2011-06-13 18:48:36 +02:00
|
|
|
/*
|
2019-08-03 19:43:48 +02:00
|
|
|
* This software is licensed under the terms of the MIT License.
|
2017-02-11 04:52:08 +01:00
|
|
|
* See COPYING for further information.
|
2011-06-13 18:48:36 +02:00
|
|
|
* ---
|
2024-05-16 23:30:41 +02:00
|
|
|
* Copyright (c) 2011-2024, Lukas Weber <laochailan@web.de>.
|
|
|
|
* Copyright (c) 2012-2024, Andrei Alexeyev <akari@taisei-project.org>.
|
2011-06-13 18:48:36 +02:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include "menu.h"
|
2024-05-17 04:41:28 +02:00
|
|
|
|
|
|
|
#include "audio/audio.h"
|
2023-04-07 05:27:56 +02:00
|
|
|
#include "eventloop/eventloop.h"
|
2024-05-17 04:41:28 +02:00
|
|
|
#include "events.h"
|
|
|
|
#include "global.h"
|
2023-06-08 01:45:48 +02:00
|
|
|
#include "replay/demoplayer.h"
|
2023-07-30 06:13:09 +02:00
|
|
|
#include "util/graphics.h"
|
2024-05-17 04:41:28 +02:00
|
|
|
#include "video.h"
|
2024-05-04 12:47:00 +02:00
|
|
|
#include "watchdog.h"
|
2011-06-13 18:48:36 +02:00
|
|
|
|
2018-05-08 10:21:01 +02:00
|
|
|
MenuEntry *add_menu_entry(MenuData *menu, const char *name, MenuAction action, void *arg) {
|
2023-09-23 13:49:26 +02:00
|
|
|
return dynarray_append(&menu->entries, {
|
|
|
|
.action = action,
|
|
|
|
.arg = arg,
|
|
|
|
.transition = menu->transition,
|
2024-09-05 06:38:58 +02:00
|
|
|
.name = name ? mem_strdup(name) : NULL,
|
2023-09-23 13:49:26 +02:00
|
|
|
});
|
2011-06-13 18:48:36 +02:00
|
|
|
}
|
|
|
|
|
2012-07-14 10:40:21 +02:00
|
|
|
void add_menu_separator(MenuData *menu) {
|
2023-09-23 13:49:26 +02:00
|
|
|
dynarray_append(&menu->entries, {});
|
2012-07-14 10:40:21 +02:00
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
void free_menu(MenuData *menu) {
|
|
|
|
if(menu == NULL) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2020-04-05 04:51:00 +02:00
|
|
|
dynarray_foreach_elem(&menu->entries, MenuEntry *e, {
|
2023-01-09 04:19:31 +01:00
|
|
|
mem_free(e->name);
|
2020-04-05 04:51:00 +02:00
|
|
|
});
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2020-04-05 04:51:00 +02:00
|
|
|
dynarray_free_data(&menu->entries);
|
2023-01-09 04:19:31 +01:00
|
|
|
mem_free(menu);
|
2011-06-13 18:48:36 +02:00
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
MenuData* alloc_menu(void) {
|
2023-01-09 04:19:31 +01:00
|
|
|
return ALLOC(MenuData, {
|
|
|
|
.selected = -1,
|
|
|
|
.transition = TransFadeBlack,
|
|
|
|
.transition_in_time = FADE_TIME,
|
|
|
|
.transition_out_time = FADE_TIME,
|
|
|
|
.fade = 1.0,
|
|
|
|
.input = menu_input,
|
|
|
|
});
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
void kill_menu(MenuData *menu) {
|
|
|
|
if(menu != NULL) {
|
|
|
|
menu->state = MS_Dead;
|
|
|
|
// nani?!
|
|
|
|
}
|
2017-02-24 22:58:27 +01:00
|
|
|
}
|
|
|
|
|
2023-04-07 05:47:47 +02:00
|
|
|
static void close_menu_finish(CallChainResult ccr) {
|
|
|
|
MenuData *menu = ccr.ctx;
|
2023-03-22 22:38:52 +01:00
|
|
|
|
2023-04-12 06:46:08 +02:00
|
|
|
if(TRANSITION_RESULT_CANCELED(ccr)) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
// This may happen with MF_AlwaysProcessInput menus, so make absolutely sure we
|
|
|
|
// never run the call chain with menu->state == MS_Dead more than once.
