qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-11-26 22:31:05 +00:00
Code Issues Projects Releases Packages Wiki Activity
quakeforge/libs/client/cl_view.c

1002 lines
26 KiB
C
Raw Normal View History

initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
/*
Clean up chase/spectator camera stuff. o Rename nq's cl_cam.c to cl_chase.c. o Split out the chase cam stuff from qw's cl_cam.c to cl_chase.c
2012-05-20 10:17:41 +00:00
cl_view.c
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
player eye positioning
Copyright (C) 1996-1997 Id Software, Inc.
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to:
Free Software Foundation, Inc.
59 Temple Place - Suite 330
Boston, MA 02111-1307, USA
*/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
move the api headers into include/QF and clean up (most of) the resulting mess. target specific files that I don't build won't compile yet. just put QF/ infront of the offending headers. Also move ver_check into libqfutils
2001-03-27 20:33:07 +00:00
#include "QF/cmd.h"
#include "QF/cvar.h"
#include "QF/msg.h"
vid libs build but probably don't work yet (I KNOW x11 doesn't) lots of build failures in the main source still
2001-04-15 08:04:15 +00:00
#include "QF/screen.h"
Merges and cleanups...
2001-05-13 22:57:27 +00:00
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
#include "QF/plugin/vid_render.h"
[scene] Move entity_t etc into scene headers I meant to do this a while ago but forgot about it. Things are a bit of a mess in that the renderer knows too much about entities, but eventually the renderer will know about only things to render (meshes, particles, etc).
2022-03-29 05:40:48 +00:00
#include "QF/scene/entity.h"
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
#include "QF/scene/scene.h"
[client] Use vec4_t in entity_state_t And clean up the mess (sort of:P)
2021-03-11 07:19:49 +00:00
#include "QF/simd/vec4f.h"
Exit early on common case, rather than go through all checks. Minor speedup, but measurable on higher framerate demos.
2001-08-07 18:53:22 +00:00
#include "compat.h"
Move and clean up clview.h Redundant or dead prototypes deleted, and the client/view.h seems a good place for the file.
2021-03-12 02:48:53 +00:00
[client] Merge chase camera and much of input Handling of view angles is a little hacky at the moment, but this gets the chase camera code and most of the common input code into one place, which will make cleaning up the camera code much easier.
2022-02-22 06:23:09 +00:00
#include "client/chase.h"
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
#include "client/entities.h"
#include "client/hud.h"
#include "client/input.h"
#include "client/view.h"
[renderer] Support multiple entity queues While there's currently only the one still, this will allow the entities to be multiply queued for multi-pass rendering (eg, shadows). As the avoidance of putting an entity in the same queue more than once relies on the entity id, all entities now come from the scene (which is stored in cl_world in the client code for nq and qw), thus the extensive changes in the clients.
2022-03-04 16:48:10 +00:00
#include "client/world.h"
[build] Move to non-recursive make There's still some cleanup to do, but everything seems to be working nicely: `make -j` works, `make distcheck` passes. There is probably plenty of bitrot in the package directories (RPM, debian), though. The vc project files have been removed since those versions are way out of date and quakeforge is pretty much dependent on gcc now anyway. Most of the old Makefile.am files are now Makemodule.am. This should allow for new Makefile.am files that allow local building (to be added on an as-needed bases). The current remaining Makefile.am files are for standalone sub-projects.a The installable bins are currently built in the top-level build directory. This may change if the clutter gets to be too much. While this does make a noticeable difference in build times, the main reason for the switch was to take care of the growing dependency issues: now it's possible to build tools for code generation (eg, using qfcc and ruamoko programs for code-gen).
2020-06-21 14:15:17 +00:00
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
/*
white space
2001-05-30 20:55:28 +00:00
The view is allowed to move slightly from it's true position for bobbing,
but if it exceeds 8 pixels linear distance (spherical, not box), the list
of entities sent from the server may not include everything in the pvs,
especially when crossing a water boudnary.
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
*/
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float scr_ofsx;
static cvar_t scr_ofsx_cvar = {
.name = "scr_ofsx",
.description =
"None",
.default_value = "0",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &scr_ofsx },
};
float scr_ofsy;
static cvar_t scr_ofsy_cvar = {
.name = "scr_ofsy",
.description =
"None",
.default_value = "0",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &scr_ofsy },
};
float scr_ofsz;
static cvar_t scr_ofsz_cvar = {
.name = "scr_ofsz",
.description =
"None",
.default_value = "0",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &scr_ofsz },
};
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float cl_rollspeed;
static cvar_t cl_rollspeed_cvar = {
.name = "cl_rollspeed",
.description =
"How quickly you straighten out after strafing",
.default_value = "200",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &cl_rollspeed },
};
float cl_rollangle;
static cvar_t cl_rollangle_cvar = {
.name = "cl_rollangle",
.description =
"How much your screen tilts when strafing",
.default_value = "2.0",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &cl_rollangle },
};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float cl_bob;
static cvar_t cl_bob_cvar = {
.name = "cl_bob",
.description =
"How much your weapon moves up and down when walking",
.default_value = "0.02",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &cl_bob },
};
float cl_bobcycle;
static cvar_t cl_bobcycle_cvar = {
.name = "cl_bobcycle",
.description =
"How quickly your weapon moves up and down when walking",
.default_value = "0.6",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &cl_bobcycle },
};
float cl_bobup;
static cvar_t cl_bobup_cvar = {
.name = "cl_bobup",
.description =
"How long your weapon stays up before cycling when walking",
.default_value = "0.5",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &cl_bobup },
};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float v_centermove;
static cvar_t v_centermove_cvar = {
.name = "v_centermove",
.description =
"How far the player must move forward before the view re-centers",
.default_value = "0.15",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_centermove },
};
float v_centerspeed;
static cvar_t v_centerspeed_cvar = {
.name = "v_centerspeed",
.description =
"How quickly you return to a center view after a lookup or lookdown",
.default_value = "500",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_centerspeed },
};
Exit early on common case, rather than go through all checks. Minor speedup, but measurable on higher framerate demos.
