xeon

Ok…so that we are talking about the same things… “scene” can mean different things to different people.

Scene in the world of 3D has multiple meanings. In 3D applications you can multiple scenes in within a file. These can be anything from product objects to entire forests. In Verge3D it has two meaningss. The first is based on the Scene Puzzle. A naming convention I wish didn’t exist…It should just be called “Load GLTF” and a more generic name as the overall content within the canvas window.

The default V3D application will load a scene containing the V3D cube a light and a camera. No puzzles. To dynamically replace the cube with something else ( another scene – meaning replace the existing geo with some new geo). We have to use the Scene Puzzle and define the GLTF that we want loaded. But the GLTF does not have any puzzles associated with it.

However if your first scene contains a HTML button named “Button” when you load in the new geo/GLTF the HTML button will remain with the same functionality you originally provided. Loading in the new GLTF does not change your puzzles. So the Button will continue to work. However if you had an object called Cube at the start and you had a set of puzzles that are specific to that Cube or reference that cube those puzzles wont work once the cube is no longer loaded.

To get around this problem. You may find it easier to load all your objects into your base scene inittially, create all your puzzles, then remove the objects you want to load in.
The puzzles objects will turn red indicating they are not loaded but they will work once the new GLTF is loaded.

The Load and Unload sample project is a bit basic and doesn’t quite describe the issues you are having but if all your puzzles are in the first project and you are just loading in and unloading GLTFs you are going to be good.

The Swiss Army Knife Demo is one way to do “overlays” but is typically frowned upon in enterprise solutions because of the iframe. The buttons / menu used in the Swiss Army Knife demo could have been built directly in the project.html file and there would be no need for an iframe. This is basically how using the HTML puzzles would work as well.

However, if you are already invested and you have your menu system working in an iframe overlay then… I would definitely consider the modal puzzle or just add one using CSS if thats an option. Link to basic concept: https://www.w3schools.com/howto/howto_css_modals.asp

In my experience since Blender’s shaders, shape keys and auto smooth functionality has change so significantly over that past few years and Blender will continue to evolve keeping track of the combination of Blender version and Verge3D version is critical for maintaining your apps. Nothing worse than just going to make a simple update and realizing you have to fix so much other stuff just to switch versions of either Blender or Verge3D.

If you don’t already…I recommend adding a comment area to your initial scripts stating what version of Blender and Verge you used on the project so at some later day you can come back read it and then – reset your system to just easily make any required updates or at least know what headaches you are about to encounter if you roll them up to the latest code.

I highly recommend keeping a copy of the blender applications you use and a copy of Verge3D by version number so you can roll forward and backward when ever you need to.

Hi Joeb,
Like all things if you open it up for comments you will probably get a mixed bag depending on who responds.

I think a lot of criticism are given just because we as artists put our work out there to show we have done something cool, unique or informative or to just get feedback. Regardless, you open yourself up for lots of things…and I commend you for it.

There are many things you could do to improve the overall user experience, performance, file sizes, redraws, etc. But some of the things you may have chose as an artistic direction and I would not know…but with respect to the materials and engravings. This is a question of your need. If you need the materials to be actual representiave of what the customer is going to get down the color, location and grain of the samples…then high resolution textures is going to a definite consideration….you have a lot of materials so thats a lot of textures.

You should be able to reduce this down a bit to just a few files though by using Atlas Maps. I would probably organize them by popularity of veneers making the most popular on 1 map and then maybe only 6 total veneer maps. The engravings you could probably get away with just one atlas map.

You may also want to place your phone cases at the orbit point of the camera. This way the person can orbit and zoom without it flying off screen. In simple terms…you may want to put your models at 0,0,0, and your camera target or look at at 0,0,0 then your orbit will be alot more stable and easier to use.

Currently you require the user to make selections then press a view in 3D button. The expereince would be alot nicer if the experience was more dynamic. Select a phone case…the correct model appears. Select a veneer it auto updates on selection. Select and engraving it appears…and eliminate the view in 3D button.

You may also want to ditch the full screen button.

Since you audience is mobile devices…. I would modify the textures, and things so that it is easy to load on your smallest devices since they are the main client. I would not worry about a PWA for this as your clients are going to use it…hopefully buy and move on…they wont come back and in the end all you want them to do is buy.

If you goal is to have the bone cause the green cube to move…. be sure to keep “Deform” selected. When you turn this off…the bone no longer has any influence over the object or vertex group.

Well…time is limited and this is a unique problem.
There are direct ways you could do this with shape keys and then based on some math make sure you dont allow a user to select a value that is a fraction of a panel. You would not use individual screens…it would just be a single poly and you could use a texture for the grid.

Alternatively….creating a dyanamic cloned based array that can change its pivot point and angle of flex is a unique problem. I have only just started looking at this problem. Perhaps, someone on here will beat me to the solution. But for now just picking away at it between jobs.

Ok…well.. good news and bad news.

The “read CSV from” puzzle does truncate leading zeros. Thats a bug or should be a bug.

BUT

You can get around this by using a list puzzle and “load data” directly into the list. This will keep the leading zeros.

See puzzles and console log attached.