<feed xmlns='http://www.w3.org/2005/Atom'>
<title>SpaghettiKart/src, branch main</title>
<subtitle>SpaghettiKart: Mario Kart 64 PC port</subtitle>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/'/>
<entry>
<title>Switch LUS to port-maintenance branch (#727)</title>
<updated>2026-08-02T05:22:35+00:00</updated>
<author>
<name>MegaMech</name>
<email>MegaMech@users.noreply.github.com</email>
</author>
<published>2026-08-02T05:22:35+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=33817cd1100ab560b16e19af34488de3b5a6b7eb'/>
<id>33817cd1100ab560b16e19af34488de3b5a6b7eb</id>
<content type='text'>
* lus update  compiles but not run

* Game runs, fix demo bug

* Switch lus to port-maintenance branch</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* lus update  compiles but not run

* Game runs, fix demo bug

* Switch lus to port-maintenance branch</pre>
</div>
</content>
</entry>
<entry>
<title>Remove As Much Code As Possible and Still Pass Tests (#732)</title>
<updated>2026-08-02T05:19:52+00:00</updated>
<author>
<name>MegaMech</name>
<email>MegaMech@users.noreply.github.com</email>
</author>
<published>2026-08-02T05:19:52+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=97f28334329d50a9cf2fb7ff1ee57ae0c18de92c'/>
<id>97f28334329d50a9cf2fb7ff1ee57ae0c18de92c</id>
<content type='text'>
* Update render_objects.c

* Remove unused HUD texture function declarations

Removed several unused function declarations related to HUD textures.

* Update skybox_and_splitscreen.c

* Remove unused set_viewport2 function declaration

Removed unused function declaration for set_viewport2.

* Update code_80005FD0.c

* Remove set_bomb_kart_spawn_positions function declaration

* Update skybox_and_splitscreen.c

* Update skybox_and_splitscreen.h

* Modify Draw function to adjust crab visibility

Commented out code to prevent premature disappearance of the crab when the camera rotates.

* Update Crab.cpp

* Update Crab.h

* Update Crab.cpp

* Delete unused texture rendering functions

Removed unused functions related to texture rendering.

* Update render_objects.h</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* Update render_objects.c

* Remove unused HUD texture function declarations

Removed several unused function declarations related to HUD textures.

* Update skybox_and_splitscreen.c

* Remove unused set_viewport2 function declaration

Removed unused function declaration for set_viewport2.

* Update code_80005FD0.c

* Remove set_bomb_kart_spawn_positions function declaration

* Update skybox_and_splitscreen.c

* Update skybox_and_splitscreen.h

* Modify Draw function to adjust crab visibility

Commented out code to prevent premature disappearance of the crab when the camera rotates.

* Update Crab.cpp

* Update Crab.h

* Update Crab.cpp

* Delete unused texture rendering functions

Removed unused functions related to texture rendering.

* Update render_objects.h</pre>
</div>
</content>
</entry>
<entry>
<title>Fix the warning-flagged real bugs across the codebase (#691, 1 of 3) (#725)</title>
<updated>2026-08-02T04:29:56+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-08-02T04:29:56+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=8629bec3205b35b6f052a5f0a0164b8132ac1304'/>
<id>8629bec3205b35b6f052a5f0a0164b8132ac1304</id>
<content type='text'>
* menu_items: fix the real-bug-tier compiler warnings (#691 batch 1)

Six fixes, all in menu_items.c, all verified by menu playtest on macOS:

- BSWAP16(*color0++) advanced the pointer twice per pixel on little-endian
  builds (the macro evaluates its argument twice) and mixed bytes from two
  different pixels. Read first, increment separately. Note: this function
  (func_8009A9FC) currently has no callers, so the corruption was latent,
  not live.
- func_8009B9D0 fell off the end on a lookup miss, returning garbage (the
  existing code comment already suspected this). The miss path now returns
  the display list head unchanged, i.e. draws nothing.
- update_ok_menu_item read an uninitialized stack slot for unknown menu
  item types; now selects an explicit no-animation value, matching what
  the garbage read did in practice.
- pause_menu_item_box_cursor: removed the three empty 'if (x2);' matching
  artifacts. The x/y/z spin state itself is untouched.
- func_800A1FB0: initialized var_s4/var_s5 against the guarded-but-warned
  switch default.
- func_800A54EC: initialized the pause cursor position pointer against its
  unreachable mode-switch default (would have been a null-deref class bug
  if mode values ever grew).

