abc518448 abc518448
关注数: 90 粉丝数: 231 发帖数: 31,471 关注贴吧数: 116
【更新公告】Balance Patch – 0.10.1386(2016-02-08) 公告链接:http://tieba.baidu.com/mo/q/checkurl?url=http%3A%2F%2Frobocraftgame.com%2Fbalance-patch-0-10-1386%2F&urlrefer=8dde5e03327c2bd4bdc977f984ddec85 翻译及原文 BALANCE 平衡: LoML 导弹 →Lock-on time set to 2.5 seconds (up from 2.0) 锁定时间变为2.5秒(原2秒) →Ammo bar usage increased to 19.16 (up from 18 | It still shoots six missiles per volley) (???)弹药条增加到19.16(原18)(满弹夹一轮仍然是6发) →Protonivity decreased to 2.9 (down from 3.5) (反)质子素率降低到2.9(原3.5) →Robot Ranking increased to 15,000 (up from 8,250) RR增加到15K(原8.25K) →Weight increased to 1,350 kg (up from 34.7 kg) 重量怎级啊到1350千克(原34.7千克) Aeroflak 防空炮 →Projectile speed doubled 炮弹飞行速度(弹道速度)翻倍 →No longer suffers from lowered accuracy while moving 移动时精度不再降低 →Health increased to 108,540 (up from 62,520) 生命值增加到108540(原62520血)(译注:提高约75%,现在生命值-CPU效率是361.8了) →Firerate increased by 20% to 0.625s, 0.525s, 0.51s, 0.5s (up from 0.75s, 0.63s, 0.606s, 0.6s) 射速提高20%,射击间隔(1~4把)变为:0.625秒,0.525秒,0.51秒,0.5秒(原0.75~0.6秒) →Protonivity reduced to 1.75 (down from 2.0) (反)质子素率降低到1.75(原2) →Weight decreased to 2,250kg (down from 4,500 kg) 重量降低到2250千克(原4500千克)(译注:降低一半) Plasma →Damage increased for most Plasma weapons: 大多数电浆炮伤害提高:Protonivity decreased to 1.84 across the board (down from 2.4) (反)质子素率一律降低到1.84(原2.4)(译注:虽说是across the board,然而根据下文,不包括Mega炮) Mega plasma protonivity decreased to 1.3 (down from 1.7) Mega电浆炮(反质子素率)降低到1.3(原1.7) PS:弹药条指的不会是这个吧:右下角的那一圈……
【简要分析】关于氢推续航以及氢气储存 最大推力时每秒氢气消耗: 大船大氢推:约6425单位(wiki:6426.7) 小船大氢推:约514单位(wiki:514.1) 两者消耗差距约12.5倍 大船小氢推:1092.5单位 小船小氢推:109.2单位 两者消耗差距约10倍 大船氢罐容量为250万单位,重8161.6千克,体积421,875升(能耗1千瓦,待机1瓦) 质量效率约为306.3单位每千克,体积效率约为5.93单位每升 小船氢罐容量为4万单位,重3161.6千克,体积15,625升(能耗1千瓦,待机1瓦) 质量效率约为12.7单位每千克,体积效率为2.56单位每升 (不过氢罐不一定装满,所以效率实际上要低) 1千克冰可由制氧机转化为9单位氢,而1千克冰的体积是0.37升 质量效率为9单位每千克,体积效率约为24.3单位每升(不过冰需要储存,所以容器质量也应计入质量效率和体积效率中,不过容器还没有确定) 大船制氧机每秒制氢501单位,能耗330千瓦,待机1千瓦(可额外储冰4000升) 小船制氧机每秒制氢166单位,能耗100千瓦,待机1千瓦(可额外储冰1000升) 列了这么多数据,来假定一个情景,一艘大船有10个大氢推,如果他需要3600秒的续航时间,那么在使用纯氢罐、纯冰储存、冰-氢混合储存的情况下,那个更优秀呢(不计物流管道) 首先需要消耗:231,300,000单位氢气,即2亿3130万单位氢气 氢罐储存:需要92.52个大船氢罐,向上取整就是93个大船氢罐,体积39,234,375升(约3923万升),质量759,028.8千克(约759吨) 冰储存:需要25,700,000千克冰,也就是2万5700吨冰……体积9,509,000升(约951万升),并且,为了满足每秒64,250单位的氢气需求,需要再安装128.2个制氧机,向上取整129个,一个制氧机2587千克,体积31,250升,129个就是333,723千克(约334吨)和4,031,250升(约403万升),同时这129个制氧机每个提供了4000升储存冰的空间,所以总共提供516,000升(51万6千升),实际货舱只需要存8,993,000升冰,用体积质量比最高的大货舱装,需要21.32个,向上取整就是22个,一个大货舱重2593.6千克,那么22个就是57,059.2千克,约57吨,但是实际体积变成了9,281,250升(约928万升),总共统计下来,用冰储存的话,整套质量是26,090,782.2千克(约26091吨),体积是13,312,500升(约1331万升) (实际上货舱的空间是可以被填满的,所以其实货舱浪费的体积可以不用在意,同样续航数字也不会定的那么死,所以所有向上取整的浪费也可以不用在意) (PS:然而冰在转化过程中质量会不断减少,如果按我的设定,每秒要转化7140千克的冰,也就是每秒转化7吨的冰,所以到后期冰的质量就不会那么悲剧了) 氢罐储存明显有着超高的质量优势,足足34.4倍的优势!