2001-08-07 18:53:22 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float v_kicktime;
static cvar_t v_kicktime_cvar = {
.name = "v_kicktime",
.description =
"How long the kick from an attack lasts",
.default_value = "0.5",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_kicktime },
};
float v_kickroll;
static cvar_t v_kickroll_cvar = {
.name = "v_kickroll",
.description =
"How much you lean when hit",
.default_value = "0.6",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_kickroll },
};
float v_kickpitch;
static cvar_t v_kickpitch_cvar = {
.name = "v_kickpitch",
.description =
"How much you look up when hit",
.default_value = "0.6",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_kickpitch },
};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
int cl_cshift_bonus;
static cvar_t cl_cshift_bonus_cvar = {
.name = "cl_cshift_bonus",
.description =
"Show bonus flash on item pickup",
.default_value = "1",
.flags = CVAR_ARCHIVE,
.value = { .type = &cexpr_int, .value = &cl_cshift_bonus },
};
int cl_cshift_contents;
static cvar_t cl_cshift_contents_cvar = {
.name = "cl_cshift_content",
.description =
"Shift view colors for contents (water, slime, etc)",
.default_value = "1",
.flags = CVAR_ARCHIVE,
.value = { .type = &cexpr_int, .value = &cl_cshift_contents },
};
int cl_cshift_damage;
static cvar_t cl_cshift_damage_cvar = {
.name = "cl_cshift_damage",
.description =
"Shift view colors on damage",
.default_value = "1",
.flags = CVAR_ARCHIVE,
.value = { .type = &cexpr_int, .value = &cl_cshift_damage },
};
int cl_cshift_powerup;
static cvar_t cl_cshift_powerup_cvar = {
.name = "cl_cshift_powerup",
.description =
"Shift view colors for powerups",
.default_value = "1",
.flags = CVAR_ARCHIVE,
.value = { .type = &cexpr_int, .value = &cl_cshift_powerup },
};
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float v_iyaw_cycle;
static cvar_t v_iyaw_cycle_cvar = {
.name = "v_iyaw_cycle",
.description =
"How far you tilt right and left when v_idlescale is enabled",
.default_value = "2",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_iyaw_cycle },
};
float v_iroll_cycle;
static cvar_t v_iroll_cycle_cvar = {
.name = "v_iroll_cycle",
.description =
"How quickly you tilt right and left when v_idlescale is enabled",
.default_value = "0.5",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_iroll_cycle },
};
float v_ipitch_cycle;
static cvar_t v_ipitch_cycle_cvar = {
.name = "v_ipitch_cycle",
.description =
"How quickly you lean forwards and backwards when v_idlescale is "
"enabled",
.default_value = "1",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_ipitch_cycle },
};
float v_iyaw_level;
static cvar_t v_iyaw_level_cvar = {
.name = "v_iyaw_level",
.description =
"How far you tilt right and left when v_idlescale is enabled",
.default_value = "0.3",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_iyaw_level },
};
float v_iroll_level;
static cvar_t v_iroll_level_cvar = {
.name = "v_iroll_level",
.description =
"How far you tilt right and left when v_idlescale is enabled",
.default_value = "0.1",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_iroll_level },
};
float v_ipitch_level;
static cvar_t v_ipitch_level_cvar = {
.name = "v_ipitch_level",
.description =
"How far you lean forwards and backwards when v_idlescale is enabled",
.default_value = "0.3",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_ipitch_level },
};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float v_idlescale;
static cvar_t v_idlescale_cvar = {
.name = "v_idlescale",
.description =
"Toggles whether the view remains idle",
.default_value = "0",
.flags = CVAR_NONE,
.value = { .type = &cexpr_float, .value = &v_idlescale },
};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
diffsnr++
2001-05-20 00:34:40 +00:00
float v_dmg_time, v_dmg_roll, v_dmg_pitch;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t v_idle_yaw;
vec4f_t v_idle_roll;
vec4f_t v_idle_pitch;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
static cshift_t cshift_empty = { {130, 80, 50}, 0};
static cshift_t cshift_water = { {130, 80, 50}, 128};
static cshift_t cshift_slime = { { 0, 25, 5}, 150};
static cshift_t cshift_lava = { {255, 80, 0}, 150};
static cshift_t cshift_bonus = { {215, 186, 60}, 50};
static cshift_t armor_blood[] = {
{ {255, 0, 0} }, // blood
{ {220, 50, 50} }, // armor + blood
{ {200, 100, 100} }, // armor > blood need two for logic
{ {200, 100, 100} }, // armor > blood need two for logic
};
static cshift_t powerup[] = {
{ { 0, 0, 0}, 0},
{ {100, 100, 100}, 100}, // IT_INVISIBILITY
{ {255, 255, 0}, 30}, // IT_INVULNERABILITY
{ {255, 255, 0}, 30}, // IT_INVULNERABILITY
{ { 0, 255, 0}, 20}, // IT_SUIT
{ { 0, 255, 0}, 20}, // IT_SUIT
{ { 0, 255, 0}, 20}, // IT_SUIT
{ { 0, 255, 0}, 20}, // IT_SUIT
{ { 0, 0, 255}, 30}, // IT_QUAD
{ { 0, 0, 255}, 30}, // IT_QUAD
{ {255, 0, 255}, 30}, // IT_INVULNERABILITY | IT_QUAD
{ {255, 0, 255}, 30}, // IT_INVULNERABILITY | IT_QUAD
{ {255, 0, 255}, 30}, // IT_INVULNERABILITY | IT_QUAD
{ {255, 0, 255}, 30}, // IT_INVULNERABILITY | IT_QUAD
{ {255, 0, 255}, 30}, // IT_INVULNERABILITY | IT_QUAD
{ {255, 0, 255}, 30}, // IT_INVULNERABILITY | IT_QUAD
};
Merges and cleanups...