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* collision: define the G_ENDDL opcode shifts, return 0 when no tyre surface found

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* replays: cast staff-ghost pointer comparisons, return 0 for empty ghost buffer

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* update_objects: return 0 from conditional step helpers, fix TLUT pointer comparison, init train draw distances

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* render_player: remove the impossible lamp range (vanilla bug, never glowed on N64 either)

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* math_util_2: return the vector, not the address of the parameter slot

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Crab: remove inner declaration shadowing the initialized objectIndex

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* editor: honor InverseMatrix failure (bool was compared against 2, always true)

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Track.h: drop dead null checks on array members

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* shells: drop always-true angle and surface gates (behavior unchanged, s16 made them tautological)

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* particles: make the no-return particle setters void

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* effects/stubs/skybox/main: align signatures with functions that return nothing

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* menus/save: type the pak status variable as s32, return BAD_READ on the fall-off path

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* audio: remove uninitialized-read matching artifacts, init isSound

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* code_80005FD0/code_80086E70: init dead bomb kart pointer and the no_init variable

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* port/engine UI: fix printf-style format types and non-literal format string

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* render_objects: feed the unused texture param to the uninitialized img walker

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Format the changed lines per .clang-format

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Address review: strip explanatory comments, drop vec3f pointer returns

Comments moved to the PR record; the one flagged worth keeping stays.
vec3f_set_xyz/normalize/cross_product return void now since the out
argument is the interface and nothing used the pointer.

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Update actors_extended.c

* Add checkbox for 'Shells Shoot Straight' option

* Update render_player.c

* Update PortMenu.cpp

* Update render_player.c

* Update actors_extended.c

* Update render_player.c

---------

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* menu_items: fix the real-bug-tier compiler warnings (#691 batch 1)

Six fixes, all in menu_items.c, all verified by menu playtest on macOS:

- BSWAP16(*color0++) advanced the pointer twice per pixel on little-endian
  builds (the macro evaluates its argument twice) and mixed bytes from two
  different pixels. Read first, increment separately. Note: this function
  (func_8009A9FC) currently has no callers, so the corruption was latent,
  not live.
- func_8009B9D0 fell off the end on a lookup miss, returning garbage (the
  existing code comment already suspected this). The miss path now returns
  the display list head unchanged, i.e. draws nothing.
- update_ok_menu_item read an uninitialized stack slot for unknown menu
  item types; now selects an explicit no-animation value, matching what
  the garbage read did in practice.
- pause_menu_item_box_cursor: removed the three empty 'if (x2);' matching
  artifacts. The x/y/z spin state itself is untouched.
- func_800A1FB0: initialized var_s4/var_s5 against the guarded-but-warned
  switch default.
- func_800A54EC: initialized the pause cursor position pointer against its
  unreachable mode-switch default (would have been a null-deref class bug
  if mode values ever grew).

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* collision: define the G_ENDDL opcode shifts, return 0 when no tyre surface found

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* replays: cast staff-ghost pointer comparisons, return 0 for empty ghost buffer

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* update_objects: return 0 from conditional step helpers, fix TLUT pointer comparison, init train draw distances

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* render_player: remove the impossible lamp range (vanilla bug, never glowed on N64 either)

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* math_util_2: return the vector, not the address of the parameter slot

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Crab: remove inner declaration shadowing the initialized objectIndex

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* editor: honor InverseMatrix failure (bool was compared against 2, always true)

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Track.h: drop dead null checks on array members

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* shells: drop always-true angle and surface gates (behavior unchanged, s16 made them tautological)

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* particles: make the no-return particle setters void