|
|
|
|
bool was_dead = (menu->state == MS_Dead);
|
|
|
|
|
2017-02-24 22:58:27 +01:00
|
|
|
menu->state = MS_Dead;
|
|
|
|
|
2020-04-05 04:51:00 +02:00
|
|
|
if(menu->selected != -1) {
|
|
|
|
MenuEntry *e = dynarray_get_ptr(&menu->entries, menu->selected);
|
2017-02-24 22:58:27 +01:00
|
|
|
|
2020-04-05 04:51:00 +02:00
|
|
|
if(e->action != NULL) {
|
|
|
|
if(!(menu->flags & MF_Transient)) {
|
|
|
|
menu->state = MS_Normal;
|
|
|
|
}
|
|
|
|
|
|
|
|
e->action(menu, e->arg);
|
|
|
|
}
|
2017-02-24 22:58:27 +01:00
|
|
|
}
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
|
|
|
|
if(!was_dead) {
|
|
|
|
run_call_chain(&menu->cc, menu);
|
|
|
|
}
|
2011-06-13 18:48:36 +02:00
|
|
|
}
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2012-08-14 17:56:53 +02:00
|
|
|
void close_menu(MenuData *menu) {
|
|
|
|
TransitionRule trans = menu->transition;
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-02-24 22:58:27 +01:00
|
|
|
assert(menu->state != MS_Dead);
|
|
|
|
menu->state = MS_FadeOut;
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-02-24 22:58:27 +01:00
|
|
|
if(menu->selected != -1) {
|
2020-04-05 04:51:00 +02:00
|
|
|
trans = dynarray_get(&menu->entries, menu->selected).transition;
|
2017-02-24 22:58:27 +01:00
|
|
|
}
|
2012-08-12 16:54:48 +02:00
|
|
|
|
2023-04-07 05:47:47 +02:00
|
|
|
CallChain cc = CALLCHAIN(close_menu_finish, menu);
|
|
|
|
|
2017-02-24 22:58:27 +01:00
|
|
|
if(trans) {
|
2023-04-07 05:47:47 +02:00
|
|
|
set_transition(
|
2017-02-24 22:58:27 +01:00
|
|
|
trans,
|
|
|
|
menu->transition_in_time,
|
|
|
|
menu->transition_out_time,
|
2023-04-07 05:47:47 +02:00
|
|
|
cc
|
2017-02-24 22:58:27 +01:00
|
|
|
);
|
|
|
|
} else {
|
2023-04-07 05:47:47 +02:00
|
|
|
run_call_chain(&cc, NULL);
|
2017-02-24 22:58:27 +01:00
|
|
|
}
|
2012-08-12 16:54:48 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
float menu_fade(MenuData *menu) {
|
2017-02-24 22:58:27 +01:00
|
|
|
return transition.fade;
|
2012-08-12 16:54:48 +02:00
|
|
|
}
|
|
|
|
|
2017-09-29 21:03:49 +02:00
|
|
|
bool menu_input_handler(SDL_Event *event, void *arg) {
|
2012-08-13 17:50:28 +02:00
|
|
|
MenuData *menu = arg;
|
2017-09-29 21:03:49 +02:00
|
|
|
TaiseiEvent te = TAISEI_EVENT(event->type);
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-09-29 21:03:49 +02:00
|
|
|
switch(te) {
|
2022-01-18 03:21:20 +01:00
|
|
|
case TE_MENU_CURSOR_DOWN: {
|
2020-06-22 16:41:03 +02:00
|
|
|
play_sfx_ui("generic_shot");
|
2022-01-18 03:21:20 +01:00
|
|
|
int orig_cursor = menu->cursor;
|
2012-08-13 17:50:28 +02:00
|
|
|
do {
|
2020-04-05 04:51:00 +02:00
|
|
|
if(++menu->cursor >= menu->entries.num_elements) {
|
2012-08-13 17:50:28 +02:00
|
|
|
menu->cursor = 0;
|
2020-04-05 04:51:00 +02:00
|
|
|
}
|
2022-01-18 03:21:20 +01:00
|
|
|
|
|
|
|
if(menu->cursor == orig_cursor) {
|
|
|
|
break;
|
|
|
|
}
|
2020-04-05 04:51:00 +02:00
|
|
|
} while(dynarray_get(&menu->entries, menu->cursor).action == NULL);
|
2017-12-28 02:35:53 +01:00
|
|
|
|
|
|
|
return true;
|
2022-01-18 03:21:20 +01:00
|
|
|
}
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2022-01-18 03:21:20 +01:00
|
|
|
case TE_MENU_CURSOR_UP: {
|
2020-06-22 16:41:03 +02:00
|
|
|
play_sfx_ui("generic_shot");
|
2022-01-18 03:21:20 +01:00
|
|
|
int orig_cursor = menu->cursor;
|
2012-08-13 17:50:28 +02:00
|
|
|
do {
|
2020-04-05 04:51:00 +02:00
|
|
|
if(--menu->cursor < 0) {
|
|
|
|
menu->cursor = menu->entries.num_elements - 1;
|
|
|
|
}
|
2022-01-18 03:21:20 +01:00
|
|
|
|
|
|
|
if(menu->cursor == orig_cursor) {
|
|
|
|
break;
|
|
|
|
}
|
2020-04-05 04:51:00 +02:00
