战团中战场人数调大后，发生定时卡顿的代码优化（全MOD通用）[更新TXT修改方法]]

aliang227 · 发表于 2022-3-2 16:54:42

本帖最后由 aliang227 于 2022-6-15 17:57 编辑

# 更新TXT修改方式，有同样问题想自行修改的汽油可以参考下。详见48楼。
# 手头上有源码，但对MS代码不是很精通的汽油，也可以参看8楼，有相对简单点的修改方法。

当玩家通过修改battle_size参数，或用战场人数调节工具修改战场人数上限，把战场人数改到2、300人的时候，野战时会有一个比较规律性的顿卡，配置高的会定时掉帧，配置低的会定时顿一下。
开始我以为是自己代码的问题，后来发现战团原版也会有这个问题。经过多次观察和排查之后，基本锁定了是在lead_charge这个mt里的这个触发器：

(3, 0, 0, [ (call_script, "script_apply_effect_of_other_people_on_courage_scores"),], [ ]), #calculating and applying effect of people on others courage scores

复制代码

引用的script代码如下：

（点击展开 / 收起）

# script_apply_effect_of_other_people_on_courage_scores
# Input: none
# Output: none
("apply_effect_of_other_people_on_courage_scores",
[
(get_player_agent_no, ":player_agent"),
(try_for_agents, ":centered_agent_no"),
(agent_is_human, ":centered_agent_no"),
(agent_is_alive, ":centered_agent_no"),
(neq, ":centered_agent_no", ":player_agent"),
(agent_get_position, pos0, ":centered_agent_no"),
(try_begin),
(agent_is_ally, ":centered_agent_no"),
(assign, ":is_centered_agent_ally", 1),
(else_try),
(assign, ":is_centered_agent_ally", 0),
(try_end),
(try_for_agents, ":agent_no"),
(agent_is_human, ":agent_no"),
(agent_is_alive, ":agent_no"),
(neq, ":centered_agent_no", ":agent_no"),
(try_begin),
(agent_is_ally, ":agent_no"),
(assign, ":is_agent_ally", 1),
(else_try),
(assign, ":is_agent_ally", 0),
(try_end),
(eq, ":is_centered_agent_ally", ":is_agent_ally"), #if centered agent and other agent is at same team then continue.
(agent_get_slot, ":agent_is_running_away_or_not", ":agent_no", slot_agent_is_running_away),
(try_begin),
(eq, ":agent_no", ":player_agent"),
(assign, ":agent_delta_courage_score", 6),
(else_try),
(agent_get_troop_id, ":troop_id", ":agent_no"),
(troop_is_hero, ":troop_id"),
#Hero Agent : if near agent (hero, agent_no) is not running away his positive effect on centered agent (centered_agent_no) fighting at his side is effected by his hit points.
(try_begin),
(neq, ":agent_is_running_away_or_not", 1), #if agent is not running away
(store_agent_hit_points, ":agent_hit_points", ":agent_no"),
(try_begin),
(eq, ":agent_hit_points", 100),
(assign, ":agent_delta_courage_score", 6),
(else_try),
(ge, ":agent_hit_points", 75),
(assign, ":agent_delta_courage_score", 5),
(else_try),
(ge, ":agent_hit_points", 60),
(assign, ":agent_delta_courage_score", 4),
(else_try),
(ge, ":agent_hit_points", 45),
(assign, ":agent_delta_courage_score", 3),
(else_try),
(ge, ":agent_hit_points", 30),
(assign, ":agent_delta_courage_score", 2),
(else_try),
(ge, ":agent_hit_points", 15),
(assign, ":agent_delta_courage_score", 1),
(end_try),
(else_try),
(assign, ":agent_delta_courage_score", 4),
(end_try),
(else_try),
#Normal Agent : if near agent (agent_no) is not running away his positive effect on centered agent (centered_agent_no) fighting at his side is effected by his hit points.
(try_begin),
(neq, ":agent_is_running_away_or_not", 1), # if agent is not running away
(store_agent_hit_points, ":agent_hit_points", ":agent_no"),
(try_begin),
(eq, ":agent_hit_points", 100),
(assign, ":agent_delta_courage_score", 4),
(else_try),
(ge, ":agent_hit_points", 75),
(assign, ":agent_delta_courage_score", 3),
(else_try),
(ge, ":agent_hit_points", 50),
(assign, ":agent_delta_courage_score", 2),
(else_try),
(ge, ":agent_hit_points", 25),
(assign, ":agent_delta_courage_score", 1),
(try_end),
(try_begin), # to make our warrior run away easier we decrease one, because they have player_agent (+6) advantage.
(agent_is_ally, ":agent_no"),
(val_sub, ":agent_delta_courage_score", 1),
(try_end),