(然而冰的质量不断减少)体积上,氢罐储存就远不如冰了,氢罐使用的体积是冰储存的2.95倍!很显然氢罐储存最适合星球起降、非战斗船只,而纯冰储存就是鸡肋,但是对于战斗船只来讲,氢罐的体积还是有点大,换成立方体就是34米的边长,而如果用3*5*6+3的方式排布的话,大方阵长45米,宽37.5米,深22.5米,也是太大(请忽略这段分析……因为本身对氢气船要求一个小时的续航就太过分了……氢气罐体积太大也是没办法) 所以如果要体积折中,质量折中,而且有一定中途自己采矿补给能力的话,还是要采取混合储存的方法,因为组合方式太多,我就不列举了,可以自己试试那种方式效率最高(建议使用程序或表格辅助计算)
【脑洞】纯属脑洞,切勿当真 从Mark的想法可以推断,Mark想让各种大小的武器和移动方式更加多样化,我沿此思路继续完善。 首先要引入一个协调控制系统,具体实现见后。 最大部件称为T5,最小部件称为T1 通用:越大的血量越多,屏障保护率越高,CPU消耗越大 武器 改变武器射击模式: 所有武器:所有武器独立装填进度,无武器生效数量上限 SMG:双击左键齐射,长按左键进行协调射击(协调射击:当所有枪依次射击时,如射击间隔大于等于0.2秒,则以射击间隔进行单次单发射击,若小于0.2秒,则通过协调改变单轮射击中单次射击的射击发数,使每次射击间隔等于0.2秒)。 电浆:双击左键齐射,长按左键进行协调射击(电浆协调射击时协调间隔为0.8秒),右键开镜时可使用鼠标滚轮调节装填速度和弹道速度(2者相乘为定值)。 磁轨:只有单发射击,射击间隔分为冷却和重新加载2部分,当单门炮射击后,所有炮进入冷却时间,并且该门炮进入装填。 性能特点: SMG:越小的机枪,弹道速度越慢,射速越快,同时稳定性越好(单发扩圈系数低),缩圈越快,但是最大散布高,伤害越低,反之亦然。 电浆:越小的电浆,其最大和最小弹道速度越大/小,伤害越低,重装时间越短,反之亦然。 磁轨:越小的磁轨,弹道速度越快,伤害越低,射程越低,扩圈系数越低,缩圈越快,反之亦然。 平衡性: 1)同类对比:(均为平滑过渡) SMG:T1机枪的理论DPS是T5机枪的1.5倍,射速是T5机枪的15倍。 电浆:T1电浆的理论DPS是T5电浆的1.8倍,射速是T5电浆的9倍。 磁轨:T1磁轨的理论DPS是T5磁轨的2倍,射速是T5磁轨的8倍。 2)不同类对比: 同T中,2.5倍磁轨DPS=1.6倍磁轨DPS=1倍机枪DPS 移动方式: 机翼:越大的升力越大,阻力也越大。 悬浮、旋翼:越大的升力越高,最大功率越大,最大动力越大,阻力越大。 轮:越大的承载能力越大,转速越慢(限速),地面摩擦系数(抓地力)越大。 蜘蛛腿:越大的承载能力越大,运动速度越慢(限速)。 履带、机械腿:越大的动力越强,速度越快(限速),但是会很重(小的会变的较轻)。 推进器:越大最大功率越大,最大推力越大。 新的阻力限速方式:(依然不物理,不科学) 推进器和悬浮、旋翼将会有功率和推力这2个属性: 最大推力决定了起步加速的速度,最大功率决定了在一定速度时推力的大小。 推力单位:kN 功率单位:mph*kN 同时悬浮、旋翼、机翼会有阻力属性: 阻力决定了平均加速度,并与功率决定了最大速度 阻力单位:kN/mph 比如,某载具有6个机翼,机翼阻力系数是0.5kN/mph 有10个推进器,推进器最大推力都是2000kN,功率是120000kN*mph 那么,速度每增加1mph,阻力增加3kN 计算极限速度(平飞) 设极速为x 平衡是阻力=推力 阻力=x*3(单位均为kN,省略不写) 推力=120000/x x*3=120000/x x*x*3=120000 3x方=120000 x方=40000 x=200 平飞极速=200mph 当然了……物理系统我觉得还是重做一下比较好……不然现在推力太不科学
【Mark】作死第四章 http://tieba.baidu.com/mo/q/checkurl?url=http%3A%2F%2Frobocraftgame.com%2Fforums%2Ftopic%2Frealising-the-vision-chapter-4-the-vision-revised%2F&urlrefer=ac4292d599f292bb03521a49b59df6ee First, a huge thank you to all your passionate feedback. With that amount passion from our core fans we simply had to take a huge step back and consider whether we were doing the right thing. We’ve spent a couple of days reading through literally every post on the forum on this subject and in other areas like Steam discussions, Reddit, Twitter etc. and brainstorming a complete rework of the plans. Our goals remain the same. - To