Difference between revisions of "Sword25/TODO"

Jump to navigation Jump to search
235 bytes added ,  15:30, 15 January 2011
Made the note about text colour even more rambling.
(Colour passing seems to be endian-aware, and swapping R and B is probably wrong because they seem to have the correct values already.)
(Made the note about text colour even more rambling.)
Line 5: Line 5:
* Graphics:
* Graphics:
** Problems with start menu:
** Problems with start menu:
**# Button texts are the wrong colour. According to [http://www.mobygames.com/game/windows/broken-sword-25-the-return-of-the-templars/screenshots/gameShotId,330004/ this screenshot on MobyGames] the text should be yellow, not light blue. The text is set by CBitmapButton:OnCreate() in menu.lua, which calls Text::setColor() in gfx/text.cpp. This is then rendered by RenderedImage::blit() in gfx/image/renderedimage.cpp, where it uses the colour to tint the text data. I do get the correct colour (or close enough) if I swap cr and cb, but that's probably not the correct fix.
**# Button texts are the wrong colour. According to [http://www.mobygames.com/game/windows/broken-sword-25-the-return-of-the-templars/screenshots/gameShotId,330004/ this screenshot on MobyGames] the text should be yellow, not light blue. The text is set by CBitmapButton:OnCreate() in menu.lua, which calls Text::setColor() in gfx/text.cpp. This is then rendered by RenderedImage::blit() in gfx/image/renderedimage.cpp, where it uses the colour to tint the text data. I do get the correct colour (or close enough) if I swap cr and cb, but that's probably not the correct fix since cr and cb already have the expected values. It's strange, though: In RenderImage::blit(), we assume we should draw the data to memory as RGBA, while in GraphicEngine::fill() it's apprently BGRA. Surely they can't both be correct?
** Problems with the exit menu:
** Problems with the exit menu:
**# Wrong shading of 'yes/no' glyphs
**# Wrong shading of 'yes/no' glyphs
415

edits

Navigation menu