(else_try),
(assign, ":agent_delta_courage_score", 2),
(end_try),
(try_end),
(try_begin),
(neq, ":agent_is_running_away_or_not", 1),
(val_mul, ":agent_delta_courage_score", 1),
(try_begin), # centered agent not running away cannot take positive courage score from one another agent not running away.
(agent_get_slot, ":agent_is_running_away_or_not", ":centered_agent_no", slot_agent_is_running_away),
(eq, ":agent_is_running_away_or_not", 0),
(val_mul, ":agent_delta_courage_score", 0),
(try_end),
(else_try),
(try_begin),
(agent_get_slot, ":agent_is_running_away_or_not", ":agent_no", slot_agent_is_running_away),
(eq, ":agent_is_running_away_or_not", 0),
(val_mul, ":agent_delta_courage_score", -2), # running away agent fears not running away agent more.
(else_try),
(val_mul, ":agent_delta_courage_score", -1),
(try_end),
(try_end),
(neq, ":agent_delta_courage_score", 0),
(agent_get_position, pos1, ":agent_no"),
(get_distance_between_positions, ":dist", pos0, pos1),
(try_begin),
(ge, ":agent_delta_courage_score", 0),
(try_begin),
(lt, ":dist", 2000), #0-20 meter
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 50),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 4000), #21-40 meter
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 40),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 7000), #41-70 meter
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 30),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 11000), #71-110 meter
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 20),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 16000), # 111-160 meter, assumed that eye can see agents friendly at most 160 meters far while fighting.
# this is more than below limit (108 meters) because we hear that allies come from further.
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 10),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(try_end),
(else_try), # negative effect of running agent on other ally agents are lower then positive effects above, to avoid starting
(try_begin), # run away of all agents at a moment. I want to see agents running away one by one during battle, not all together.
(lt, ":dist", 200), #1-2 meter, # this would create better game play.
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 15),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 400), #3-4 meter,
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 13),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 600), #5-6 meter
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 11),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 800), #7-8 meter
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 9),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 1200), #9-12 meters
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 7),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 2400), #13-24 meters
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 5),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 4800), #25-48 meters
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 3),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(else_try),
(lt, ":dist", 9600), #49-98 meters, assumed that eye can see agents running away at most 98 meters far while fighting.
(agent_get_slot, ":agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(val_mul, ":agent_delta_courage_score", 1),
(val_add, ":agent_courage_score", ":agent_delta_courage_score"),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":agent_courage_score"),
(try_end),
(try_end),
(try_end),
(try_end),
]), #ozan