maximize creative options – but not at the cost of removing all the depth in the combat engine that users love - To allow all skill levels of players from casual to Invictus Rex to enjoy Robocraft - To increase the % of players who install and then become engaged Robocraft players, to grow our community So, here are our plans. A complete revisit of Chapter 3, we hope you like it… - The damage model will not change, it will remain as it is now, cubes and functionals falling off as they do now - We want to rework all the weapons to remove hitscan, add proper area effect, and improve Nano and Rail, more on that in a later Chapter - We will remove the Pilot Seat, you will be dead when only a % of your Robot remains NB: For those of you who were looking forward to SCU’s, we still plan to add features like their abilities in the form of ‘modules’ that have a CPU cost - We also want to significantly increase the CPU costs of functional parts, we believe this change can improve the game significantly and make building design deeper - At the same time we will increase the total CPU you can spend at all Player Levels, you will be able to fit more Chassis cubes but less functional cubes - Redundancy is not a bad thing if you have to make a hard choice over having redundancy, more damage, more speed, or more health - With CPU costs being higher we feel we can add in a Weapon Select feature, since having more than one type of weapon on your – Robot will be a tougher choice, since you will sacrifice redundancy or health or speed for flexibility - This should make choices more intuitive for new players too as it’s a classic game choice of how to I spend my resources, they can learn about tri-forcing later - We also plan to add in an improved match-making algorithm that puts less skilled builders and fighters together and more skilled builders and fighters together - We will be adding Paint Tools to the game along with more shapes, I will cover the Paint system in more detail in a later Chapter I want to talk to you about how we plan to add Tutorials to the game and a much improved progression system designed to help new users learn the game once they install, but I will do that in a later Chapter also (as previously planned). Finally, I want to explain to you that we are working hard to fix many bugs, Rudder Bug, Falling through terrain, Tower glitch, memory leaks, and lag, and many more. Our next launch will also bring in the release of our Unity 5 engine which initially we hope will improve performance by using multiple cores of your PC, and later will allow us to improve the physics engine so that really physical interactions such as realistic collisions in battle can be added to the game. There is even more to tell you, but for now I will close out as this is already a wall of text. 翻译稍后