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* effects/stubs/skybox/main: align signatures with functions that return nothing

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* menus/save: type the pak status variable as s32, return BAD_READ on the fall-off path

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* audio: remove uninitialized-read matching artifacts, init isSound

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* code_80005FD0/code_80086E70: init dead bomb kart pointer and the no_init variable

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* port/engine UI: fix printf-style format types and non-literal format string

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* render_objects: feed the unused texture param to the uninitialized img walker

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Format the changed lines per .clang-format

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Address review: strip explanatory comments, drop vec3f pointer returns

Comments moved to the PR record; the one flagged worth keeping stays.
vec3f_set_xyz/normalize/cross_product return void now since the out
argument is the interface and nothing used the pointer.

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Update actors_extended.c

* Add checkbox for 'Shells Shoot Straight' option

* Update render_player.c

* Update PortMenu.cpp

* Update render_player.c

* Update actors_extended.c

* Update render_player.c

---------

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>macOS: package the game as a proper SpaghettiKart.app bundle (#724)</title>
<updated>2026-08-02T01:35:51+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-08-02T01:35:51+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=aa6c9c10caf3630ca37cb45bafc0485b76605445'/>
<id>aa6c9c10caf3630ca37cb45bafc0485b76605445</id>
<content type='text'>
* macOS: package the game as a proper SpaghettiKart.app bundle

Turns the bare Spaghettify executable into a self-contained SpaghettiKart.app:

- cmake/macos/apple_bundle.cmake: bundle metadata (using the existing
  Info.plist, which already described a bundle), app icon, runtime resources
  (config.yml, yamls, meta, spaghetti.o2r) into Contents/Resources, dylib
  relinking for local Homebrew builds (no-op with static vcpkg deps), and
  ad-hoc codesigning so the app launches without 'damaged app' warnings
- App icon: on Xcode 26+ the Icon Composer package (SpaghettiKartIcon.icon,
  drawn from the existing HM ship-kart icon.png art) compiles with actool
  into an Assets.car, giving the native Liquid Glass look on macOS 26+, plus
  a flattened icns for older systems; older Xcode falls back to the previous
  sips/iconutil flat icon. Supersedes the CreateOSXIcons target.
- Game.cpp: default SHIP_HOME to ~/Library/Application Support/SpaghettiKart
  on macOS so first-run extraction, saves, config, and mods land there
  instead of scattering into the user's home folder; a user-set SHIP_HOME
  still wins
- Info.plist: NSHighResolutionCapable for proper Retina rendering
- CI: mac jobs now ship the .app (ditto zip so the executable bit survives
  artifact upload)
- the yamls-next-to-exe copy is scoped to non-Darwin; on macOS those files
  live in Contents/Resources instead

* macOS: ship prebuilt icon artifacts instead of compiling them at build time

Commits the compiled Assets.car (Liquid Glass icon for macOS 26+) and the
flattened SpaghettiKart.icns, both built from the in-repo Icon Composer
package. The build now just copies them into the bundle, which removes the
actool machinery and the Xcode 26 requirement entirely: every build on any
machine, including CI, gets the real icon.

Also adds a staleness guard: the artifacts carry a hash stamp of the
icon.png they were made from, and if icon.png ever changes without them
being regenerated, the build automatically generates a flat icns from the
current icon.png instead, so updating the logo never requires any macOS
work and an outdated icon can never ship. Regeneration steps are documented
in apple_bundle.cmake.

* Rename the icon layer to ship-kart-layer.png

Group 2.png was the design tool's default layer name; name it for what it
is. Verified the Icon Composer package still compiles with actool after the
rename; the committed Assets.car is unaffected (compiled, self-contained).