|
|
|
} while(dynarray_get(&menu->entries, menu->cursor).action == NULL);
|
2017-12-28 02:35:53 +01:00
|
|
|
|
|
|
|
return true;
|
2022-01-18 03:21:20 +01:00
|
|
|
}
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-09-29 21:03:49 +02:00
|
|
|
case TE_MENU_ACCEPT:
|
2020-06-22 16:41:03 +02:00
|
|
|
play_sfx_ui("shot_special1");
|
2020-04-05 04:51:00 +02:00
|
|
|
if(dynarray_get(&menu->entries, menu->cursor).action) {
|
2012-08-13 17:50:28 +02:00
|
|
|
menu->selected = menu->cursor;
|
|
|
|
close_menu(menu);
|
|
|
|
}
|
2017-12-28 02:35:53 +01:00
|
|
|
|
|
|
|
return true;
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-09-29 21:03:49 +02:00
|
|
|
case TE_MENU_ABORT:
|
2020-06-22 16:41:03 +02:00
|
|
|
play_sfx_ui("hit");
|
2012-08-13 17:50:28 +02:00
|
|
|
if(menu->flags & MF_Abortable) {
|
|
|
|
menu->selected = -1;
|
|
|
|
close_menu(menu);
|
|
|
|
}
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-12-28 02:35:53 +01:00
|
|
|
return true;
|
2017-09-29 21:03:49 +02:00
|
|
|
|
2017-12-28 02:35:53 +01:00
|
|
|
default:
|
|
|
|
return false;
|
|
|
|
}
|
2012-01-07 13:20:16 +01:00
|
|
|
}
|
|
|
|
|
2011-06-13 18:48:36 +02:00
|
|
|
void menu_input(MenuData *menu) {
|
2017-09-29 21:03:49 +02:00
|
|
|
events_poll((EventHandler[]){
|
|
|
|
{ .proc = menu_input_handler, .arg = menu },
|
2017-12-26 12:07:40 +01:00
|
|
|
{ NULL }
|
2017-09-29 21:03:49 +02:00
|
|
|
}, EFLAG_MENU);
|
2011-06-13 18:48:36 +02:00
|
|
|
}
|
|
|
|
|
2017-02-24 22:58:27 +01:00
|
|
|
void menu_no_input(MenuData *menu) {
|
2017-09-29 21:03:49 +02:00
|
|
|
events_poll(NULL, 0);
|
2017-02-24 22:58:27 +01:00
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
static LogicFrameAction menu_logic_frame(void *arg) {
|
2017-10-04 07:07:04 +02:00
|
|
|
MenuData *menu = arg;
|
2017-02-11 04:52:08 +01:00
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
if(menu->state == MS_Dead) {
|
|
|
|
return LFRAME_STOP;
|
|
|
|
}
|
|
|
|
|
2017-12-26 12:07:40 +01:00
|
|
|
if(menu->logic) {
|
|
|
|
menu->logic(menu);
|
|
|
|
}
|
|
|
|
|
|
|
|
menu->frames++;
|
2017-10-04 07:07:04 +02:00
|
|
|
|
2024-05-04 12:47:00 +02:00
|
|
|
if(watchdog_signaled()) {
|
|
|
|
menu->selected = -1;
|
|
|
|
close_menu(menu);
|
|
|
|
}
|
|
|
|
|
2017-10-04 07:07:04 +02:00
|
|
|
if(menu->state != MS_FadeOut || menu->flags & MF_AlwaysProcessInput) {
|
|
|
|
assert(menu->input);
|
|
|
|
menu->input(menu);
|
|
|
|
} else {
|
|
|
|
menu_no_input(menu);
|
2011-06-13 18:48:36 +02:00
|
|
|
}
|
|
|
|
|
2017-12-26 09:56:21 +01:00
|
|
|
update_transition();
|
2017-12-26 12:07:40 +01:00
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
return LFRAME_WAIT;
|
2017-12-26 12:07:40 +01:00
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
static RenderFrameAction menu_render_frame(void *arg) {
|
2017-12-26 12:07:40 +01:00
|
|
|
MenuData *menu = arg;
|
2017-10-04 07:07:04 +02:00
|
|
|
assert(menu->draw);
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
set_ortho(SCREEN_W, SCREEN_H);
|
2017-10-04 07:07:04 +02:00
|
|
|
menu->draw(menu);
|
2017-12-26 09:56:21 +01:00
|
|
|
draw_transition();
|
2017-12-26 12:07:40 +01:00
|
|
|
return RFRAME_SWAP;
|
2017-10-04 07:07:04 +02:00
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
static void menu_end_loop(void *ctx) {
|
|
|
|
MenuData *menu = ctx;
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2023-06-08 01:45:48 +02:00
|
|
|
if(menu->flags & MF_NoDemo) {
|
|
|
|
demoplayer_resume();
|
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
if(menu->state != MS_Dead) {
|
|
|
|
// definitely dead now...
