Regular Letter Rendering

This letter rendering method is the default one and it is generally the most stable because it was implemented from the start. Every letter is rendered with this method unless single is explicitly toggled on. It involves rendering each letter as individual TextMesh objects (called letter slots) with a maximum of 500 letters that can be rendered using this method alone.

Its main advantage is its compatibility: it works with every features because having each letter be its own TextMesh allows fine control over its meshes. This allows several font effects and other rendering Commands to work correctly while Single Letter Rendering would be incompatible with them.

Its main problem is its inefficiency because it renders each letter in its own object and it each of them uses its own material instance instead of sharing the same one for a given font. This creates a ton of allocations/garbage collection every time rendering needs to happen with several letters on the screen which can lead to poor performance especially on embedded consoles such as the Nintendo Switch. Additionally, it limits more the amount of text that can be shown on screen at once since it is limited to 500.

Rendering process

• In Dialogue mode, sets tailtarget to be talking
• If in Dialogue mode and the current letter isn't the last one, play the current bleep sound with the current pitch and volume settings and accumulate the letter in the backtracking system.
• Get the first free letter slot and store it in letter
• If a slot was available
  • Set the appropriate layer of the letter (this changes the camera that renders the letter):
    • 15 (3DUI) if triui is enabled (will render using the 3DGUI camera)
    • 5 (UI) if the tridimensional parameter is false (will render using the GUICamera, this is the default)
    • 0 (Default) if the tridimensional parameter is true (will render using the main camera of the game)
  • Set the letter's tag to Letter
  • Reserve the letter slot to be the current char, with a font of fonttype, with the parent being the textholder at (current offset, current line - 0.1, 0.0) + (0.0, -0.1), using the current text color, a sort of sort and a size of (size.x, size.y, 1.0) * 0.07
  • If we are using a dropshadow
    • Sets the ds field to the first free slot and set its layer to the letter's layer
    • Reserve the ds slot to be letter's text, with a font of fonttype, with the parent being the textholder at the letter's localPosition + the dropshadow offset, using a half transparent black color if fadeletter is disabled and clear color otherwise, a sort of the current sort and a size of the letter's scale.
    • If fadeletter is true, starts a fade of ds's MeshRender to half transparent over the course of 200 frames
    • Increment the letter's MeshRenderer's sortingOrder so it is rendered above the ds one
  • If any of the font effects is enabled, add a FontEffects to letter with the current effects desired (also sets superglitch)
  • Add GetLetterOffset of the current character with font fonttype and size.x to current offset
• Sets maxlength to the current offset if it is larger than the current value
• If center is enabled, sets the textholder's localPosition to (position.x - maxlenght / 2.0, position.y, position.z)
• Decide whether to do the letter yield. The conditions are:
  • We are in a minibubble's inner call in which case, it will always do the yield
  • If not, we need to be in Dialogue mode
  • If we are, the current speed needs to be above 0
  • If it is, this is dependant on the Text advance: if skiptext isn't active OR noskip is active, then the yield will happen
• If we decided to yield, do it for Speed seconds
• An additional yield could be done of 0.15 seconds except on some conditions using the same base condition than the general yield (with the exception that this ignores noskip). The additional conditions to do this yield are that the current letter passed to charIsPunctuation is true AND it's not among |, ), ¿, ¡, ', /, ¿, ) or ¡ AND It's not among the last 2 letters