tool button), so that the user could switch enable/disable snapping quickly.
A new option to automatically set the cell width for the Excel Writer.
Can we consolidate the paging interface between the Forum and Ideas? At present they are different and require different user behaviours to navigate.Ideally I'd like to see a page selection option on both sites along the lines of:Page 1, 2, 3, 4, ... n-3, n-2, n-1, n <Enter page>Currently in Ideas there are abut 650 pages and we have to click "Next" 50 times to get to page 51. Just imagine trying to get ot page 640! Once in the middle of the list, if we hit the browser back button (common behaviour in these sorts of searches), instead of the "Return to list" button, we are returned to the start of the list with the prospect of yet another 50 clicks to get back to where were.In the Questions Forum I'd like the same page behaviour, but I'd also like to see, in a sidebar perhaps, My Questions, Questions I've Answered, Questions I've Commented On, so they can be quickly accessed.
Native support for AMQP
When uploading a file on S3 with S3Connector, it is possible to define metadata such as content-type, which is great.Why not allowing the same upon folder upload for the files located in the folder ?I have a large number of images to upload from a directory tree and I'd need them to be flagged as "content-type: image/png" instead of "content-type: binary/octet-steam" which is the current default.
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.