复制代码

我们在做MOD的时候，对于原版的代码一般都尽量不会动他的框架，更多的是改改判定和数值，因为默认原版的代码要比自己写的好一些。但是这个署名 #ozan 的代码确实不敢恭维。
我们先看看他这个script到底想干嘛：
1，以一个agent为中心，判断周围agent和他的关系。第一个雷点：这个是两个agent循环嵌套，相信有过一定MOD开发经验的作者都会尽量避免这种搞法，因为条件限制不足的话必卡。150战场就是150*150=25500次循环，300人战场就是90000次循环。

2，判断是否在一个队伍，如果是的话就根据周围agent自身的血量，来计算其勇气值(courage_score)。第二个雷点：这个值相同时间内是定的，即使要算，单列出来算只需要循环agent数量次，但是他放在第二个循环内，每次就计算了agent数量个平方次，也就是只需要算300次的东西他给算了90000次。

3，按照中心agent和周围agent的是否逃跑状态，来分为四种情况：
  1）中心agent在逃跑，周围agent没逃跑：增加中心agent相应勇气值。也就是让逃跑的人再回头继续打架。
  2）中心agent没逃跑，周围agent没逃跑：什么事情也没发生。这是大多数情况，也就是每3秒都在干这个没意义的运算，然后啥事没有。搁这儿挖矿呢……
  3）中心agent没逃跑，周围agent在逃跑：减少两倍中心agent的相应勇气值。也就是没逃跑的被逃跑的人动摇了。
  4）中心agent在逃跑，周围agent在逃跑：减少中心agent相应勇气值。但是他这块写错了，几个局部变量的混用估计把自己弄糊涂了，这个情况实际是不会生效的，全算在情况3里头了。

4，根据两者的距离来计算倍率，越近影响越大。而远了的呢？照算不误，只是不加进来。

归纳一下他的想法，agent逃跑时会影响周围一定范围内的人，周围的人如果没跑，那人减勇气，自己加勇气；周围的人如果在跑，两边都减勇气。
因此优化的途径也是把他代码和思路的雷点避开掉：
1，有逃跑的人才开算，而且只循环逃跑的人，没跑的我管他干嘛；
2，把一些不必要放在二次循环的内容单拿出来算，用的时候直接调就行，减少运算量；
3，修正两个人都在跑时的勇气计算

因此我照此思路修改后的代码如下：

1，mt里把原来那个触发器注释掉，修改为以下内容。也就是直到有人跑的时候才开始计算，并把跑的人数和agent_no预先保存下来，以减少循环和计算。
说明一下， slot_item_temp_slot 这个临时slot需要自己预先定义一下，因为ms里没有数组，我就用item随便定义了一个临时slot当数组。当然，如果你用troop/agent之类的都无所谓，看个人习惯了。

（点击展开 / 收起）

2，修改之前的script_apply_effect_of_other_people_on_courage_scores，或者新加个script_apply_effect_of_other_people_on_courage_scores_new以调用。思路也是如上所述：

1，预先计算每个人的勇气值，保存在临时slot里（slot_agent_temp_slot，需要自己先定义下）

2，增加一个入参，只对逃跑的agent周围进行循环，减少循环次数；

3，利用1.168增加的可选范围agent循环，用(try_for_agents, ":agent_no",pos2,160*1000), 将循环范围限制到其判定的最大范围内。但是得记得先(set_fixed_point_multiplier,1000), 然后避开pos0；

4，因为只从逃跑的agent角度出发了，因此以他为中心，判定时就只有两种情况了：

1）周围的agent没跑，那么他自己会扣除中心agent计算出来的勇气值；中心agent则会自己增加周围没跑的agent计算出来勇气值；
2）周围的agent也在跑，那么只让中心agent扣除周围agent计算出来的勇气值；因为周围agent也在逃跑的话，他之后也会循环到，没必要两人都扣，这样就会重复扣2次了。