对白虎的评价 我不对作者作任何评论,单说载具: 首先我觉得定位有些模糊 1.说偷袭爆发吧,MkX枪1把+7把Mk10枪什么鬼,这不是自己降低自己DPS么,而且极速才230……我记得我的650块中悬和牛盾的重滑板都能飚255+ 2.说正面对射吧,感觉不太实际啊,过于脆皮 3.说防空支援吧,装底枪防空输出大降,侧枪支援的时候还得转向,而且这个载具面对炸逼过于脆弱 第二呢,我认为你所说的优异性能有部分失实 1.首先这游戏不要叫防御力,那叫续航,而续航对布局和装甲量都有要求,你这种布局,明显和你的偏向不符,在正面堆甲,而完全忽略底部和顶部防御,很有可能被人抓住机会一轮带走,别跟我说机动优势,2个人群殴你你面向谁?而且所谓的10层甲,我看不见得有效,敌人的火力不总是和你在同一平面上的,有高度差就会让你的防御被轻松绕过 第二攻击力的话,虽说MkX枪输出是不错,但是火力徐航也很重要,当你被对面袭击的时候,很可能你会丧失部分战斗力,事情不会像想象中的那样总是单挑,很有可能车轮战甚至群殴,而你的机动虽然有优势,然而你不会飞,也就是说你无法随时回基地回血 2.平衡力……现在悬浮你只要搞好升心质心位置,基本就是不倒翁…… 3.敏捷,这玩意呢其实大部分载具的敏捷没有太大差距,只是有着极速差距,说到机动,230的极速并不是很优越,爬85度坡也不是很突出的性能,所以优势不大,不过敏捷性对手感有很大影响,这载具个人认为开起来应该挺爽 @DemonHunter鹏
论星球上重力炮的效果(原版情况) 在1g重力环境下! 1.重力炮防空 效果会非常的差,就以自动炮台寻敌距离的800m来说 如果初速度为104m/s,那么800m的速度用能量观点计算就是: v为初速度,v0为800m速度 E初=1/2mv^2 E800m=1/2mv0^2+mgh 二者相等 则m*1/2v^2=m*(1/2v0^2+gh) 1/2v0^2=1/2v^2-gh v0^2=v^2-2gh 带入得 v0^2=-4,880 ……好吧,这……他都打不到800m 也可以用匀加速运动的一元二次方程 s=vt+1/2gt^2 求根公式得:(104±√10816-15696)/9.81 还是一样的悲剧 所以重力炮最大射高(仰角90度)为:H=v^2/2g【能量守恒嘛,v1^2=2gh】 即551.27421m 2.重力迫击炮 无高度差最大射程是防空炮射高的一倍(1,102.54842m) 有高度差情况比较复杂,但是因为SE神奇的限速(假设为104m/s),所以比较容易计算 (在初速=限速时的情况!) 分成2部分 v为初速,α为抛射角(α0为变量),h为高度差,t0为无高度差部分时间,t1为高度差部分时间,v0为限速(104),v1为水平方向速度分量(变量) 一: s0=v*cos(α)*t0 又t0=2v*sin(α)/g s0=v*cos(α)*2v*sin(α)/g 二: s1=v1*t1 t1=2h/(v*sin(α)+v0)【就是高度差除以平均速度,因为在有限速情况下,当下坠速度达到限速时,其他速度分量为0,所以可视为最大时间】 sin(α0)=(v*sin(α)+gt1)/v0【这个比的值是不断增加的,最大为1】 (v=v0) 【而cos(α0)=√1-sin(α0)^2】 【也就是说,v1/v0=√1-[(v*sin(α)+gt1)/v0]^2】 【所以v1={√1-[(v*sin(α)+gt1)/v0]^2}/v0】 【v,v0,g,α是已知量】 很显然,这是个非匀变速运动…… 而通过游戏机制得知……这游戏每秒计算60次速度 所以,通过简单的重复计算可求得s1 进而求出s0+s1的最大值…… 结论:任何在飞行过程中重力与初速度方向不相同或相近的动能武器,在星球上都是废品……
【Mark的研究】Graphics performance = retention issues 原文:http://tieba.baidu.com/mo/q/checkurl?url=http%3A%2F%2Frobocraftgame.com%2Fforums%2Ftopic%2Fgraphics-performance-retention-issues%2F&urlrefer=359c5a3ae0eb50b29f8e1cbc199c6a3e Hi, So, as you know we’ve been focusing on frame-rate and performance recently. The reason is that we noticed that our ’1 day retention’ (that is the % of users who install the game and then play again the next day) had fallen from a world class 48% to a low 32% around the point of the launch of Respawned and Overclocked. Naturally we assumed