@tadema said:

Hello agamemnus, I think you need to put a lot more work into the model, at the moment it's confusing. You state it's a research building but it has no doors or windows! no sense of scale so it looks like a big blue LEGO model.
John

It's to keep people in! Heh. Thanks for the input guys. I'll consider some doors/windows!

Come on guys... do you like it? Hate it? Does it look like a good basket concept?

It is a research building, untextured. I have done several variations and I present two here. Each step is (likely) a floor.

research building new 3g.skp

research building new 3h.skp

It took me an hour to code, because I haven't coded in Ruby and the Sketchup API for a while. And, actually, I plan to use this sort of script for more than one model. Setting up 64 scenes would probably take me 2 hours, and it is very annoying to be doing the same repetitive work for that long, so it would take even more time. I guess it depends on if it takes you a month or an hour to make that script..

@yorik1984 said:

Any scens can make uniq combination of layers. Every layer can be visible or invisible of any scens. Prepare scens before export. Do not reinvent the wheel.
Repeat

@unknownuser said:

Learn basic SU tools!!!

And I repeat, you didn't read the post.
8 directions
8 rotor rotations
64 scenes for all the combinations

..or making a script? Go make your 64 scenes. I'll make the script.

I see. That's an interesting trick, but it does not allow you to combine layer variations. Feeling yorik1984 needs to read more thoroughly before yelling at people.

My model doesn't have any animation.. I think my Ruby script probably works better than your "basic" tools.

Edit: besides which .... "image set"? It doesn't exist. Are you pulling my leg?

That is quite shameful. I did not know that the ToS has changed so radically. Even if it did, I agree -- the buyers of the design probably don't know it was stolen.

@renderiza said:

@agamemnus said:

I created some code to render a helicopter. Unfortunately, for some blasted reason, most of the scenes are not being rendered correctly. Help!

Can you be more specific when you say "rendered correctly" because I am not 100% sure what you mean by this. Do you mean the scene style rendering options where not updated?

Yes... all of the camera didn't update, so the image doesn't render as intended. Had to do "view.camera = pages.selected_page.camera" to get it to work correctly.

That is very nice. I like the proportions.

I created some code to render a helicopter. Unfortunately, for some blasted reason, most of the scenes are not being rendered correctly. Help!

Edit: ok, never mind. I added "view.camera = pages.selected_page.camera" .. that seemed to fix things. That's pretty odd that the new camera target isn't being copied. Is it a Sketchup 8 bug?

Fixed file:

require 'sketchup.rb'

module Render_isometric
 def Render_isometric.do_render
  model = Sketchup.active_model
  pages = model.pages
  layers = model.layers
  view = model.active_view
  # Set up and get pointers to the appropriate layers.
  image_base_name = "osprey"
  
  scenes_extension   = ["bottom", "bottom-left", "left", "top-left", "top", "top-right", "right", "bottom-right"]
  rotation_extension = ["rotation 1", "rotation 2", "rotation 3", "rotation 4", "rotation 5", "rotation 6", "rotation 7", "rotation 8"]
  rotors_extension   = ["rotors forward", "rotors up"]
  for i in (0 .. scenes_extension.length - 1)
   pages.selected_page = pages[scenes_extension[i]]
   view.camera = pages.selected_page.camera
   pages.selected_page.transition_time = 0
   pages.selected_page.delay_time = 0
   for j in (0 .. rotation_extension.length - 1)
    current_rotation_layer = layers[rotation_extension[j]]
    current_rotation_layer.visible = true
    for k in (0 .. rotors_extension.length - 1)
     current_rotors_layer = layers[rotors_extension[k]]
     current_rotors_layer.visible = true
     keys = {
      ;filename    => "c;/vehicles/" + image_base_name + "-" + scenes_extension[i] + "-" + rotation_extension[j] + "-" + rotors_extension[k] + ".png",
      ;width       => 256,
      ;height      => 256,
      ;antialias   => true,
      ;transparent => true
     }
     model = Sketchup.active_model
     view = model.active_view
     view.write_image keys
     current_rotors_layer.visible = false
    end
    current_rotation_layer.visible = false
   end
  end
 end
end #end module


Render_isometric_menu = UI.menu("Plugins").add_submenu("Render Isometric")
Render_isometric_menu.add_item("Do Render") {Render_isometric.do_render}

I already made a script to do just that. Running into problems with that, however.. seems like the scenes aren't being .. calibrated... properly.

If I can select the layers in Layout, then I can create the 64 different views in Layout, then export to png.

@dave r said:

Create the scenes you need in SU with the appropriate layers turned on or off as needed. If you need to change what is visible in the viewport in LayOut, make the changes to the scene in SU and update the reference in LO.

Ah, that is going to be a problem. I am creating a rendered vehicle image (helicopter) to use in a game, with 64 different views. That's 8 directions, plus 8 rotor rotations. That would be 64 scenes. Probably better to start working on my custom render plugin I guess!

Is there any way to select the model's layers for content, instead of just the view?

Secondary (slightly unrelated) question: is there a place where I can modify a group/object's position and size in a text box (something that it is always visible, like a regular setting... not like the bottom text-box copied from Sketchup)?

Interesting thought. We ended up doing it differently, but I now need that for something else. Thanks, Chris. I will try this method. Now I just need to figure out how to stick Sketchup into CentOS!!!

I just read the things they offer, and in my opinion, they are not really looking at the fundamentals... plus they are flat out duplicating some things apparently? (like their extension warehouse) I guess Google ran out of "cool" ideas, then sold it to Trimble. Now Trimble is running out of "cool" ideas.

Here is what I'd love to have in a new version, which they keep missing but I think is the most important though mundane:

1. Faster Ruby plugins... must-have.
2. Ability to selective disable/enable different inferencing mechanisms (red/blue/green, same-line, etc)
3. Built-in sub-layer support.
4. A new mechanism for rotation that lets us automatically rotate from a specific point (or, the center of the bounding box..) with a pre-defined axis (select your axis from a list).
5. More exposure of the program's controls to the Ruby API...
6. More polygon fidelity (so I can zoom in without the polys evaporating).. might require 64-bit...

I like it.

Anyone?

Thanks to whoever moved the thread. We will probably be doing this in 2D via some CSS, but it would be nice to know if this sort of thing is possible.

Renderiza:
Here is an example. You might be a company selling a common household item or trinket with an engraving, and you want to show your customers how that item will potentially look like.