5，根据两者距离乘以系数，最后附加到相应勇气slot里。

（点击展开 / 收起）

# script_apply_effect_of_other_people_on_courage_scores_new
# Input: arg1 = now rout agent number, should have saved to slot_item_temp_slot before
# Output: none
# Write by aliang227
("apply_effect_of_other_people_on_courage_scores_new",
[
(store_script_param, ":rout_num", 1),
(get_player_agent_no, ":player_agent"),
#count agent_delta_courage_score first, save to slot_agent_temp_slot
(try_for_agents, ":agent_no"),
(agent_is_human, ":agent_no"),
(agent_is_alive, ":agent_no"),
(agent_set_slot,":agent_no",slot_agent_temp_slot,0),#ini
(agent_get_slot, ":agent_is_running_away_or_not", ":agent_no", slot_agent_is_running_away),
(try_begin),
(eq, ":agent_no", ":player_agent"),
(assign, ":agent_delta_courage_score", 6),
(else_try),
(agent_get_troop_id, ":troop_id", ":agent_no"),
(troop_is_hero, ":troop_id"),
#Hero Agent : if near agent (hero, agent_no) is not running away his positive effect on centered agent (centered_agent_no) fighting at his side is effected by his hit points.
(try_begin),
(neq, ":agent_is_running_away_or_not", 1), #if agent is not running away
(store_agent_hit_points, ":agent_hit_points", ":agent_no"),
(try_begin),
(eq, ":agent_hit_points", 100),
(assign, ":agent_delta_courage_score", 6),
(else_try),
(ge, ":agent_hit_points", 75),
(assign, ":agent_delta_courage_score", 5),
(else_try),
(ge, ":agent_hit_points", 60),
(assign, ":agent_delta_courage_score", 4),
(else_try),
(ge, ":agent_hit_points", 45),
(assign, ":agent_delta_courage_score", 3),
(else_try),
(ge, ":agent_hit_points", 30),
(assign, ":agent_delta_courage_score", 2),
(else_try),
(ge, ":agent_hit_points", 15),
(assign, ":agent_delta_courage_score", 1),
(try_end),
(else_try),
(assign, ":agent_delta_courage_score", 4),
(try_end),
(else_try),
#Normal Agent : if near agent (agent_no) is not running away his positive effect on centered agent (centered_agent_no) fighting at his side is effected by his hit points.
(try_begin),
(neq, ":agent_is_running_away_or_not", 1), # if agent is not running away
(store_agent_hit_points, ":agent_hit_points", ":agent_no"),
(try_begin),
(eq, ":agent_hit_points", 100),
(assign, ":agent_delta_courage_score", 4),
(else_try),
(ge, ":agent_hit_points", 75),
(assign, ":agent_delta_courage_score", 3),
(else_try),
(ge, ":agent_hit_points", 50),
(assign, ":agent_delta_courage_score", 2),
(else_try),
(ge, ":agent_hit_points", 25),
(assign, ":agent_delta_courage_score", 1),
(try_end),
(try_begin), # to make our warrior run away easier we decrease one, because they have player_agent (+6) advantage.
(agent_is_ally, ":agent_no"),
(val_sub, ":agent_delta_courage_score", 1),
(try_end),
(else_try),
(assign, ":agent_delta_courage_score", 2),
(try_end),
(try_end),
(agent_set_slot,":agent_no",slot_agent_temp_slot,":agent_delta_courage_score"),
(try_end),
#try for running agents
(set_fixed_point_multiplier,1000),
(try_for_range,":agent_item_no",0,":rout_num"),
(item_get_slot,":centered_agent_no",":agent_item_no",slot_item_temp_slot),
#check again
(agent_is_human, ":centered_agent_no"),
(agent_is_alive, ":centered_agent_no"),
(neg|agent_is_routed,":centered_agent_no"),
(agent_get_position, pos2, ":centered_agent_no"),
(agent_get_team,":centered_agent_team",":centered_agent_no"),
(agent_get_slot, ":centered_agent_courage_score", ":centered_agent_no", slot_agent_courage_score),
(try_for_agents, ":agent_no",pos2,160*1000),#max dis is 160m, avoid using pos0
(agent_is_human, ":agent_no"),
(agent_is_alive, ":agent_no"),
(neq, ":centered_agent_no", ":agent_no"),
(agent_get_team,":agent_team", ":agent_no"),
#if centered agent and other agent is at same team then continue.
(neg|teams_are_enemies,":centered_agent_team",":agent_team"),
(agent_get_slot,":agent_delta_courage_score",":agent_no",slot_agent_temp_slot),