2001-05-13 22:57:27 +00:00
"Merge" V_CalcRoll.
2012-06-02 04:43:25 +00:00
#define sqr(x) ((x) * (x))
header/whitespace sweep continues.
2001-08-28 23:05:45 +00:00
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
float
Move and clean up clview.h Redundant or dead prototypes deleted, and the client/view.h seems a good place for the file.
2021-03-12 02:48:53 +00:00
V_CalcRoll (const vec3_t angles, vec4f_t velocity)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
Clean up some diff noise.
2012-05-21 13:00:58 +00:00
float side, sign, value;
vec3_t forward, right, up;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
AngleVectors (angles, forward, right, up);
side = DotProduct (velocity, right);
sign = side < 0 ? -1 : 1;
side = fabs (side);
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
value = cl_rollangle;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (side < cl_rollspeed)
side = side * value / cl_rollspeed;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
else
side = value;
return side * sign;
}
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static float
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcBob (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vec4f_t velocity = vs->velocity;
White space cleanups.
2011-06-19 10:04:49 +00:00
float cycle;
"Merge" V_CalcRoll.
2012-06-02 04:43:25 +00:00
static double bobtime;
White space cleanups.
2011-06-19 10:04:49 +00:00
static float bob;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (!vs->bob_enabled)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return 0;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->onground == -1)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return bob; // just use old value
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
bobtime += vs->frametime;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
cycle = bobtime - (int) (bobtime / cl_bobcycle) *
cl_bobcycle;
cycle /= cl_bobcycle;
if (cycle < cl_bobup)
cycle = cycle / cl_bobup;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
else
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
cycle = 1 + (cycle - cl_bobup) / (1.0 - cl_bobup);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
"Merge" V_CalcRoll.
2012-06-02 04:43:25 +00:00
// bob is proportional to velocity in the xy plane
sw32. Flee in terror. This'll break compilation for everything until -sdl32 compiles. sw32 wont work for a while either, so don't get overly excited yet.
2001-08-25 02:47:11 +00:00
// (don't count Z, or jumping messes it up)
[client] Use vec4_t in entity_state_t And clean up the mess (sort of:P)
2021-03-11 07:19:49 +00:00
velocity[2] = 0;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
bob = sqrt (dotf (velocity, velocity)[0]) * cl_bob;
"Merge" V_CalcRoll.
2012-06-02 04:43:25 +00:00
bob = bob * 0.3 + bob * 0.7 * sin (cycle * M_PI);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
if (bob > 4)
bob = 4;
else if (bob < -7)
bob = -7;
return bob;
}
void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_StartPitchDrift (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->laststop == vs->time) {
sw32. Flee in terror. This'll break compilation for everything until -sdl32 compiles. sw32 wont work for a while either, so don't get overly excited yet.
2001-08-25 02:47:11 +00:00
return; // something else is keeping it from drifting
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
sw32. Flee in terror. This'll break compilation for everything until -sdl32 compiles. sw32 wont work for a while either, so don't get overly excited yet.
2001-08-25 02:47:11 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->nodrift || !vs->pitchvel) {
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
vs->pitchvel = v_centerspeed;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->nodrift = false;
vs->driftmove = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
}
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
static void
V_StartPitchDrift_f (void *data)
{
V_StartPitchDrift (data);
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_StopPitchDrift (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->laststop = vs->time;
vs->nodrift = true;
vs->pitchvel = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
/*
V_DriftPitch
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
Moves the client pitch angle towards vs->idealpitch sent by the server.
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
If the user is adjusting pitch manually, either with lookup/lookdown,
mlook and mouse, or klook and keyboard, pitch drifting is constantly
stopped.
Drifting is enabled when the center view key is hit, mlook is released
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
and lookspring is non 0, or when
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
*/
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_DriftPitch (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
Diff reduction.
2011-06-18 13:22:47 +00:00
float delta, move;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
float forwardmove = vs->movecmd[0];
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (noclip_anglehack || vs->onground == -1 || !vs->drift_enabled) {
vs->driftmove = 0;
vs->pitchvel = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
}
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// don't count small mouse motion
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->nodrift) {
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (fabs (forwardmove) < cl_forwardspeed)
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->driftmove = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
else
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->driftmove += vs->frametime;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (vs->driftmove > v_centermove) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_StartPitchDrift (vs);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
return;
}
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
delta = vs->idealpitch - vs->player_angles[PITCH];
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
if (!delta) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->pitchvel = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
}
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
move = vs->frametime * vs->pitchvel;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
vs->pitchvel += vs->frametime * v_centerspeed;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
if (delta > 0) {
if (move > delta) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->pitchvel = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
move = delta;
}
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
vs->player_angles[PITCH] += move;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
} else if (delta < 0) {
if (move > -delta) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->pitchvel = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
move = -delta;
}
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
vs->player_angles[PITCH] -= move;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
}
header/whitespace sweep continues.
2001-08-28 23:05:45 +00:00
/* PALETTE FLASHES */
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
void
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
V_ParseDamage (qmsg_t *net_message, viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
White space cleanups.
2011-06-19 10:04:49 +00:00
float count, side;
int armor, blood;
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
vec4f_t origin = vs->player_origin;
vec_t *angles = vs->player_angles;
White space cleanups.
