Sorry for the no updates - been hella busy
Decks are cleared and some progress is being made.
Sorry for the no updates - been hella busy
Decks are cleared and some progress is being made.
The whole spocklers team be like:
Great work Jonathon, is it available for testing? Is this an FME only solution or does it require a separate speckle connector as well?
Not yet available no. I don’t think it will be long before I can share alpha goodies. I have a sort of draft ‘roadmap’ to release, but this is a spare-time effort rn.
Loosely the plan is:
Because It has to feel like doing things the “FME way” I don’t think a strong UI will come quickly. Really my focus is looking at getting as many core conversions working first.
If you were content with just reading Display Meshes with data (kinda like what you already have working) a version 0.1 could probably be possible sooner. What did you have in mind?
I currently have only a vague grasp on the writing side in terms of what might be the use-cases… does an FME Speckle writer in fact attempt to have separate writers for Revit/QGIS/Rhino etc. commits a la the rest of FME, or…?
Let us know if we can help in any way on the Speckle-side of things!
Let us know if we can help in any way on the Speckle-side of things!
48hour days?
FWIW and in case it’s of use to anyone, I have stubbed out docstrings for most of the FME Python API
Entirely done to speed up my work on the Object translations, but sucked time.
Enjoy.
Revisited the mesh translation since the new displayValue
change.
I need some time to look into translating the Metal-based materials into Specular-based ones (perhaps the reverse of something in speckle-sharp)
Well, I don’t know FME, but this looks amazing! What’s up with the glTF viewer?
FME converts anything to anything (almost) gltf was a quickly I didn’t need to launch any boomer cad software to demo.
I’ve been using FME for 1+ year in conjunction with Dynamo to create Revit models from CAD floor plans and a spreadsheet of level elevations. FME does all the “interpreting” and outputs a spreadsheet for Dynamo to generate a model. This includes, walls, floors, doors, columns, area boundaries, and fixtures, currently. I plan to expand to all disciplines and more complicated assemblies, anything where vector geometry references objects (Family Types, etc.). FME is very powerful and the Safe peeps are great.
My goal is to generate models very quickly (< 1 Hr) at little/no cost. Maybe Speckle is the missing piece between FME and Revit? I will be checking it out.
Would you be looking to use FME to create “native” Revit objects in Speckle and then read the streams in Revit?
For now, you’ll be able to stream in Speckle to Revit whatever geometry FME puts there, but building Revit types…
… is what I have been hovering around not doing. Mostly because Kits and object extensions as such are C#. I’ll be honest I’m vague about it all but if there was a transpiling of the Objects library to specklepy this would make things easier. ??? ¯\_(ツ)_/¯
Hi @jonathon,
I don’t know if it’s possible to create Revit objects from FME, as I thought the API only works in their app. Safe has been working on a Revit Writer, but it remains to be seen if it happens.
What I can do now is manipulate geometry and data to give the Revit API what it wants. I’m not afraid of C#, there is a lot of info out there. My coding skills are limited regardless of the language!
But, after learning more about IFC and Speckle, I am going in another direction. My frustration with Revit has overshadowed it’s benefits, and I am realizing open-source solutions are the way to future-proof our data.
Looking at the Speckle code, they have most things mapped, so I think it’s only a matter of time before it’s possible to go direct from Excel or FME to Revit and back.
@Rob at a recent Community Standup showed us how he mapped our C# classes in SpecklePy, so we might be not too far from being able to create Revit objects from FME directly!
When is the next one of those?
Yet to be scheduled! Probably around the last week of April or first of May
Thanks for tagging!
@jonathon
In my case, I mapped the C# classes to Python classes, using Python code. I’m not familiar with FME, but the example might still be of use. Regarding that @teocomi, you mentioned that a tutorial on this class mapping would be welcome, which could maybe help in this case.
@m-clare and me could actually work on this tutorial. What would be the prerequisites for that, do we need to deliver you the information only, or should we generate the full docs ourselves? As this is a bit off-topic here, would be nice if you can send a message to us with some advice, if you’re still interested of course.
Thanks in advance.
Of course we’re interested - there are no prerequisites, you can prep it in whatever format you’d like (word, gdoc, notion etc) and then we’ll just copy-paste it in our blog. Alternatively, video formats are great too!
If you need any help along the way just ping us and we’ll help
To bump or not to bump, that’s the question (previous post was 10 months ago).
What is the status of the FME connector? Any plans on opening the repo?
There is some interest internally to explore this connector a bit more and a potential use case on a big project.