Skip to main content
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [eavp-dev] Geometry Properties View screenshots

Jay,


Yes, those parts are bugs I was hoping for some guidance on.


For 8, neither Kasper or I could find the source code for or any documentation or discussion about the issue when we looked over the summer. It's easy to create a simple example that shows the issue, though. Essentially, somewhere the EMF data structures are checking whether or not an object is set to another list or variable specified in the EMF model and removing it from that position when it's put in another one. So when a Vertex is added to a Triangle it is removed from all other Triangles. I tried implementations with both a list of Vertices or three explicit Vertex class variables, getting the same result.


My intuition is that this is because of persistence to XML. EMF doesn't want the same object to appear twice in the tree that gets written to the file, so they make sure it can only appear in one place. 


I'll put in a bug for replacing the table with a NatTable.


For 15, I double checked and may have been over-complicating this. I may just be able to set up the ICEGeometryPage as an ITabbedPropertiySheetContributer to the same ID. I'll try it out and apply the same thing to the ICEMeshPage(which currently has the same issue) if so.


Robert

From: eavp-dev-bounces@xxxxxxxxxxx <eavp-dev-bounces@xxxxxxxxxxx> on behalf of Jay Jay Billings <jayjaybillings@xxxxxxxxx>
Sent: Thursday, September 08, 2016 6:43 AM
To: eavp developer discussions
Subject: Re: [eavp-dev] Geometry Properties View screenshots
 

Robert,

Excellent work!

Both 8 and 15 are confusing and signs like bugs. Could you explain them more please?

In 8, I don't understand why EMF is causing such a restriction. Can you point to the code that causes the duplicity restriction?

In 8, have you looked at using SWT.Virtual for the table or considered using a NatTable? Creating a table should only take as long as is needed to draw the few entries that are actually displayed.

15 eludes me completely. Could you provide more information on the problem here? Certainly the properties view doesn't care if something is an item or not.

Jay


On Sep 7, 2016 11:19 AM, "Smith, Robert W." <smithrw@xxxxxxxx> wrote:

All,


Here are some screenshots showing the functionality of the new Properties View for the Geometry Editor along with some comments about how it works. The GeometryEditorNewLayout file is annotated to demonstrate the list.


  1. Importing a file with multiple shapes will create a union named after the source file.

  2. Now that shapes’ color is editable, selected objects are no longer drawn in red.

  3. The Transformation View has been removed to make more room for the Properties View.

  4. When an operation is selected, all its descendents will be listed as additional tabs in breadth first order.

  5. The Data section on the left controls properties for the model. Changes here should be reflected in any modified files

  6. This is a section for hard coded options available to all shapes. Changing the center will move the shape about the scene.

  7. The Mesh Properties group contains editable properties, such as radius for a sphere. It is blank for imported shapes.

  8. The Triangle Mesh Data displays the coordinates for each triangle in the mesh. The table is initially hidden and will only be displayed when the user clicks the show button, as it can take some time to make a table for large meshes. Triangle data is not editable at the moment, due to EMF data structures preventing us from using the same Vertex for multiple triangles. Changing vertices directly then would cause only a single triangle to change, creating a hole in the mesh instead of deforming it.

  9. The Display section on the right controls how the shape is presented graphically. These are temporary adjustments for visualization purposes and should not be reflected in the data itself. For example, setting a cube’s sideLength in the Data section makes the cube bigger, while setting its scale in the Display section only magnifies the cube in the scene.

  10. DisplayOptions publish their configurable settings to option groups. Each option group will display a set of controls appropriate to the type of data provided. Multiple DisplayOptions can publish to the same option group if their operations overlap. The Display section will be automatically populated from the DisplayOptions present on the shape.

  11. The check boxes next to each option group will enable or disable the option(s). This allows the user to easily return to the default display without having to lose their temporary settings. While deactivated, the shape will be displayed as though that DisplayOption were not applied to it.

  12. The Color group publishes three options, red, green, and blue, and provides text boxes for the user to edit them. The settings are integers in the range 0-255, and the text boxes can also be controlled with the up and down arrows, with ctrl, shift, or page up/down providing slower or faster changes to the value. When a union’s color option is activated, it will override its children's colors.

  13. The Scale group publishes a single option, a double named scale, that controls the magnification of the shape

  14. The Opacity group is an example of multiple options contributing to the same group. The opacity option has values of 100 for “Opaque”, 0 for “Transparent”, and 100 as default. The wireframe option has values of false for “Opaque”, true for “Wireframe”, and false as default. The result is a combo box with three options that configures the values for both. While activated, a union’s opacity will override its children's.

  15. Currently, the Geometry Editor itself will not allow for the use of the Properties View, as it is only an Item in ICE instead of a devoted view. The pin button will keep the view's current content when the Shape Tree View loses focus (such as when moving the camera), but has to be deselected every time the selected shape is changed, making it less than ideal.


​Robert

_______________________________________________
eavp-dev mailing list
eavp-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/eavp-dev

Back to the top