2011-06-19 10:04:49 +00:00
vec3_t from, forward, right, up;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
This prepares for moving the net_svc code to a branch allowing both network code cleanups and general performance work to be developed in relative peace. While cleaning up the networking code /is/ important, fixing QF's perfomance issues is of much higher priority.
2001-11-07 08:24:56 +00:00
armor = MSG_ReadByte (net_message);
blood = MSG_ReadByte (net_message);
Diff reduction.
2011-06-18 13:22:47 +00:00
MSG_ReadCoordV (net_message, from);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
This prepares for moving the net_svc code to a branch allowing both network code cleanups and general performance work to be developed in relative peace. While cleaning up the networking code /is/ important, fixing QF's perfomance issues is of much higher priority.
2001-11-07 08:24:56 +00:00
count = blood * 0.5 + armor * 0.5;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
if (count < 10)
count = 10;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (cl_cshift_damage
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
|| (vs->force_cshifts & INFO_CSHIFT_DAMAGE)) {
cshift_t *cshift = &vs->cshifts[CSHIFT_DAMAGE];
int percent = cshift->percent;
*cshift = armor_blood[(2 * (armor > blood)) + (armor > 0)];
cshift->percent = percent + 3 * count;
cshift->percent = bound (0, cshift->percent, 150);
Frame-rate independent cshift fades. As per mh's post on inside3d.com. I don't see any difference, but it still fades well, so... :) http://forums.inside3d.com/viewtopic.php?f=12&t=4899
2012-06-11 12:01:45 +00:00
cshift->initialpct = cshift->percent;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
cshift->time = vs->time;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// calculate view angle kicks
Yet more diff noise reduction.
2012-06-02 10:46:57 +00:00
VectorSubtract (from, origin, from);
This prepares for moving the net_svc code to a branch allowing both network code cleanups and general performance work to be developed in relative peace. While cleaning up the networking code /is/ important, fixing QF's perfomance issues is of much higher priority.
2001-11-07 08:24:56 +00:00
VectorNormalize (from);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Yet more diff noise reduction.
2012-06-02 10:46:57 +00:00
AngleVectors (angles, forward, right, up);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
This prepares for moving the net_svc code to a branch allowing both network code cleanups and general performance work to be developed in relative peace. While cleaning up the networking code /is/ important, fixing QF's perfomance issues is of much higher priority.
2001-11-07 08:24:56 +00:00
side = DotProduct (from, right);
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
v_dmg_roll = count * side * v_kickroll;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
This prepares for moving the net_svc code to a branch allowing both network code cleanups and general performance work to be developed in relative peace. While cleaning up the networking code /is/ important, fixing QF's perfomance issues is of much higher priority.
2001-11-07 08:24:56 +00:00
side = DotProduct (from, forward);
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
v_dmg_pitch = count * side * v_kickpitch;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
v_dmg_time = v_kicktime;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
V_cshift_f (void)
{
cshift_empty.destcolor[0] = atoi (Cmd_Argv (1));
cshift_empty.destcolor[1] = atoi (Cmd_Argv (2));
cshift_empty.destcolor[2] = atoi (Cmd_Argv (3));
cshift_empty.percent = atoi (Cmd_Argv (4));
}
/*
V_BonusFlash_f
When you run over an item, the server sends this command
*/
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_BonusFlash_f (void *data)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
viewstate_t *vs = data;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (!cl_cshift_bonus
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
&& !(vs->force_cshifts & INFO_CSHIFT_BONUS))
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->cshifts[CSHIFT_BONUS] = cshift_bonus;
vs->cshifts[CSHIFT_BONUS].initialpct = vs->cshifts[CSHIFT_BONUS].percent;
vs->cshifts[CSHIFT_BONUS].time = vs->time;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
/*
V_SetContentsColor
Underwater, lava, etc each has a color shift
*/
void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_SetContentsColor (viewstate_t *vs, int contents)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (!cl_cshift_contents
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
&& !(vs->force_cshifts & INFO_CSHIFT_CONTENTS)) {
vs->cshifts[CSHIFT_CONTENTS] = cshift_empty;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
}
switch (contents) {
case CONTENTS_EMPTY:
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->cshifts[CSHIFT_CONTENTS] = cshift_empty;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case CONTENTS_LAVA:
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->cshifts[CSHIFT_CONTENTS] = cshift_lava;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
case CONTENTS_SOLID:
case CONTENTS_SLIME:
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->cshifts[CSHIFT_CONTENTS] = cshift_slime;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
break;
default:
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->cshifts[CSHIFT_CONTENTS] = cshift_water;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
}
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcPowerupCshift (viewstate_t *vs)
Merge most of cshift calcs from sw and gl to r_view.c
2001-05-21 19:53:57 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->cshifts[CSHIFT_POWERUP] = powerup[vs->powerup_index];
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
}
/*
V_CalcBlend
LordHavoc made this a real, true alpha blend. Cleaned it up
a bit, but otherwise this is his code. --KB
*/
Move and clean up clview.h Redundant or dead prototypes deleted, and the client/view.h seems a good place for the file.
2021-03-12 02:48:53 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcBlend (viewstate_t *vs)
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
{
Diff reduction.
2011-06-18 13:22:47 +00:00
float a2, a3;
float r = 0, g = 0, b = 0, a = 0;
int i;
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
for (i = 0; i < NUM_CSHIFTS; i++) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
a2 = vs->cshifts[i].percent / 255.0;
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
if (!a2)
continue;
a2 = min (a2, 1.0);
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
r += (vs->cshifts[i].destcolor[0] - r) * a2;
g += (vs->cshifts[i].destcolor[1] - g) * a2;
b += (vs->cshifts[i].destcolor[2] - b) * a2;
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
a3 = (1.0 - a) * (1.0 - a2);
a = 1.0 - a3;
Merge most of cshift calcs from sw and gl to r_view.c
2001-05-21 19:53:57 +00:00
}
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
// LordHavoc: saturate color
if (a) {
a2 = 1.0 / a;
r *= a2;
g *= a2;
b *= a2;
}
[client] Move cshift info from viddef to viewstate It makes much more sense for cshift state to be in viewstate because it is very much an effect rather than anything specific to a renderer.