---------

Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* macOS: package the game as a proper SpaghettiKart.app bundle

Turns the bare Spaghettify executable into a self-contained SpaghettiKart.app:

- cmake/macos/apple_bundle.cmake: bundle metadata (using the existing
  Info.plist, which already described a bundle), app icon, runtime resources
  (config.yml, yamls, meta, spaghetti.o2r) into Contents/Resources, dylib
  relinking for local Homebrew builds (no-op with static vcpkg deps), and
  ad-hoc codesigning so the app launches without 'damaged app' warnings
- App icon: on Xcode 26+ the Icon Composer package (SpaghettiKartIcon.icon,
  drawn from the existing HM ship-kart icon.png art) compiles with actool
  into an Assets.car, giving the native Liquid Glass look on macOS 26+, plus
  a flattened icns for older systems; older Xcode falls back to the previous
  sips/iconutil flat icon. Supersedes the CreateOSXIcons target.
- Game.cpp: default SHIP_HOME to ~/Library/Application Support/SpaghettiKart
  on macOS so first-run extraction, saves, config, and mods land there
  instead of scattering into the user's home folder; a user-set SHIP_HOME
  still wins
- Info.plist: NSHighResolutionCapable for proper Retina rendering
- CI: mac jobs now ship the .app (ditto zip so the executable bit survives
  artifact upload)
- the yamls-next-to-exe copy is scoped to non-Darwin; on macOS those files
  live in Contents/Resources instead

* macOS: ship prebuilt icon artifacts instead of compiling them at build time

Commits the compiled Assets.car (Liquid Glass icon for macOS 26+) and the
flattened SpaghettiKart.icns, both built from the in-repo Icon Composer
package. The build now just copies them into the bundle, which removes the
actool machinery and the Xcode 26 requirement entirely: every build on any
machine, including CI, gets the real icon.

Also adds a staleness guard: the artifacts carry a hash stamp of the
icon.png they were made from, and if icon.png ever changes without them
being regenerated, the build automatically generates a flat icns from the
current icon.png instead, so updating the logo never requires any macOS
work and an outdated icon can never ship. Regeneration steps are documented
in apple_bundle.cmake.

* Rename the icon layer to ship-kart-layer.png

Group 2.png was the design tool's default layer name; name it for what it
is. Verified the Icon Composer package still compiles with actool after the
rename; the committed Assets.car is unaffected (compiled, self-contained).

---------

Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Update SkySnow.cpp (#728)</title>
<updated>2026-07-29T05:05:44+00:00</updated>
<author>
<name>MegaMech</name>
<email>MegaMech@users.noreply.github.com</email>
</author>
<published>2026-07-29T05:05:44+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=b1eefb7361e5cd5196ac59f5b73218285705150c'/>
<id>b1eefb7361e5cd5196ac59f5b73218285705150c</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>Fix garbage frame flash on cinematic camera cuts (#720)</title>
<updated>2026-07-27T05:00:47+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-07-27T05:00:47+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=5b28472d477bab101dee2a0f469fe2aee2c58a01'/>
<id>5b28472d477bab101dee2a0f469fe2aee2c58a01</id>
<content type='text'>
* Fix garbage frame flash on cinematic camera cuts

Frame interpolation blended matrices across instant camera teleports in
attract/demo and post-race cameras, flashing 1-2 frames of sheared
geometry. Complete the existing camera-epoch mechanism: flag a cut in
func_80019890 when the camera moves &gt;100 units, and have the
interpolator snap that frame to the new view instead of blending.

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Retrigger CI (transient 502 downloading libogg from gitlab.xiph.org)

* Name the cinematic shot dispatcher

Review feedback on #720: func_80019890 -&gt; camera_start_cinematic_shot.
It starts whichever shot D_80164680 has selected for a camera, dispatching
to the per-shot setup that teleports the camera to the shot's opening
position, which is why the cut detection lives there.

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Apply fix to freecam

Added FrameInterpolation_DontInterpolateCamera calls to improve camera behavior when changing target players.

---------

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* Fix garbage frame flash on cinematic camera cuts

Frame interpolation blended matrices across instant camera teleports in
attract/demo and post-race cameras, flashing 1-2 frames of sheared
geometry. Complete the existing camera-epoch mechanism: flag a cut in
func_80019890 when the camera moves &gt;100 units, and have the
interpolator snap that frame to the new view instead of blending.

