Clearing Out Pending GitHub Bugs — Handmade Hero — Episode Guide

0:10Set the stage for clearing out pending GitHub bugs¹

0:58ctray.cpp: Fix "enterring" typo

2:17readme.txt: Update links to point to the new locations

5:05"You never upgrade anything during the development of the game"^α

7:10Determine to enable the game to compile without warnings or errors in VS 2015

7:45simple_preprocessor.cpp: Make ParseMember() cast everything to int

10:00On the two different versions of the CRT's maths routines

11:42handmade_intrinsics.h: Make AbsoluteValue() call fabsf()

12:16handmade_platform.h: Determine to enable AlignPow2() to compile without warning

16:54handmade_platform.h: Make AlignPow2() add (Value-Value) to force integral promotion to the size of Value

18:37handmade_entity.cpp: Make DEBUGPickEntity() cast Entity->ID.Value to a umm before converting to pointer

19:29win32_handmade.cpp: Make WinMain() cast the GlobalBackbuffer values to u32

19:53Fix many "hides function parameter" and "hides previous local declaration" warnings

33:50Close that issue and request help for the CLANG compatibility issue

37:12Determine to reduce the PeekMessage latency

38:11On message queues and how Windows implements theirs

43:37On PeekMessage()

47:36win32_handmade.cpp: Enable Win32ProcessPendingMessages() to make PeekMessage() circumvent WM_PAINT and WM_MOUSEMOVE

51:43Run the game to see that it all still works and close that issue

53:54Determine to fix the PushClipRect and glScissor issue

54:30win32_handmade.cpp: Pass a lower resolution to Win32ResizeDIBSection() and run the game to see that we're in bad shape

56:25On food and feeding the world^β

59:48win32_handmade.cpp: Switch to the software renderer and run the game to determine that the problem is only in the hardware renderer

1:01:08Compare the code for the software and hardware renderers

1:06:32Run the game in windowed mode, and resize it to exhibit the bug

1:09:03Consult the glScissor documentation²

1:09:59Consult the glViewPort documentation³ and endeavour to understand the difference between the renderers

1:21:23Run the game and eyeball how the cliprect is scaling

1:25:04handmade_opengl.cpp: Determine to enable OpenGLRenderCommands() to perform the glScissor() in the correct space

1:27:43Blackboard: glScissor coords

1:29:03Blackboard: Normalized device coordinates

1:31:02Blackboard: Calculating our correct glScissor coordinates

1:33:30handmade_opengl.cpp: Enable OpenGLRenderCommands() to correctly scale the region passed to glScissor()

1:39:52Run the game to see that the hardware renderer's cliprect is now correct, and close that issue

1:42:06handmade_render.cpp: Enable DrawRectangle() to AND the StartClipMask and EndClipMask together for rectangles <= 3 pixels wide

1:48:02handmade_entity.cpp: Toggle off rendering of Traversables, run the game and see the effects of the DrawRectangle() change

1:52:35handmade_render.cpp: Change DrawRectangle() to test == 4

1:54:17Recompile and run to see that the drawing of narrow rectangles is now correct

1:55:10handmade_render.cpp: Consider removing the ClipMask from DrawRectangleQuickly(), before closing the issue with it still intact

1:58:25Reflect on the day's work

1:58:58Q&A

🗩

1:58:58Q&A

🗩

1:58:58Q&A

🗩

1:59:11@kknewkles It's rolled over 'round these parts already so Happy New 2017, everyone

🗪

1:59:11@kknewkles It's rolled over 'round these parts already so Happy New 2017, everyone

🗪

1:59:11@kknewkles It's rolled over 'round these parts already so Happy New 2017, everyone

🗪

1:59:35@Jim0_o In the compression code there is an assert(LiteralCount == 0) it seems to fire on some files, should the literal buffer be flushed after the loop in case it didn't end on a run?

🗪

1:59:35@Jim0_o In the compression code there is an assert(LiteralCount == 0) it seems to fire on some files, should the literal buffer be flushed after the loop in case it didn't end on a run?

🗪

1:59:35@Jim0_o In the compression code there is an assert(LiteralCount == 0) it seems to fire on some files, should the literal buffer be flushed after the loop in case it didn't end on a run?

🗪

2:00:55simple_compressor.cpp: Enable RLECompress() and LZCompress() to correctly flush the literal buffer

2:03:06@thesizik Are you going to consider resizing the debug display for lower display resolutions, or would that not really be worthwhile since it's not part of the final game?

🗪

2:03:06@thesizik Are you going to consider resizing the debug display for lower display resolutions, or would that not really be worthwhile since it's not part of the final game?

🗪

2:03:06@thesizik Are you going to consider resizing the debug display for lower display resolutions, or would that not really be worthwhile since it's not part of the final game?

🗪

2:07:50@phillip_green A wee bit off-topic, but have you ever considered setting up a green screen / chromakey for your stream?

🗪

2:07:50@phillip_green A wee bit off-topic, but have you ever considered setting up a green screen / chromakey for your stream?

🗪

2:07:50@phillip_green A wee bit off-topic, but have you ever considered setting up a green screen / chromakey for your stream?

🗪

2:08:49@tukimitzoo What are your 2017 resolutions for Handmade Hero?

🗪

2:08:49@tukimitzoo What are your 2017 resolutions for Handmade Hero?

🗪

2:08:49@tukimitzoo What are your 2017 resolutions for Handmade Hero?

🗪

2:09:22@mtsmox Shouldn't the cliprect only be rescaled at the same point that the offset is done? Only the main buffer should have it

🗪

2:09:22@mtsmox Shouldn't the cliprect only be rescaled at the same point that the offset is done? Only the main buffer should have it

🗪

2:09:22@mtsmox Shouldn't the cliprect only be rescaled at the same point that the offset is done? Only the main buffer should have it

🗪

2:10:36@hayai No high production values? Stream looks good and there's a frame around the scare-cam and a nice little fade-in timer. That seems like more effort than a lot of what I see

🗪

2:10:36@hayai No high production values? Stream looks good and there's a frame around the scare-cam and a nice little fade-in timer. That seems like more effort than a lot of what I see

🗪

2:10:36@hayai No high production values? Stream looks good and there's a frame around the scare-cam and a nice little fade-in timer. That seems like more effort than a lot of what I see

🗪

2:12:34@thezoq2 So that black box is a window, not just an overlay in OBS?

🗪

2:12:34@thezoq2 So that black box is a window, not just an overlay in OBS?

🗪

2:12:34@thezoq2 So that black box is a window, not just an overlay in OBS?

🗪

2:14:04@jessermeyer Please consider demonstrating NVidia's cmdlist extension as a way of thinking about modern graphics dev

🗪

2:14:04@jessermeyer Please consider demonstrating NVidia's cmdlist extension as a way of thinking about modern graphics dev

🗪

2:14:04@jessermeyer Please consider demonstrating NVidia's cmdlist extension as a way of thinking about modern graphics dev

🗪

2:15:10Close it down with a glimpse into 2017

🗩

2:15:10Close it down with a glimpse into 2017

🗩

2:15:10Close it down with a glimpse into 2017

🗩

Keyboard Navigation

Global Keys

Menu toggling

In-Menu and Index Controls

Quotes and References Menus and Index

Quotes, References and Credits Menus

Filter Menu

Filter and Link Menus

Credits Menu