2022-02-28 07:55:12 +00:00
vs->cshift_color[0] = min (r, 255.0) / 255.0;
vs->cshift_color[1] = min (g, 255.0) / 255.0;
vs->cshift_color[2] = min (b, 255.0) / 255.0;
vs->cshift_color[3] = bound (0.0, a, 1.0);
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
}
Frame-rate independent cshift fades. As per mh's post on inside3d.com. I don't see any difference, but it still fades well, so... :) http://forums.inside3d.com/viewtopic.php?f=12&t=4899
2012-06-11 12:01:45 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_DropCShift (cshift_t *cs, double time, float droprate)
Frame-rate independent cshift fades. As per mh's post on inside3d.com. I don't see any difference, but it still fades well, so... :) http://forums.inside3d.com/viewtopic.php?f=12&t=4899
2012-06-11 12:01:45 +00:00
{
if (cs->time < 0) {
cs->percent = 0;
} else {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
cs->percent = cs->initialpct - (time - cs->time) * droprate;
Frame-rate independent cshift fades. As per mh's post on inside3d.com. I don't see any difference, but it still fades well, so... :) http://forums.inside3d.com/viewtopic.php?f=12&t=4899
2012-06-11 12:01:45 +00:00
if (cs->percent <= 0) {
cs->percent = 0;
cs->time = -1;
}
}
}
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_PrepBlend (viewstate_t *vs)
Massive whitespace cleanup. Lots of trailing whitespace and otherwise blank lines.
2012-05-21 23:23:22 +00:00
{
Diff reduction.
2011-06-18 13:22:47 +00:00
int i, j;
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (cl_cshift_powerup
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
|| (vs->force_cshifts & INFO_CSHIFT_POWERUP))
V_CalcPowerupCshift (vs);
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
[renderer] Clean up a lot of recalc_refdef use I think the widespread use of recalc_refdef (and force_fullscreen) was the result of a rushed merge of the renderer and video code (I do seem to remember sprinkling them around). This cleans the two out of the client code.
2022-03-18 15:56:30 +00:00
qboolean cshift_changed = false;
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
for (i = 0; i < NUM_CSHIFTS; i++) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->cshifts[i].percent != vs->prev_cshifts[i].percent) {
[renderer] Clean up a lot of recalc_refdef use I think the widespread use of recalc_refdef (and force_fullscreen) was the result of a rushed merge of the renderer and video code (I do seem to remember sprinkling them around). This cleans the two out of the client code.
2022-03-18 15:56:30 +00:00
cshift_changed = true;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->prev_cshifts[i].percent = vs->cshifts[i].percent;
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
}
for (j = 0; j < 3; j++) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->cshifts[i].destcolor[j] != vs->prev_cshifts[i].destcolor[j])
header/whitespace sweep continues.
2001-08-28 23:05:45 +00:00
{
[renderer] Clean up a lot of recalc_refdef use I think the widespread use of recalc_refdef (and force_fullscreen) was the result of a rushed merge of the renderer and video code (I do seem to remember sprinkling them around). This cleans the two out of the client code.
2022-03-18 15:56:30 +00:00
cshift_changed = true;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->prev_cshifts[i].destcolor[j] = vs->cshifts[i].destcolor[j];
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
}
}
}
// drop the damage value
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_DropCShift (&vs->cshifts[CSHIFT_DAMAGE], vs->time, 150);
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
// drop the bonus value
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_DropCShift (&vs->cshifts[CSHIFT_BONUS], vs->time, 100);
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
[renderer] Clean up a lot of recalc_refdef use I think the widespread use of recalc_refdef (and force_fullscreen) was the result of a rushed merge of the renderer and video code (I do seem to remember sprinkling them around). This cleans the two out of the client code.
2022-03-18 15:56:30 +00:00
if (!cshift_changed)
Kill off gl_view.c and sw_view.c. Renderer merge is more or less complete now. Software and gl renderers now calculate color shifts the same way, and sw looks a bit different now. Powerups don't shift enough in sw right now, IMO.
2001-08-10 00:28:57 +00:00
return;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcBlend (vs);
Merge most of cshift calcs from sw and gl to r_view.c
2001-05-21 19:53:57 +00:00
}
header/whitespace sweep continues.
2001-08-28 23:05:45 +00:00
/* VIEW RENDERING */
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
CalcGunAngle (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
vec4f_t rotation = Transform_GetWorldRotation (vs->camera_transform);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
//FIXME make child of camera
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
transform_t wep_form = Entity_Transform (vs->weapon_entity);
Transform_SetWorldRotation (wep_form, rotation);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_BoundOffsets (viewstate_t *vs)
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
{
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
vec4f_t offset = Transform_GetWorldPosition (vs->camera_transform);
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
offset -= vs->player_origin;
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
// absolutely bound refresh reletive to entity clipping hull
// so the view can never be inside a solid wall
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
offset[0] = bound (-14, offset[0], 14);
offset[1] = bound (-14, offset[1], 14);
offset[2] = bound (-22, offset[2], 30);
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
Transform_SetWorldPosition (vs->camera_transform,
vs->player_origin + offset);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
}
static vec4f_t
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
idle_quat (vec4f_t axis, float cycle, float level, double time)
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
{
vec4f_t identity = { 0, 0, 0, 1 };
if (!level || !cycle) {
return identity;
}
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
float scale = sin (time * cycle);
float ang = scale * level * v_idlescale;
[qw] Actually fix the idle scale Oops. I really need to get cl_view merged.
