- 好友
- 1
- 在线时间
- 0 小时
- 最后登录
- 2026-2-1
随仆
- UID
- 3840361
- 第纳尔
- 0
- 精华
- 0
- 互助
- 0
- 荣誉
- 0
- 贡献
- 0
- 魅力
- 0
- 注册时间
- 2026-1-13
 鲜花( 0)  鸡蛋( 0)
|
发表于 2026-1-14 22:34:36
|
显示全部楼层
I noticed something similar and solved it differently than the original post for v162 Phantasy Calradia "2025" / "Chiaroscuro steam workshop (somewhat unpublished), which then became Paradigm Chronomancer.
My source code is published as OSP but all my mods in general are labelled "LSP" because all the artwork came from OSP kits which have conditions on their use - non commercial and thank them in credits.
The quick answer is I set both a legacy mode as was and a modified mode where instead of try_for_agent type checks to see who was nearby, tallied all the injuries or kills from the last 7 seconds and then applied a blanket adjust to each side, reset my counters and continued, making sure to flush before any reinforcement wave. This gave far less bandwidth consumed at the cost of local pockets of runners -- since I did not know where any pockets were, results applied to the whole team at middle distance. This led to an all-or-nothing sort of response, but I offset that by having a more detailed scheme for assigning courage at agent spawn so that not everyone would panic at the same time and veterans would panic much less than farmers and raw low tier recruits.
I won't burn a half day collecting code snippets and splicing them together, but do notice the interplay between module_mission_templates and formAI* and formations*
I also changed handling for how to auto-assign ranged versus cavalry versus melee if that sort of formations tweaking is of use to anyone. The thing is -- my mods are super huge, and probably not the stablest ever, as I am fond of quick experiments rather than just doing something once and never touching it again. Thus I have weekly mod updates, when I can think clearly. You'll find them at Steam workshop, as I detest Moddb when I update weekly or even every 3 days.
The original coding here is good, just there are more ways to solve the problem and having a switch users could set let them stopwatch the difference under similar conditions.
As for people saying they had "x" number of agents on their side fighting "x" on the other in two different mods, different mods have a different number of middle ware timer driven routines. I have quite a few firing to handle magic or effects like blindness or damage or heal over time, while some other mod might have nothing there and go "faster" supporting more agents. Also don't forget every horse is ANOTHER agent, during try_for_agents. I think do fairly well at spreading my CPU use across a spectrum of time, so it SEEMS like I am able to support a little more. I also modify my reinforcements code, which might be frustrating to follow.
But I am not the best nor most popular modder here. There is always something nice to be said for simplicity.
I was mostly looking for different answers than ordinary in my experiments -- and for me mods are all experiments. I really don't care about pictures -- it is IDEAS that are immortal. Go ahead and think Warband is dead; to me I find nothing interesting in Banterlord. But that's just me. Technically, it SHOULD be superior. Except it has yet to be so, in my eyes. |
|
[复制链接]
显身卡