that it was the introduction of that game mode that caused the drop in retention, and various other changes we had made since so it took us a long time to see the ‘real’ reason. Once we’d AB tested the old and new game modes and found that the new game mode actually has better retention than the old game mode (slightly) it forced us to look at other things. Many of your were reporting a loss in frame-rate and whilst we were concerned about it we didn’t believe initially that frame-rate would cause such a significant drop in retention (and users). So, we added the ‘automatic GFX setting’ feature which scores your PC and sets the GFX setting automatically based on your score. This is supposed to ensure that you get a descent frame-rate from your first play – ideally you enjoy the game, get hooked and keep playing forever. Adding this automated setting and making the recent optimisations to the Mothership increased the 1 day retention from 32% to 40% overnight. This is like ‘wow’. I mean, you can introduce a whole new game mode that changes the game fundamentally from a deathmatch to a respawning game with Towers and stuff and it doesn’t make a difference to retention, but removing the planet outside the window and adding an automated GFX setting can make it jump by 8%. That is incredible. It gets more interesting. We also attached some analytics so that we can measured the different retentions of people with different quality PC’s. Here are the results: PC’s autoset to ‘Fastest’ – 51% played the next day PC’s autoset to ‘Fast’ – 41% played the next day PC’s autoset to ‘Normal’ – 33% played the next day PC’s autoset to ‘Good’ – 27% played the next day PC’s autoset to ‘Fantastic’ – 35% played the next day PC’s autoset to ‘Beautiful’ – 44% played the next day NB: What’s weird is that the crap PC’s and the Ninja PC’s have the best retention, and the ones in the middle are the worst. Our theory is that our automatic GFX setter is not grading the PC’s correctly in the middle and pushing them on too high setting too fast. So, we have 2 conclusions: A) We need to optimise the game more so everyone gets a good smooth experience B) We need to put the old Mothership back (or something cool like it) for people with good PC’s C) We need to adjust our automatic quality setter to make sure it doesn’t put mid range PC’s on too higher GFX settings We will be doing some more AB testing on the autosetter to see how grading the PC’s differently for new users affects this retention and i’ll report back. 这是Mark对图形质量和玩家留存率之间关系的研究……Mark表示自动设置工作良好,Mark表示会让高特效的人重新见到那个星球贴图
