glyph-brush-0.6

• GlyphBrushBuilder removed initial_cache_size, gpu_cache_scale_tolerance, gpu_cache_position_tolerance, gpu_cache_align_4x4 public fields replaced by gpu_cache_builder field. This change allows the following changes.
• Add GlyphBrush::to_builder method to construct GlyphBrushBuilders from GlyphBrush.
• Add GlyphBrushBuilder::replace_fonts, GlyphBrushBuilder::rebuild methods. Along with to_builder these may be used to rebuild a GlyphBrush with different fonts more conveniently.
• Replace hashbrown with rustc-hash + std::collections these are the same in 1.36.
• Update rusttype -> 0.8. Compatible with rusttype 0.6.5 & 0.7.9.

glyph-brush-0.5.4

• Semver trick re-export glyph_brush 0.6 without GlyphBrushBuilder.
• Export GlyphBrushBuilderNext returned by GlyphBrush::to_builder.

gfx-glyph-0.16

• Remove deprecated GlyphBrush::draw_queued (use use_queue()).
• Update glyph_brush -> 0.6.

glyph-brush-layout-0.1.7

• Update xi-unicode -> 0.2.

glyph-brush-0.5.3

• Fix queue_pre_positioned cache check for position & scale changes.

glyph-brush-0.5.2

• Disallow GlyphBrushBuilder direct construction.

glyph-brush-0.5.1

• Add GlyphCruncher::glyph_bounds & glyph_bounds_custom_layout functions. These return section bounding boxes in terms of the font & glyph's size metrics, which can be more useful than the pixel rendering bounds provided by pixel_bounds.
• Add GlyphBrushBuilder::gpu_cache_align_4x4 for rusttype gpu_cache align_4x4 option. delegate_glyph_brush_builder_fns! includes this for downstream builders.
• Update hashbrown -> 0.5.

glyph-brush-layout-0.1.6

• Fix missing line breaks for multi-byte breaking chars like Chinese characters.

glyph-brush-0.5

• Removed screen dimensions from process_queued arguments. to_vertex function arguments also no longer include screen dimensions. Vertices should now be given pixel coordinates and use an appropriate projection matrix as a transform.

  This approach simplifies glyph_brush code & allows the vertex cache to survive screen resolution changes. It also makes pre-projection custom transforms much easier to use. See usage changes in the opengl example & gfx_glyph.

gfx-glyph-0.15

• New API for drawing queued glyphs. Depth buffer usage is now optional.

  // v0.14
  glyph_brush.draw_queued(encoder, color, depth)?;
  // v0.15
  glyph_brush.use_queue().depth_target(depth).draw(encoder, color)?;

• Depth test now defaults to Only draw when the fragment's output depth is less than or equal to the current depth buffer value, and update the buffer. Instead of Always pass, never write. This is because depth buffer interaction is now optional.
• Custom transform usages are now expected to provide the orthographic projection, whereas before this projection was pre-baked. The shader also now inverts the y-axis to be more in-line with other APIs. Previous usages can be technically converted with:

  // v0.14
  glyph_brush.draw_queued_with_transform(custom_transform, ..);
  
  // v0.15
  glyph_brush
      .use_queue()
      .transform(invert_y * custom_transform * gfx_glyph::default_transform(&gfx_color))
      .draw(..);

  The new style allows easier pre-projection transformations, like rotation, as before only post-projection transforms were possible. Draws without custom transforms are unchanged, they now internally use gfx_glyph::default_transform.
• Deprecated GlyphBrush::draw_queued in favour of use_queue() draw builder usage.
• Removed GlyphBrush::draw_queued_with_transform in favour of use_queue() draw builder usage.
• Update glyph_brush -> 0.5.