(agent_get_slot,":centered_agent_delta_courage_score",":centered_agent_no",slot_agent_temp_slot),
(neq, ":agent_delta_courage_score", 0),#not 0
(store_mul,":agent_delta_courage_score_pos",":agent_delta_courage_score",1),
(store_mul,":agent_delta_courage_score_neg",":agent_delta_courage_score",-1),
#distance count
(agent_get_position, pos1, ":agent_no"),
(get_distance_between_positions, ":dist", pos2, pos1),
(try_begin),
(lt, ":dist", 2000), #0-20 meter
(val_mul, ":agent_delta_courage_score_pos", 50),
(else_try),
(lt, ":dist", 4000), #21-40 meter
(val_mul, ":agent_delta_courage_score_pos", 40),
(else_try),
(lt, ":dist", 7000), #41-70 meter
(val_mul, ":agent_delta_courage_score_pos", 30),
(else_try),
(lt, ":dist", 11000), #71-110 meter
(val_mul, ":agent_delta_courage_score_pos", 20),
(else_try),
(lt, ":dist", 16000), # 111-160 meter, assumed that eye can see agents friendly at most 160 meters far while fighting.
# this is more than below limit (108 meters) because we hear that allies come from further.
(val_mul, ":agent_delta_courage_score_pos", 10),
(else_try),
(val_mul, ":agent_delta_courage_score_pos", 0),
(try_end),
# negative effect of running agent on other ally agents are lower then positive effects above, to avoid starting
(try_begin), # run away of all agents at a moment. I want to see agents running away one by one during battle, not all together.
(lt, ":dist", 200), #1-2 meter, # this would create better game play.
(val_mul, ":agent_delta_courage_score_neg", 15),
(val_mul, ":centered_agent_delta_courage_score", 15),
(else_try),
(lt, ":dist", 400), #3-4 meter,
(val_mul, ":agent_delta_courage_score_neg", 13),
(val_mul, ":centered_agent_delta_courage_score", 13),
(else_try),
(lt, ":dist", 600), #5-6 meter
(val_mul, ":agent_delta_courage_score_neg", 11),
(val_mul, ":centered_agent_delta_courage_score", 11),
(else_try),
(lt, ":dist", 800), #7-8 meter
(val_mul, ":agent_delta_courage_score_neg", 9),
(val_mul, ":centered_agent_delta_courage_score", 9),
(else_try),
(lt, ":dist", 1200), #9-12 meters
(val_mul, ":agent_delta_courage_score_neg", 7),
(val_mul, ":centered_agent_delta_courage_score", 7),
(else_try),
(lt, ":dist", 2400), #13-24 meters
(val_mul, ":agent_delta_courage_score_neg", 5),
(val_mul, ":centered_agent_delta_courage_score", 5),
(else_try),
(lt, ":dist", 4800), #25-48 meters
(val_mul, ":agent_delta_courage_score_neg", 3),
(val_mul, ":centered_agent_delta_courage_score", 3),
(else_try),
(lt, ":dist", 9600), #49-98 meters, assumed that eye can see agents running away at most 98 meters far while fighting.
(val_mul, ":agent_delta_courage_score_neg", 1),
(val_mul, ":centered_agent_delta_courage_score", 1),
(else_try),
(val_mul, ":agent_delta_courage_score_neg", 0),
(val_mul, ":centered_agent_delta_courage_score", 0),
(try_end),
(agent_get_slot, ":agent_is_running_away_or_not", ":agent_no", slot_agent_is_running_away),
#pos or neg
(try_begin),#agent is not running
(neq,":agent_is_running_away_or_not",1),
(assign,":centered_agent_courage_score_add",":agent_delta_courage_score_pos"),
(store_mul,":agent_courage_score_add",":centered_agent_delta_courage_score",-2),
(else_try),#agent is running
(assign,":centered_agent_courage_score_add",":agent_delta_courage_score_neg"),
(assign,":agent_courage_score_add",0),
(try_end),
#set agent score
(agent_get_slot, ":agent_courage_score", ":agent_no", slot_agent_courage_score),
(val_add,":agent_courage_score",":agent_courage_score_add"),
(agent_set_slot, ":agent_no", slot_agent_courage_score, ":agent_courage_score"),
#set center agent score
(val_add,":centered_agent_courage_score",":centered_agent_courage_score_add"),
(try_end),
(agent_set_slot, ":centered_agent_no", slot_agent_courage_score, ":centered_agent_courage_score"),
(try_end),
]),