2021-07-24 02:38:28 +00:00
float c = cos (ang * M_PI / 360);
float s = sin (ang * M_PI / 360);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
return axis * s + identity * c;
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
/*
V_AddIdle
Idle swaying
*/
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_AddIdle (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t roll = idle_quat ((vec4f_t) { 1, 0, 0, 0},
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
v_iroll_cycle, v_iroll_level, vs->time);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t pitch = idle_quat ((vec4f_t) { 0, 1, 0, 0},
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
v_ipitch_cycle, v_ipitch_level, vs->time);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t yaw = idle_quat ((vec4f_t) { 0, 0, 1, 0},
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
v_iyaw_cycle, v_iyaw_level, vs->time);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t rot = normalf (qmulf (yaw, qmulf (pitch, roll)));
// rotate the view
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
vec4f_t rotation = Transform_GetWorldRotation (vs->camera_transform);
Transform_SetWorldRotation (vs->camera_transform, qmulf (rot, rotation));
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
// counter-rotate the weapon
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
transform_t wep_form = Entity_Transform (vs->weapon_entity);
rot = qmulf (qconjf (rot), Transform_GetWorldRotation (wep_form));
Transform_SetWorldRotation (wep_form, rot);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
/*
V_CalcViewRoll
Roll is induced by movement and damage
*/
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcViewRoll (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
vec_t *angles = vs->player_angles;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vec4f_t velocity = vs->velocity;
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec3_t ang = { };
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
ang[ROLL] = V_CalcRoll (angles, velocity);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
if (v_dmg_time > 0) {
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
ang[ROLL] += v_dmg_time / v_kicktime * v_dmg_roll;
ang[PITCH] += v_dmg_time / v_kicktime * v_dmg_pitch;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
v_dmg_time -= vs->frametime;
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
}
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->flags & VF_DEAD) { // VF_GIB will also set VF_DEAD
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
ang[ROLL] = 80; // dead view angle
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t rot;
Update vec3_t/vec4f_t hacks to work with clang Still work with gcc, of course, and I still need to fix them properly, but now they're actually slightly easier to find as they all have vec_t and FIXME on the same line.
2022-03-30 15:07:20 +00:00
AngleQuat (ang, (vec_t*)&rot);//FIXME
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
vec4f_t rotation = Transform_GetWorldRotation (vs->camera_transform);
Transform_SetWorldRotation (vs->camera_transform, qmulf (rotation, rot));
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcIntermissionRefdef (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
// vs->player_entity is the player model (visible when out of body)
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
entity_t ent = vs->player_entity;
entity_t view;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
float old;
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
transform_t transform = Entity_Transform (ent);
vec4f_t origin = Transform_GetWorldPosition (transform);
vec4f_t rotation = Transform_GetWorldRotation (transform);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Diff reduction.
2011-06-18 13:22:47 +00:00
// view is the weapon model (visible only from inside body)
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
view = vs->weapon_entity;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
Transform_SetWorldPosition (vs->camera_transform, origin);
Transform_SetWorldRotation (vs->camera_transform, rotation);
[client] Clean up some entity component accesses
2022-11-02 00:53:52 +00:00
renderer_t *renderer = Ent_GetComponent (view.id, scene_renderer, view.reg);
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
renderer->model = NULL;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// always idle in intermission
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
old = v_idlescale;
v_idlescale = 1;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_AddIdle (vs);
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
v_idlescale = old;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcRefdef (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
Much diff noise reduction.
2012-05-31 23:20:09 +00:00
// view is the weapon model (visible only from inside body)
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
entity_t view = vs->weapon_entity;
Diff reduction.
2011-06-18 13:22:47 +00:00
float bob;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
static float oldz = 0;
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
vec4f_t forward = {}, right = {}, up = {};
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
vec4f_t origin = vs->player_origin;
vec_t *viewangles = vs->player_angles;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[client] Clean up some entity component accesses
2022-11-02 00:53:52 +00:00
renderer_t *renderer = Ent_GetComponent (view.id, scene_renderer, view.reg);
animation_t *animation = Ent_GetComponent (view.id, scene_animation, view.reg);
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_DriftPitch (vs);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
bob = V_CalcBob (vs);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Much diff noise reduction in V_CalcRefdef. NQ's client-side onground now works similarly to QW's: -1 = not on ground, 0+ = on ground (plane num?) but only 0 is used in NQ as the necessary information is not sent over the network.
2012-06-02 04:45:52 +00:00
// refresh position
[qw] Move view hight adjustment out of screen update cl_screen.c is that little bit closer to being merged.
2022-11-02 00:52:35 +00:00
if (vs->flags & VF_GIB) {
origin[2] += 8; // gib view height
} else if (vs->flags & VF_DEAD) {
origin[2] += -16; // corpse view height
} else {
origin[2] += vs->height + bob;
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
sw32. Flee in terror. This'll break compilation for everything until -sdl32 compiles. sw32 wont work for a while either, so don't get overly excited yet.
2001-08-25 02:47:11 +00:00
// never let it sit exactly on a node line, because a water plane can
// disappear when viewed with the eye exactly on it.
audit the usage of "only" There are still a few iffy places (notably around certain prepositions), but the relevant sentences are now much easier to read.
2010-01-13 06:42:26 +00:00
// server protocol specifies to only 1/8 pixel, so add 1/16 in each axis
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
origin += (vec4f_t) { 1.0/16, 1.0/16, 1.0/16, 0};
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
vec4f_t rotation;
Update vec3_t/vec4f_t hacks to work with clang Still work with gcc, of course, and I still need to fix them properly, but now they're actually slightly easier to find as they all have vec_t and FIXME on the same line.