【重开贴】模块式构建复合装甲 禁止在除规定回复楼层外回复,谢谢配合 首先来了解复合装甲最基本的结构: ①基础装甲 1.斜面斜面拥有4个连接点,2个是相对的,2个是相邻的,全部使用相同方向的斜块构建的装甲块/板有2面具有全部连接点,4面具有一半连接点 2.斜角斜角拥有3个连接点,全部相邻,这一特性决定了复合装甲不能以斜角为主要基板,即使勉强而为,结构也很脆弱,并且强迫症不能用斜角作为中轴材料……又因为复合装甲要求有大量外连接点,所以一般作为内部特殊隔离使用,被广泛替代斜面使用在装甲节当中 这里有一种闲的蛋疼才用的纯斜角装甲板,实际效果=0,虽说可以无限嵌套虽说大规模嵌套之后强度不弱,但是这坑坑洼洼的外表……呵呵3.圆润缺角圆润缺角拥有5个连接点,缺失的连接点与放置连接点相对,利用缺失一个而且只缺失一个连接点的特性,往往可以很简单的构成装甲板 下面就是复合装甲中第二级的总成,规模仅次于装甲体: ②装甲板 1.波浪装甲板波浪装甲板是由2种方向的斜面交错编织而成,特性是具有3面完全连接点,1面无连接点,2面半连接点,这种特性导致波浪装甲板不能单独层叠成为装甲体 伤害传导特性图(吸收伤害特性)对比普通装甲:可以看见效果不明显,对于同样长度的普通装甲,对于吸收伤害没有多大帮助 但是波浪装甲得益于极大的传导优先级延长,因此具有优异的单向抗断性能,但是,某一方向的抗断性意味着另一方向更为脆弱 抗断实际模拟对照图:蓝色箭头表示对于垂直与该方向连接的断裂有较好的抵抗力,红色反之分析详见http://tieba.baidu.com/p/3798253448?pn=2 2.波浪装甲条—横竖斜面复合装甲节复合装甲板该种装甲板在内部舍弃了一些横向连接点(垂直对称轴方向),增加了外连接点,使所有面都有连接点,沿高度轴方向的连接点使该装甲板可以层叠,建议层叠方式:镜像层叠 伤害传导特性:类似与波浪装甲板,但是单方向抗断性不突出 注释:1)横向斜面:相对的连接点在左右的斜面 2)竖向斜面:相对的连接点在上下的斜面(在前后的为纵向) 3.波浪装甲条—竖—横三角装甲节复合装甲板与波浪装甲条—横竖斜面装甲条复合装甲板类似,但是并没有使用复合装甲节,而是交错使用了2种装甲节 (未完待续……)
【水】【译】收益机制(译自wiki) 首先,对于一个收益截图 比如:那么第一个是CPU DESTROYED:顾名思义即CPU摧毁数 第二个是 IN PROTECTION:你攻击正在攻击你队友的敌人时摧毁的CPU 第三个是 IN DEFENSE:你攻击靠近你方基地的敌人时摧毁的CPU 第四个是 KILLS:击杀数 第五个是 ASSISTS:协助数,打超过25%伤害然后被抢人头或5%血量治疗 第六个是 HEAL CPU:CPU治疗数 第七个是 ALLIES 100% HEALED:100%治疗数(治好数),队友的同一条命只能得一个 第八个是 SCOUTING:带雷达后,第一个(全队第一个用雷达侦查到的)用雷达侦查到的敌人数 第九个是 SPOTTING:按Q发现敌人(似乎可以重复获得)数 第十个是 PROTONIUM DESTROYED:摧毁质子素水晶数 第十一个是 TIER HANDICAP:残疾人低保(雾) 具体收益系数分为3个部分 1.基础系数 这个系数上限为1 基本上你只要参加战斗了就能拿到(甚至你受到伤害也有)获得的奖励次数乘以该奖励类型的系数就是这个基础系数了 2.奖励系数 这个系数没有上限 鼓励你参加激烈的战斗这个系数受到占领基地和侦查的影响更大,而且不会因为承受伤害而增加 注:Base Segments即占领基地的段数,就是TEAM DEATH模式顶上那个条,总共4段 3.delta scalar 不明…… 最终收益为: (基础系数+(奖励系数×delta scalar))×(胜利奖励或者英勇奖励)【不明】
【伤害传导】装甲解析 目前的传导机制是正6面体传导,但是依然遵守连接才会传导的原则,所以伤害导轨依然有效(甚至比以前更强了) 正投影差不多是这样的(伤害击中了1号),顺序是从小到大,在同优先级中随机选择,当全部同优先级都被摧毁选择下一优先级我们可以推导一下,假如击中的不是中间,而是边缘,那么将会是这样的:我们发现,如果要从边缘打断一个厚度为1,宽度为5的板子,我们最多需要破坏1+3+5+7+9个方块(最少1+3+5+7+1),也就是25个方块即5^2,也就是说,对于厚度为a,宽度为b的板子【设b>a】,那么我们最多需要破坏a*b^2个方块,可以看到,摧毁方块数与宽度的平方成正比,与厚度成正比,那么,一个最少方块的最难被打断的板子,一定有a=1,b趋近最大值 (如果是最少摧毁数,那么:最少破坏(b-1)^2*a+1个方块,也就是和宽度-1的平方成正比,和厚度成正比) 那么怎样才能更容易打断一块板子呢 当然是打中间,可以看到,打击一块a=1,b=5的板子,如果我们攻击边缘最多需要摧毁5^2个方块,而显而易见的,攻击中间只用摧毁1+4+8个方块(最少为1+4+2),也就是13个方块即(5+1)*3/2+(3+1)*2/2【也可以看做(5^2+1)/2】,对于一块厚度为a,宽度为b的板子,我们需要摧毁a*((b+1)/2*(b+1)/2+(b-1)/2*(b-1)/2)也就是a*(((b+1)/2)^2+((b-1)/2)^2),解出来得a*(b^2+1)/2,写成标准形式就是0.5*a*b^2+0.5,很显然与宽度的平方成正比,和厚度成正比,与之前结果相同,只是在击打正中间的情况下,需要摧毁的方块少了一半左右,所以磁轨童鞋们,打人打中间啊 以上就是在简单情况下的分析 接下来是复杂的分析 1)伤害导轨情况下的伤害传导
首页 1 2 3 4 下一页