复制代码

至此，应该修复与优化完成了，再进游戏开大战场时，应该不会再有这种3秒一次的卡顿了。如果还有的话，那就是其他代码造成的了。
因为是战团原版代码的问题，因此全MOD应该都是通用的，各位作者可随意使用。
虽然现在战团也差不多快到寿命的尾巴了，但给作品留个疙瘩还是有点难受，可以的话还是尽量避免掉吧。
另外最近还发现另外一个各类MOD可优化的点：如果你的装备贴图dds尺寸不是2的次幂的话，每次看到时载入贴图时将会发生顿卡，可能是会自动拉伸一下的原因，同时发生这种现象的贴图越多，画面会越卡。解决方式就是排查下textures文件夹的所有文件，把所有不合规格的贴图调整为256、512、1024之类的尺寸就好了。似乎只限装备贴图，图片那种mesh载入图之类的好像影响不大。

最后，如果优化完毕后的代码有什么粗心所致的疏忽或逻辑问题，也希望能指正探讨，大家一起学习。

aliang227 · 发表于 2022-6-15 17:48:42

本帖最后由 aliang227 于 2022-6-18 15:54 编辑

补充一下TXT简单的修改思路吧，因为原意还是想和各位MOD作者讨论，然后直接从源码角度修改。
但确实有很多MOD的玩家也有这样的需求，而且对MS并不了解，或者没有所玩MOD的代码，只需要直接的TXT修改方法。这里还是对比了下供大家参考吧

1，打开MOD文件夹下的“mission_templates.txt”文件，ctrl+f 先搜索 “mst_lead_charge”这句

2，再ctrl+f搜索 “3.000000 0.000000 0.000000 1 1 1”，应该可以搜到类似这句：

3.000000 0.000000 0.000000 1 1 1 936748722493063474 0

因为每个MOD的script顺序不一，红色标记的这个数也可能不一样，所以还得自己确认清楚，现在这个数是原版native的。

3，然后有这样两种简单修改办法，任选其一：

a) 直接简单粗暴屏蔽这句，也就是逃跑的人不会影响周围人的士气，以及逃跑的人不会再反过头来战斗。估计实际影响也没那么大？

可将之前那句替换为（注意红色标记处需对应你MOD里所搜到的相应数字）：

3.000000 0.000000 100000000.000000 1 1 1 936748722493063474 0

也就是把触发器中的 3,0,0改为 3,0,ti_once，只会触发一次

b) 有人逃跑的时候才触发，平时不触发。这种相对折衷一点，有人跑的时候战场人数应该也不多了，应该不会卡的那么明显

可将之前那句替换为（注意红色标记处需对应你MOD里所搜到的相应数字）：

3.000000 0.000000 0.000000 10 2133 2 1224979098644774912 0 12 1 1224979098644774913 31 2 1224979098644774912 0 1704 1 1224979098644774913 1702 1 1224979098644774913 2147485347 1 1224979098644774913 545 3 1224979098644774913 15 1 2133 2 1224979098644774912 1 3 0 31 2 1224979098644774912 1 1 1 1 936748722493063474

当然根本解决方法还是源码修改了再编译更好些。乱舞水浒MOD已经改过了，其他一些国内MOD如果你搜不到上面那句的话，可能也是已经改过了，不用再做修改了。

上述修改未经测试，各位可以自行尝试之后再做反馈，多谢。

15653910782 · 发表于 2022-3-2 19:19:25

顶一个，大佬，有没有TXT的修改方法

aliang227 · 发表于 2022-3-2 22:53:36

15653910782 发表于 2022-3-2 19:19
顶一个，大佬，有没有TXT的修改方法

TXT修改比较困难，不过可以退而求其次只把触发器的前置条件加上，也就是有人逃跑的时候才做后续计算，那时候战场人数也不多了，卡顿应该可以缓解不少。

nimoadna · 发表于 2022-3-11 19:26:30

楼主辛苦，很赞！

武安apk43 · 发表于 2022-3-13 00:12:43

大佬威武！来自: Android客户端

mikulalala · 发表于 2022-3-13 21:22:41

楼主大大，我玩的时候会很规律的卡顿，我感觉就是这个问题，可不可以请教一下我自己怎么修改呢，5900x结果玩潘德只能开240人。

mikulalala · 发表于 2022-3-13 22:15:30

mikulalala 发表于 2022-3-13 21:22
楼主大大，我玩的时候会很规律的卡顿，我感觉就是这个问题，可不可以请教一下我自己怎么修改呢，5900x结果 ...

我不会制作mod。。不知道怎么修改mod的代码。

aliang227 · 发表于 2022-3-13 22:29:02

mikulalala 发表于 2022-3-13 21:22
楼主大大，我玩的时候会很规律的卡顿，我感觉就是这个问题，可不可以请教一下我自己怎么修改呢，5900x结果 ...