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Retrigger CI (transient 502 downloading libogg from gitlab.xiph.org)

* Name the cinematic shot dispatcher

Review feedback on #720: func_80019890 -&gt; camera_start_cinematic_shot.
It starts whichever shot D_80164680 has selected for a camera, dispatching
to the per-shot setup that teleports the camera to the shot's opening
position, which is why the cut detection lives there.

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

* Apply fix to freecam

Added FrameInterpolation_DontInterpolateCamera calls to improve camera behavior when changing target players.

---------

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Sky: paint the full window width (#717)</title>
<updated>2026-07-27T04:29:12+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-07-27T04:29:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=24ac81951ea9b40a431b07a3c6395beab636cd97'/>
<id>24ac81951ea9b40a431b07a3c6395beab636cd97</id>
<content type='text'>
The gradient quad's projection path does not preserve the outward
rounding of the Rect edge getters, which can leave the last window
column unpainted at some aspect ratios (visible as a black strip at the
screen edge that only sky sprites drew into). Overshoot the quad a few
virtual pixels past both edges; the excess clips at the window.

Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The gradient quad's projection path does not preserve the outward
rounding of the Rect edge getters, which can leave the last window
column unpainted at some aspect ratios (visible as a black strip at the
screen edge that only sky sprites drew into). Overshoot the quad a few
virtual pixels past both edges; the excess clips at the window.

Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Don't let the attract timer hijack an in-flight menu fade (#723)</title>
<updated>2026-07-27T04:18:15+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-07-27T04:18:15+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=9ffdb8342375c5642bed38d30cae4a209df70fe5'/>
<id>9ffdb8342375c5642bed38d30cae4a209df70fe5</id>
<content type='text'>
Pressing start on the splash screen in the last moments before the demo
timeout could play the enter chime and then drop into the demo reel with
no music: the attract timer (gMenuTimingCounter &gt;= 0x12D) fired while the
press-start fade was in flight and overwrote gMenuFadeType to the demo
value, so the fade completed as a demo entry. Vanilla bug; roughly a
0.7-second window every attract cycle.

Only arm the demo when no fade is in flight. A natural timeout (idle, no
fade running) behaves exactly as before.

Play-verified: start pressed repeatedly around the timeout boundary always
lands in the main menu with music; the demo reel still plays when idle.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Pressing start on the splash screen in the last moments before the demo
timeout could play the enter chime and then drop into the demo reel with
no music: the attract timer (gMenuTimingCounter &gt;= 0x12D) fired while the
press-start fade was in flight and overwrote gMenuFadeType to the demo
value, so the fade completed as a demo entry. Vanilla bug; roughly a
0.7-second window every attract cycle.

Only arm the demo when no fade is in flight. A natural timeout (idle, no
fade running) behaves exactly as before.

Play-verified: start pressed repeatedly around the timeout boundary always
lands in the main menu with music; the demo reel still plays when idle.</pre>
</div>
</content>
</entry>
<entry>
<title>Exit cleanly after teardown instead of racing static destructors (#718)</title>
<updated>2026-07-23T18:23:53+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-07-23T18:23:53+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=322cd52c167842f22d818066ef18ce7b25439075'/>
<id>322cd52c167842f22d818066ef18ce7b25439075</id>
<content type='text'>
Normal quits ran the full static-destructor chain, where destruction order is
unspecified and something logs through spdlog after its statics are destroyed,
intermittently segfaulting on exit (same class as upstream issue #689). All
user-visible state (config, saves) is already persisted by GameEngine::Destroy,
so end the process with _Exit like the extraction error paths already do.

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Normal quits ran the full static-destructor chain, where destruction order is
unspecified and something logs through spdlog after its statics are destroyed,
intermittently segfaulting on exit (same class as upstream issue #689). All
user-visible state (config, saves) is already persisted by GameEngine::Destroy,
so end the process with _Exit like the extraction error paths already do.