2022-03-30 15:07:20 +00:00
AngleQuat (vs->player_angles, (vec_t*)&rotation);//FIXME
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
Transform_SetWorldRotation (vs->camera_transform, rotation);
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_CalcViewRoll (vs);
V_AddIdle (vs);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// offsets
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
//FIXME semi-duplicates AngleQuat (also, vec3_t vs vec4f_t)
Update vec3_t/vec4f_t hacks to work with clang Still work with gcc, of course, and I still need to fix them properly, but now they're actually slightly easier to find as they all have vec_t and FIXME on the same line.
2022-03-30 15:07:20 +00:00
AngleVectors (viewangles, (vec_t*)&forward, (vec_t*)&right, (vec_t*)&up);//FIXME
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
// don't allow cheats in multiplayer
// FIXME check for dead
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->voffs_enabled) {
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
origin += scr_ofsx * forward + scr_ofsy * right + scr_ofsz * up;
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
}
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_BoundOffsets (vs);
Pretend to allow screen offsets. Because qw's maxclients is always > 1, they'll never work, though :)
2012-05-31 23:19:22 +00:00
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// set up gun position
[client] Get the basic chase camera working The more advanced modes are rather broken (continuous spinning), but they may have been for a while. The bulk of the various changes were due to renaming viewstate's origin and angles to make their meaning more explicit.
2022-03-01 02:43:23 +00:00
vec4f_t gun_origin = vs->player_origin;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
CalcGunAngle (vs);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
gun_origin += (vec4f_t) { 0, 0, vs->height, 0 };
gun_origin += forward * bob * 0.4f + (vec4f_t) { 0, 0, bob, 0 };
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
sw32. Flee in terror. This'll break compilation for everything until -sdl32 compiles. sw32 wont work for a while either, so don't get overly excited yet.
2001-08-25 02:47:11 +00:00
// fudge position around to keep amount of weapon visible
// roughly equal with different FOV
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (hud_sbar == 0 && *r_data->scr_viewsize >= 100) {
A modification to the gun location hack code.. The QW HUD was supposed to give you a little more screen real-estate. It does, but you still have a big gun blocking most of your screen in the center. If you turned off the HUD, you got more screen and less gun. That now happens whenever the HUD is displayed. This is a temporary measure. At a later point, the gun will always be at the bottom of the view area regardless of fov.
2001-02-21 19:13:03 +00:00
;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
} else if (*r_data->scr_viewsize == 110) {
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
gun_origin += (vec4f_t) { 0, 0, 1, 0};
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
} else if (*r_data->scr_viewsize == 100) {
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
gun_origin += (vec4f_t) { 0, 0, 2, 0};
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
} else if (*r_data->scr_viewsize == 90) {
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
gun_origin += (vec4f_t) { 0, 0, 1, 0};
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
} else if (*r_data->scr_viewsize == 80) {
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
gun_origin += (vec4f_t) { 0, 0, 0.5, 0};
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
}
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
model_t *model = vs->weapon_model;
if (vs->flags & (VF_GIB | VF_DEAD)) {
[nq,qw] Reset view animation pose when model changes I think this bug has been haunting us since we introduced lerping. I do remember chasing bad pose indices 20 years ago and never finding the cause. I guess we never thought to check the view model.
2021-11-30 16:10:59 +00:00
model = NULL;
}
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
if (renderer->model != model) {
animation->pose2 = -1;
[renderer] Clean up entity_t to a certain extent This is the first step towards component-based entities. There's still some transform-related stuff in the struct that needs to be moved, but it's all entirely client related (rather than renderer) and will probably go into a "client" component. Also, the current components are directly included structs rather than references as I didn't want to deal with the object management at this stage. As part of the process (because transforms use simd) this also starts the process of moving QF to using simd for vectors and matrices. There's now a mess of simd and sisd code mixed together, but it works surprisingly well together.
2021-03-09 14:52:40 +00:00
}
[scene] Make entity_t just an entity id for ECS This puts the hierarchy (transform) reference, animation, visibility, renderer, active, and old_origin data in separate components. There are a few bugs (crashes on grenade explosions in gl/glsl/vulkan, immediately in sw, reasons known, missing brush models in vulkan). While quake doesn't really need an ECS, the direction I want to take QF does, and it does seem to have improved memory bandwidth a little (uncertain). However, there's a lot more work to go (especially fixing the above bugs), but this seems to be a good start.
2022-10-23 01:32:09 +00:00
renderer->model = model;
animation->frame = vs->weaponframe;
renderer->skin = 0;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// set up the refresh position
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
Transform_SetWorldRotation (vs->camera_transform,
qmulf (vs->punchangle, rotation));
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
Merges and cleanups...
2001-05-13 22:57:27 +00:00
// smooth out stair step ups
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
if ((vs->onground != -1) && (gun_origin[2] - oldz > 0)) {
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
float steptime;
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
steptime = vs->frametime;
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
oldz += steptime * 80;
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
if (oldz > gun_origin[2])
oldz = gun_origin[2];
if (gun_origin[2] - oldz > 12)
oldz = gun_origin[2] - 12;
origin[2] += oldz - gun_origin[2];
gun_origin[2] += oldz - gun_origin[2];
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
} else {
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
oldz = gun_origin[2];
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
}
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
Transform_SetWorldPosition (vs->camera_transform, origin);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
{
// FIXME sort out the alias model specific negation
vec3_t ang = {-viewangles[0], viewangles[1], viewangles[2]};
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
CL_TransformEntity (view, 1, ang, gun_origin);
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
make gcc more anal about prototypes, string constants and function visibility (ie, global functions must have a prototype)
2003-01-06 18:28:13 +00:00
static void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
DropPunchAngle (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vec4f_t punch = vs->punchangle;
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
float ps = magnitude3f (punch)[0];
if (ps < 1e-3) {
// < 0.2 degree rotation, not worth worrying about
//ensure the quaternion is normalized
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->punchangle = (vec4f_t) { 0, 0, 0, 1 };
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
return;
}
[nq,qw] Clean up some bogus diffs in cl_view The diffs were caused by white space and a redundant line.
