You can optimize the flow if you add different wall categories with parameters. for example a window wall class where you can set “window distance height from floor and left and right edge” in the general settings and width as input in de editing after selecteing or pre placing. same for doors
The wall openings for windows and doors are already movable and fully parametric. Right now I’m working on wall serialization in the world and the intersection tools. I already have parametric doors implemented that appear automatically inside the openings, but I’m starting to think it might be better to let users attach pre-made static meshes instead. That would probably give more flexibility and make the workflow simpler for custom assets.
Yes i can see that! looking good . What i mean is a different wall category, where you for example have your normal wall, a wall with window, a wall with door, a wall with open passage etc, a wall with full windows.. With pre set up values and interactive input values. for example most windows have the same distance from floor (pre set up value), small walls with door (door is centered). If someone want a different door style click existing window or door en change style. It speeds up the workflow and makes it scalable.
placenormal wall/ next wall with window/ next wall with door. then you dont have to do the extra round of placing windows and doors or openings, and adjusting them by eye and then proberlly have to change settings anyway to size and align them.
Yeah, this is actually already on my roadmap. The idea of having predefined wall types with embedded doors/windows/openings makes a lot of sense workflow-wise.
Right now though, most of my focus is going into the math side and learning Unreal Engine properly. My long-term goal is to start building my projects fully inside the engine, and over time that goal is becoming more and more achievable.
A proper parametric wall system like the one you describe is definitely a direction I want to move toward. I also have plans to make tools for slabs, columns, and roofs, but everything in its own time.
Thats the way to go, and it has real commercial potential thats for sure. you solved already a major chunk of the math parametric walls the orientatian and snapping system works already great. i’m currently working on a AR reconstruction app. and ecnounter the same mathmatical issues you will have. Centralize reusuable math in compoments and functions for modularity and scalabilty of your app.So your placable objects can be stored and loaded from a external database and using meta data to get the correct propperties. Then your project is future proof if it grows with hundreds of objects
Thanks a lot for the advice, I really appreciate it.
I’m already following this direction. I’ve centralized most of the reusable math into shared functions and components, and I use that logic across the whole system for snapping, orientation, alignment, and parametric wall behavior.
I’ve also built a centralized mesh builder that generates the elements through Procedural Mesh Components instead of hardcoding every object separately. The goal is to keep the whole system modular, reusable, and scalable as it grows.
To be honest, I’m mainly building this project for personal use, so initially I wasn’t even thinking about it as a commercial product. I just wanted a powerful workflow and toolset for myself. But as the system keeps evolving, I’m starting to realize it could potentially become something bigger in the future.
Right now I’m focused on stabilizing the core architecture and making everything reliable enough for real production use. Hopefully within the next month I’ll be able to put it into real-world use and properly stress test it in actual projects.
I’ve actually been working on this for around a year and a half mostly without much validation or feedback, so I genuinely appreciate your opinion and advice. It really helps hearing this from someone working on similar technical problems.
Your feedback really confirms that I’m moving in the right direction, so thanks again for the valuable advice.
Yup thats the best foundation of a app. Real world application based on real world experience by a user that actually needs it. Not a app that started off as a idea “this might be benificial” .
It got me curious whether my tool could fit into your workflow.
If you can send me the walls as lines with global position/rotation, start and end points (or just length), thickness, and height, I can try to reconstruct the building directly from a JSON file inside Unreal.
For openings, I just need width, height, and X/Z offsets relative to the start of the wall.
I still have to write a few more complex functions for the automatic joining when multiple walls meet at the same point, and after that we can test it.
Could become an interesting experiment between AR reconstruction and procedural architectural generation.
I’m not there yet im currently working on a interpretation layer that uses raw ARdata to display valid candidates for the placing system. but you idea is not far fetch if you target architects and stylist, cause they draw their layouts often in cad based programs with the use of splines or other 3d object lines
I was working on my app and came up for a idea in your app lol. Now you your flow is first a opening cut and then you place a window or door , why not make the window object a cut on its own ? Cause you will eliminate one extra process. so instedad of first making a cut directly fuse the window in the wall en let de window object make the cut and placement.
The idea is actually great, it honestly never crossed my mind to approach it that way.
I already developed a procedural door generation system, but supporting new door variations was becoming too time-consuming, so I temporarily froze that part of the project.
Your approach would improve the workflow a lot by eliminating the separate opening/cut step, so thanks for the suggestion.
Yesterday I finally finished the wall graph and intersection system, and right now I’m mostly cleaning bugs — there turned out to be quite a lot of them
People who don’t write unit and integration tests — and don’t validate their inputs/outputs — suffer
Once I stabilize the current system, I’ll definitely try implementing your idea as well. Thanks again!
Yes and you can combine it with the previous idea i mentioned. For example wall category window (has one dummy window) user select the style ->> want a extra window in that same wall—> same style applied and makes a cut+window and style in once.
Yeah, that’s actually a really good idea and it can be combined nicely. Just let me get to that point and I’ll implement it
Right now I’m fighting with a broken graph and I still can’t figure out where I messed it up.
I finally found my issue — it was a really stupid mistake. I was calling an old function and the graph kept breaking whenever the wall height changed.
It took me 5 hours of refactoring and adding validators before I caught it. Hopefully you manage to find your issue too.
And honestly, I’m not even a programmer by background — I only started learning programming about a year and 9 months ago, so I’m still learning through moments like this
Great. yup same here wish i had your skills, im learning on the fly with a hallucinating chat gpt , wish i had your skills than my app would be finished already haha. my skill is understanding of processes and my brain is a non stop idea producer lol, I can easily tap in someone else his idea.
first month coding and unreal its hella confusing sometimes
Haha trust me, I was exactly the same in the beginning. Everything felt complicated and weird at first, but little by little you start getting used to how Unreal and coding work together.
And honestly, if you already understand processes and systems, that’s a huge advantage for this kind of work.
If you run into a specific problem, just ask — I’ll gladly help if I can
will certainly do we walk a similar path with a lot of overlaps, and you seems to already fixed a lot of problems that i will, no doubt encounter in the near future
Just don’t give up. Be persistent. A year and a half ago I didn’t even know what a variable was, or what cross and dot products were. And I still managed to build a tool that lets me model walls inside Unreal in a style similar to Archicad and Revit. So it’s really not as hard as it seems.
, wish i had your skills than my app would be finished already haha. my skill is understanding of processes and my brain is a non stop idea producer lol, I can easily tap in someone else his idea.