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Fix the reset button doing nothing or looping the start screen when pressed repeatedly (#719)</title>
<updated>2026-07-23T18:21:36+00:00</updated>
<author>
<name>quarrel07</name>
<email>178681861+quarrel07@users.noreply.github.com</email>
</author>
<published>2026-07-23T18:21:36+00:00</published>
<link rel='alternate' type='text/html' href='https://git.dog6.net/SpaghettiKart/commit/?id=78a012ce06b8756870e18230c3957c00cc2a66da'/>
<id>78a012ce06b8756870e18230c3957c00cc2a66da</id>
<content type='text'>
* Make the reset button reliable from any screen

Pressing reset repeatedly near app start could do nothing, fade the music
with no reset, or bounce the press-start screen back to itself. Three
defects in the old handler:

- It ran from the ImGui widget, racing the menu state machine; a press
  landing mid-fade was re-advanced by the in-flight transition (the
  press-start loop).
- It always wrote gGamestateNext = MAIN_MENU_FROM_QUIT; once the game was
  already in that state a repeat write is swallowed by the != guard in
  main.c, so nothing happened.
- CM_ResetAudio ran unconditionally, so the music faded even when the
  reset was swallowed.

The widget now only sets an atomic request flag; ApplyPendingReset (top of
push_frame, on the game loop) performs the reset. It alternates between
the two identical FROM_QUIT gamestates so every press trips the gamestate
switch, and re-enters the menus through the intro's own transition
protocol (FADE_MODE_LOGO -&gt; setup_menus rebuild + fresh fade-in), which
replaces any in-flight transition. Audio only fades when a reset actually
executes. The dead gSkipIntro switch in the old handler (an unconditional
override below it always won) is gone; behavior is unchanged: reset lands
on the logo intro, or the start menu in debug mode.

Play-verified: hammering reset during the splash screens restarts the
logo every press; reset mid-fade after press-start no longer loops back;
resets from the main menu and mid-race work as before.

* Honor gSkipIntro when picking the post-reset screen

Review feedback on #719: restore the gSkipIntro switch from the old
handler instead of hard-coding the logo intro, keeping the debug-mode
override on top (no else), matching the boot-time logic in main().

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

---------

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* Make the reset button reliable from any screen

Pressing reset repeatedly near app start could do nothing, fade the music
with no reset, or bounce the press-start screen back to itself. Three
defects in the old handler:

- It ran from the ImGui widget, racing the menu state machine; a press
  landing mid-fade was re-advanced by the in-flight transition (the
  press-start loop).
- It always wrote gGamestateNext = MAIN_MENU_FROM_QUIT; once the game was
  already in that state a repeat write is swallowed by the != guard in
  main.c, so nothing happened.
- CM_ResetAudio ran unconditionally, so the music faded even when the
  reset was swallowed.

The widget now only sets an atomic request flag; ApplyPendingReset (top of
push_frame, on the game loop) performs the reset. It alternates between
the two identical FROM_QUIT gamestates so every press trips the gamestate
switch, and re-enters the menus through the intro's own transition
protocol (FADE_MODE_LOGO -&gt; setup_menus rebuild + fresh fade-in), which
replaces any in-flight transition. Audio only fades when a reset actually
executes. The dead gSkipIntro switch in the old handler (an unconditional
override below it always won) is gone; behavior is unchanged: reset lands
on the logo intro, or the start menu in debug mode.

Play-verified: hammering reset during the splash screens restarts the
logo every press; reset mid-fade after press-start no longer loops back;
resets from the main menu and mid-race work as before.

* Honor gSkipIntro when picking the post-reset screen

Review feedback on #719: restore the gSkipIntro switch from the old
handler instead of hard-coding the logo intro, keeping the debug-mode
override on top (no else), matching the boot-time logic in main().

Co-Authored-By: Claude Fable 5 &lt;noreply@anthropic.com&gt;

---------

Co-authored-by: Claude Fable 5 &lt;noreply@anthropic.com&gt;
Co-authored-by: MegaMech &lt;MegaMech@users.noreply.github.com&gt;</pre>
</div>
</content>
</entry>
</feed>