|
|
|
|
menu->state = MS_Dead;
|
|
|
|
run_call_chain(&menu->cc, menu);
|
2011-06-25 12:41:40 +02:00
|
|
|
}
|
2017-02-11 04:52:08 +01:00
|
|
|
|
2017-02-24 22:58:27 +01:00
|
|
|
if(menu->end) {
|
|
|
|
menu->end(menu);
|
|
|
|
}
|
2017-02-11 04:52:08 +01:00
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
free_menu(menu);
|
|
|
|
}
|
|
|
|
|
|
|
|
void enter_menu(MenuData *menu, CallChain next) {
|
|
|
|
if(menu == NULL) {
|
|
|
|
run_call_chain(&next, NULL);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
menu->cc = next;
|
|
|
|
|
|
|
|
if(menu->begin != NULL) {
|
|
|
|
menu->begin(menu);
|
|
|
|
}
|
|
|
|
|
2023-06-08 01:45:48 +02:00
|
|
|
if(menu->flags & MF_NoDemo) {
|
|
|
|
demoplayer_suspend();
|
|
|
|
}
|
|
|
|
|
Emscripten compatibility (#161)
* Major refactoring of the main loop(s) and control flow (WIP)
run_at_fps() is gone 🦀
Instead of nested blocking event loops, there is now an eventloop API
that manages an explicit stack of scenes. This makes Taisei a lot more
portable to async environments where spinning a loop forever without
yielding control simply is not an option, and that is the entire point
of this change.
A prime example of such an environment is the Web (via emscripten).
Taisei was able to run there through a terrible hack: inserting
emscripten_sleep calls into the loop, which would yield to the browser.
This has several major drawbacks: first of all, every function that
could possibly call emscripten_sleep must be compiled into a special
kind of bytecode, which then has to be interpreted at runtime, *much*
slower than JITed WebAssembly. And that includes *everything* down the
call stack, too! For more information, see
https://emscripten.org/docs/porting/emterpreter.html
Even though that method worked well enough for experimenting, despite
suboptimal performance, there is another obvious drawback:
emscripten_sleep is implemented via setTimeout(), which can be very
imprecise and is generally not reliable for fluid animation. Browsers
actually have an API specifically for that use case:
window.requestAnimationFrame(), but Taisei's original blocking control
flow style is simply not compatible with it. Emscripten exposes this API
with its emscripten_set_main_loop(), which the eventloop backend now
uses on that platform.
Unfortunately, C is still C, with no fancy closures or coroutines.
With blocking calls into menu/scene loops gone, the control flow is
reimplemented via so-called (pun intended) "call chains". That is
basically an euphemism for callback hell. With manual memory management
and zero type-safety. Not that the menu system wasn't shitty enough
already. I'll just keep telling myself that this is all temporary and
will be replaced with scripts in v1.4.
* improve build system for emscripten + various fixes
* squish menu bugs
* improve emscripten event loop; disable EMULATE_FUNCTION_POINTER_CASTS
Note that stock freetype does not work without
EMULATE_FUNCTION_POINTER_CASTS; use a patched version from the
"emscripten" branch here:
https://github.com/taisei-project/freetype2/tree/emscripten
* Enable -Wcast-function-type
Calling functions through incompatible pointers is nasal demons and
doesn't work in WASM.
* webgl: workaround a crash on some browsers
* emscripten improvements:
* Persist state (config, progress, replays, ...) in local IndexDB
* Simpler HTML shell (temporary)
* Enable more optimizations
* fix build if validate_glsl=false
* emscripten: improve asset packaging, with local cache
Note that even though there are rules to build audio bundles, audio
does *not* work yet. It looks like SDL2_mixer can not work without
threads, which is a problem. Yet another reason to write an OpenAL
backend - emscripten supports that natively.
* emscripten: customize the html shell
* emscripten: force "show log" checkbox unchecked initially
* emscripten: remove quit shortcut from main menu (since there's no quit)
* emscripten: log area fixes
* emscripten/webgl: workaround for fullscreen viewport issue
* emscripten: implement frameskip
* emscripter: improve framerate limiter
* align List to at least 8 bytes (shut up warnings)
* fix non-emscripten builds
* improve fullscreen handling, mainly for emscripten
* Workaround to make audio work in chromium
emscripten-core/emscripten#6511
* emscripten: better vsync handling; enable vsync & disable fxaa by default
2019-03-09 20:32:32 +01:00
|
|
|
eventloop_enter(menu, menu_logic_frame, menu_render_frame, menu_end_loop, FPS);
|
2011-06-25 12:41:40 +02:00
|
|
|
}
|