具体表现形式就是每3秒会顿卡一次，自己改的话可以简化一下，只是把这个触发器

(3, 0, 0, [ (call_script, "script_apply_effect_of_other_people_on_courage_scores"),], [ ]),

复制代码

替换为

(3, 0, 0, [
(assign,":rout_num",0),
(try_for_agents, ":agent"),
(eq,":rout_num",0),
(agent_is_human, ":agent"),
(agent_is_alive, ":agent"),
(neg|agent_is_routed,":agent"), #这句估计都可以不要？
(agent_slot_eq, ":agent", slot_agent_is_running_away,1),
(assign,":rout_num",1),
(try_end),
(eq,":rout_num",1),
], [(call_script, "script_apply_effect_of_other_people_on_courage_scores"),]),

复制代码

可能都能缓解不少。按说拿原版的ms编译后再比对下TXT应该就行了，但我机器码看的实在眼花，看有没有擅长做TXT修改的尝试下吧。

mikulalala · 发表于 2022-3-14 17:35:59

aliang227 发表于 2022-3-13 22:29
具体表现形式就是每3秒会顿卡一次，自己改的话可以简化一下，只是把这个触发器

感谢回复，我应该在哪里找到并修改这些代码？我只有一点点C#的基础，昨天翻找mod文件也没有发现有类似这种代码的文件。

幼稚园殺手 · 发表于 2022-3-14 20:24:31

mikulalala 发表于 2022-3-14 17:35
感谢回复，我应该在哪里找到并修改这些代码？我只有一点点C#的基础，昨天翻找mod文件也没有发现有类似这 ...

Module_system里，一般Mod都不开源，所以找不到这些源码文件，但是你可以修改战团的，只需要下载战团的源码就行了

huagao · 发表于 2022-3-14 20:29:19

15653910782 发表于 2022-3-2 19:19
顶一个，大佬，有没有TXT的修改方法

emmm，是要改哪个mod呢

huagao · 发表于 2022-3-14 20:38:34

15653910782 发表于 2022-3-2 19:19
顶一个，大佬，有没有TXT的修改方法

如果只是战团native的话可以自己修改就行了，如果是某个mod的话，因为不同mod的参数id可能会不一样所以得告知一下目标。

mikulalala · 发表于 2022-3-14 20:49:04

幼稚园殺手发表于 2022-3-14 20:24
Module_system里，一般Mod都不开源，所以找不到这些源码文件，但是你可以修改战团的，只需要下载战团的源 ...

请问有教程吗，下载了源码，修改好了具体又是怎么应用到游戏里的？

mikulalala · 发表于 2022-3-14 21:14:29

huagao 发表于 2022-3-14 20:38
如果只是战团native的话可以自己修改就行了，如果是某个mod的话，因为不同mod的参数id可能会不一样所以得 ...

我想改潘德

mikulalala · 发表于 2022-3-14 21:48:01

刚才我在维京mod里开了一把400人的自定义战斗就没有这种卡顿、

幼稚园殺手 · 发表于 2022-3-15 07:22:14

mikulalala 发表于 2022-3-14 21:14
我想改潘德

首先你不要三连，其次这些MOD都是闭源的，所以源代码你是找不到的，你只能找战团的ms，战团的MS在制作区置顶有

chunbo3321 · 发表于 2022-3-15 20:13:23

TXT能改么？

djman · 发表于 2022-3-20 09:46:30

本帖最后由 djman 于 2022-3-20 10:02 编辑

战团还有很多可以优化的地方，上次看到一个提高加载速度的帖子

君悦大大牛逼 · 发表于 2022-3-22 16:31:52

这个可以直接批量编译到潘德里面吗？我用的modsystem1171

君悦大大牛逼 · 发表于 2022-3-22 18:04:11

试了一下，发现native400人以上可以流畅运行，但是潘德还是一样3秒卡一次。。（用的旧档，因为潘德没有自定义战斗）

野蛮人y · 发表于 2022-3-28 18:26:26

原script旨意是：
# 跑的降未跑的2士气反之，不跑的加跑的1士气  （平衡士气）
# 跑的降跑的1士气

在第一个try_for_agent 加个限制条件"跑的" 筛选就行，不需要用数组记录一遍逃跑的，再提出来循环，一样要套循环;
逃跑的对不逃跑的影响只有两种，除了玩家（这种情况不存在），英雄4士气普通2士气，
你新组的script跑的对跑的消极影响情况那里并没有乘以-1，后面又用的val_add
      (try_begin),#agent is not running
         (neq,":agent_is_running_away_or_not",1),
         (assign,":centered_agent_courage_score_add",":agent_delta_courage_score_pos"),
         (store_mul,":agent_courage_score_add",":centered_agent_delta_courage_score",-2),
      (else_try),#agent is running
         (assign,":centered_agent_courage_score_add",":agent_delta_courage_score_neg"),
         (assign,":agent_courage_score_add",0),
      (try_end),
      ……
      #set center agent score
      (val_add,":centered_agent_courage_score",":centered_agent_courage_score_add"),

