less hacky solution to resetting movesequence on a level change

git-svn-id: svn://svn.icculus.org/twilight/trunk/darkplaces@6931 d7cf8633-e32d-0410-b094-e92efae38249

diff --git a/sv_main.c b/sv_main.c
index 0bebd163e50400fc9e18d756d9d8de048dbcd53f..c9a237a4d53ce33185fcb219a3514aa86d77bce7 100644 (file)
--- a/sv_main.c
+++ b/sv_main.c
@@ -447,6 +447,16 @@ void SV_SendServerinfo (client_t *client)
        MSG_WriteByte (&client->netconnection->message, 1);
 
        client->spawned = false;                // need prespawn, spawn, etc
+
+       // clear movement info until client enters the new level properly
+       memset(&client->cmd, 0, sizeof(client->cmd));
+       client->movesequence = 0;
+#ifdef NUM_PING_TIMES
+       for (i = 0;i < NUM_PING_TIMES;i++)
+               client->ping_times[i] = 0;
+       client->num_pings = 0;
+#endif
+       client->ping = 0;
 }
 
 /*

diff --git a/sv_user.c b/sv_user.c
index ce1fcf17b5243552dd616f3dc6bcb7bdfcc09824..f4ea2cff108f05307edf58b408924779411d699f 100644 (file)
--- a/sv_user.c
+++ b/sv_user.c
@@ -445,20 +445,6 @@ void SV_ReadClientMove (void)
        usercmd_t newmove;
        usercmd_t *move = &newmove;
 
-       // reset stale client move state if a level change has occurred
-       if (host_client->cmd.receivetime > sv.time + 0.1 || !host_client->spawned)
-       {
-               memset(&host_client->cmd, 0, sizeof(host_client->cmd));
-               host_client->movesequence = 0;
-#ifdef NUM_PING_TIMES
-               for (i=0;i < NUM_PING_TIMES;i++)
-                       host_client->ping_times[i] = 0;
-               host_client->num_pings = 0;
-#endif
-               host_client->ping = 0;
-               host_client->clmovement_disabletimeout = realtime + 0.1;
-       }
-
        memset(move, 0, sizeof(*move));
 
        if (msg_badread) Con_Printf("SV_ReadClientMessage: badread at %s:%i\n", __FILE__, __LINE__);

diff --git a/todo b/todo
index ef0ea44ce1cf2f3ce541fc01a7228a1d3fac5138..f5325af4ac8d43fac3452e764cdbef392b3083d4 100644 (file)
--- a/todo
+++ b/todo
@@ -62,7 +62,6 @@
 -f bug darkplaces physics: GAME_TAOV: Vigil's movement isn't working properly, the qc uses MOVETYPE_STEP and clears FL_ONGROUND every frame and moves using velocity, this is causing a landing sound every frame and causing the player to slide down minor slopes very quickly, this did not occur in Quake, and seems that it must be related to a velocity_z check or FL_ONGROUND check in the MOVETYPE_STEP physics code (RenegadeC, xaGe)
 0 bug darkplaces client: can't move mouse around in nexuiz menu if vid_mouse is 0
 0 bug darkplaces client: it has been reported that sometimes level changes on quakeworld servers don't load a map, this may be related to downloading? (Baker)
-0 bug darkplaces client: name (and probably other userinfo properties) are not being set when entering a qw server?
 0 bug darkplaces client: when going through a teleporter the cl_movement prediction still interpolates the move (div0)
 0 bug darkplaces docs: host_maxfps is gone, correct the darkplaces.txt and host.c cvar description for host_framerate
 0 bug darkplaces loader: make rtlight entity loader support q3map/q3map2 lights properly, they use a spawnflag for LINEAR mode, by default they use 1/(x*x) falloff (Carni, motorsep)
@@ -533,6 +532,7 @@ d bug darkplaces client: make "wait" command wait fornext network frame somehow
 d bug darkplaces client: make envmap command work with the corrected layout
 d bug darkplaces client: make server queries use a queue to avoid flooding out queries too fast (Willis)
 d bug darkplaces client: missing bolt/beam models should not produce warnings
+d bug darkplaces client: name (and probably other userinfo properties) are not being set when entering a qw server?
 d bug darkplaces client: on crctf proquake servers the scoreboard does not contain exactly matching player names (READY is sometimes appended), the ping report and status parsing should ignore text after the player name
 d bug darkplaces client: quakeworld servers often stuffcmd the cvars topcolor, bottomcolor, pants, team, skin, noaim, so commands for these need to be added (topcolor/bottomcolor will modify _cl_color, the others can be real cvars)
 d bug darkplaces client: seta commands create cvars that are not saved to config because they match their 'default' value