2022-02-25 02:39:11 +00:00
float pc = punch[3];
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
float ds = 0.0871557427 * vs->frametime;
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
float dc = sqrt (1 - ds * ds);
float s = ps * dc - pc * ds;
float c = pc * dc + ps * ds;
if (s <= 0 || c >= 1) {
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
vs->punchangle = (vec4f_t) { 0, 0, 0, 1 };
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
} else {
punch *= s / ps;
punch[3] = c;
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
/*
V_RenderView
The player's clipping box goes from (-16 -16 -24) to (16 16 32) from
the entity origin, so any view position inside that will be valid
*/
void
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
V_RenderView (viewstate_t *vs)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (!vs->active) {
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
vec4f_t base = { 0, 0, 0, 1 };
Transform_SetWorldPosition (vs->camera_transform, base);
Transform_SetWorldRotation (vs->camera_transform, base);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
return;
[renderer] Remove more old fields from entity_t The only transform related field remaining is old_origin. This also brings the renderer closer to using simd (lots of stuff to fix still, though).
2021-03-19 11:18:45 +00:00
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
if (vs->decay_punchangle) {
DropPunchAngle (vs);
}
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
if (vs->intermission) { // intermission / finale rendering
V_CalcIntermissionRefdef (vs);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
} else {
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (vs->chase && chase_active) {
[client] Fix the chase cameras They no longer spin like crazy. I don't know how, but I must have broken something over the years as I'm sure Seth had the code working (and I seem to remember seeing it working). In the process, clean up a lot of the angle mess.
2022-03-01 05:44:53 +00:00
Chase_Update (vs->chasestate);
} else {
V_CalcRefdef (vs);
}
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
}
void
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
V_Init (viewstate_t *viewstate)
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
{
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
Cmd_AddDataCommand ("bf", V_BonusFlash_f, viewstate,
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
"Background flash, used when you pick up an item");
[client] Merge nq and qw cl_view They've been near-identical for years, now they're only one. It proved necessary to start merging the HUD code which for now is just a few cvar declarations (not even init), but that should be a separate set of commits.
2022-02-28 03:12:51 +00:00
Cmd_AddDataCommand ("centerview", V_StartPitchDrift_f, viewstate,
[nq,qw] Clean up cl_view's use of the client struct This is a huge step towards merging cl_view.
2022-02-25 06:48:57 +00:00
"Centers the player's "
Diff reduction.
2011-06-18 13:22:47 +00:00
"view ahead after +lookup or +lookdown\n"
sw32. Flee in terror. This'll break compilation for everything until -sdl32 compiles. sw32 wont work for a while either, so don't get overly excited yet.
2001-08-25 02:47:11 +00:00
"Will not work while mlook is active or freelook is 1.");
Whitespace.
2001-09-10 17:32:22 +00:00
Cmd_AddCommand ("v_cshift", V_cshift_f, "This adjusts all of the colors "
"currently being displayed.\n"
"Used when you are underwater, hit, have the Ring of "
"Shadows, or Quad Damage. (v_cshift r g b intensity)");
[renderer] Pass in a camera transform More r_data cleanup. Things could be better still, but this is a start.
2022-02-28 07:59:38 +00:00
[ecs] Move ECS core into its own library While the libraries are probably getting a little out of hand, the separation into its own directory is probably a good thing as an ECS should not be tied to scenes. This should make the ECS more generally useful.
2022-10-26 06:22:09 +00:00
viewstate->camera_transform = Transform_New (cl_world.scene->reg,
nulltransform);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
void
V_Init_Cvars (void)
{
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
Cvar_Register (&v_centermove_cvar, 0, 0);
Cvar_Register (&v_centerspeed_cvar, 0, 0);
Cvar_Register (&v_iyaw_cycle_cvar, 0, 0);
Cvar_Register (&v_iroll_cycle_cvar, 0, 0);
Cvar_Register (&v_ipitch_cycle_cvar, 0, 0);
Cvar_Register (&v_iyaw_level_cvar, 0, 0);
Cvar_Register (&v_iroll_level_cvar, 0, 0);
Cvar_Register (&v_ipitch_level_cvar, 0, 0);
Cvar_Register (&v_idlescale_cvar, 0, 0);
Cvar_Register (&scr_ofsx_cvar, 0, 0);
Cvar_Register (&scr_ofsy_cvar, 0, 0);
Cvar_Register (&scr_ofsz_cvar, 0, 0);
Cvar_Register (&cl_rollspeed_cvar, 0, 0);
Cvar_Register (&cl_rollangle_cvar, 0, 0);
Cvar_Register (&cl_bob_cvar, 0, 0);
Cvar_Register (&cl_bobcycle_cvar, 0, 0);
Cvar_Register (&cl_bobup_cvar, 0, 0);
Cvar_Register (&v_kicktime_cvar, 0, 0);
Cvar_Register (&v_kickroll_cvar, 0, 0);
Cvar_Register (&v_kickpitch_cvar, 0, 0);
Cvar_Register (&cl_cshift_bonus_cvar, 0, 0);
Cvar_Register (&cl_cshift_contents_cvar, 0, 0);
Cvar_Register (&cl_cshift_damage_cvar, 0, 0);
Cvar_Register (&cl_cshift_powerup_cvar, 0, 0);
initial checkin of most recent newtree and nuq(?) source
2001-02-19 21:15:25 +00:00
}
Reference in a new issue Copy permalink