aliang227 · 发表于 2022-3-28 20:52:30

野蛮人y 发表于 2022-3-28 18:26
原script旨意是：
# 跑的降未跑的2士气反之，不跑的加跑的1士气（平衡士气）
# 跑的降跑的1士气

第一个问题，不知道你是指的原来的script还是我新加的，如果是指原来的话，只把第一个循环限制为逃跑的应该行不通，因为避开了这个情况：
3）中心agent没逃跑，周围agent在逃跑：减少两倍中心agent的相应勇气值。

如果指新加的话，既然是优化，当然还是得尽量彻底，首先mt里的循环既然发现了，就可以直接提出来作为较为可控的循环范围。try_for_agents这个循环效率很低，即使加个“跑的”，仍免不了判断是否为human，是否活着，再判断逃跑与否的slot，就算放在第一句，循环也会把全部agent过一遍。我这里也只是提供一种思路，很多附带战场ai控制的MOD可能就是这类循环写的过多造成的人多就会卡顿的现象。

第二个问题，确实是我疏忽了，在赋值的时候

(assign,":agent_delta_courage_score_pos",":agent_delta_courage_score"),

(assign,":agent_delta_courage_score_neg",":agent_delta_courage_score"),
写成这样应该更合适，可读性更好一些
(store_mul,":agent_delta_courage_score_pos",":agent_delta_courage_score",1),

(store_mul,":agent_delta_courage_score_neg",":agent_delta_courage_score",-1),

命名的时候加了个"_neg" 就以为负过了，惭愧，感谢提出，已经重新修改过了。

aliang227 · 发表于 2022-3-28 20:59:03

君悦大大牛逼发表于 2022-3-22 18:04
试了一下，发现native400人以上可以流畅运行，但是潘德还是一样3秒卡一次。。（用的旧档，因为潘德没有自定 ...

native的ms应该不能编译到未开源的mod里头吧

而且自定义战斗测试应该和游戏战场是没关系的，因为用的不是一个mt，没有士气之类的设定，所以本身应该就不会发生这种顿卡

韩其泰 · 发表于 2022-3-29 09:41:56

大佬请问一下，这些是放在哪些文件里面的，我有源代码文件

天生就像珍珠贝 · 发表于 2022-4-1 16:16:53

原来是这个原因啊，感谢大佬，可惜汉匈决战没有源代码

天生就像珍珠贝 · 发表于 2022-4-1 16:17:12

大佬们，攻城的时候怎么不进入城镇和府衙场景，一次性打完

貫一 · 发表于 2022-4-14 17:58:28

本帖最后由貫一于 2022-4-14 18:03 编辑

示例
slot_item_temp_slot = 236 #
slot_agent_temp_slot = 237 #
#236,237，是我自己用的数值，可以用其它数值。前提是这数值是module_constants.py里不重复的，没被使用过的数值，理论上数值越小越好，前提是不重复。

我一个外行人不太懂，但是感觉是楼主在写解决战场卡顿的代码，不是对于大地图卡顿。

chunbo3210 · 发表于 2022-4-16 15:06:21

看来txt修改貌似比较麻烦呐

chunbo3321 · 发表于 2022-4-22 23:56:17

蹲一蹲看哪位大神会用TXT修改这个

by永远的 · 发表于 2022-4-28 03:04:17

凉大，为什么人一多就容易闪退

		自动登录	找回密码
密码			注册(Register!)

[版主推荐] 战团中战场人数调大后，发生定时卡顿的代码优化（全MOD通用）[更新TXT修改方法]]

评分

鲜花鸡蛋

评分

鲜花鸡蛋

鲜花鸡蛋

鲜花鸡蛋

维京征服正版勋章

骑砍中文站APP会员勋章

骑士美德之英勇勋章[杰出会员活跃勋章]

霸主正版勋章

元老骑士勋章

战团正版勋章

原版正版勋章

火与剑正版勋章

拿破仑正版勋章

剧本设计大奖赛

骑友真人秀Ⅷ勋章

骑友真人秀Ⅶ勋章

骑友真人秀Ⅸ勋章

汉匈决战正版勋章