Official KaM Remake Ideas topic

[pawel95]

There was a discussion about it between Da Revolution and Lewin here somewhere, but I couldn´t find it anywhere So I just wanted to push it.

"Please let the serfs bring wood first to the buildings, then stone".

This was implemented (accidently) already in an older KaM Remake Version. Now I wanted to ask for it, because I quite often see in mp games, that me and my ally is building the schoolhouse for example, same as me. However his serfs have a bad day and bring first all stones to the house, then the wood I think because of that, there are some bigger differences watchable in the early game.

So the idea was, to let serfs bring everytime wood first to a hosue, then stones. EXCEPT when there is no more wood (Like in woodproblems in midgame).


I talked with Krom already about it, he thought it could be cool. So I just wanted to push that "forgotten" idea from Da Revo again

[Killer!!]

but later at the game it could be a problem, because if you have to much serfssome will deliver the timber to the building and others just don't do anything because the timber isn't tjere at that moment

[pawel95]

but later at the game it could be a problem, because if you have to much serfssome will deliver the timber to the building and others just don't do anything because the timber isn't tjere at that moment

LoL? Where is the problem? It just says "First wood then stone" that means, when you have like 11 serfs and school costs like 6 wood and 5 stones, all serfs should start to move at one time

Sorry when that wasn´t clear. Only when you have like 5 serfs(That´s why i said, that most problem are the first min of the game) then, they sometimes bring like 5 stones first

[Lewin]

Ok it's implemented: "Please let the serfs bring wood first to the buildings, then stone". Thanks for reminding me.

[Nissarin]

BTW: some time ago we had similar discussion regarding serfs bringing construction materials and "depositing" them not only at the "door" but also next to it (in case builder is blocking it) just like in the original.

[Ben]

Thank you Nissarin! This really needs to be implemented. I've been screwed over so many times because of this missing feature.

I previously posted a more detailed post here: viewtopic.php?f=26&t=1017&p=23760#p23760

[Lewin]

BTW: some time ago we had similar discussion regarding serfs bringing construction materials and "depositing" them not only at the "door" but also next to it (in case builder is blocking it) just like in the original.

We haven't implemented that (yet) because it's not that simple to do. There's no other place in the game where you need a unit to walk to with 1 tile of another tile that is not walkable (the doorway). It would require special cases in our pathfinding and walking code. A long time ago we used to support this feature, but it was removed when we rewrote our walking/pathfinding code to fix bugs and add other more important features.

We might implement it eventually but I'd say it's low priority given that the current system is fine most of the time, it would be non-trivial to implement, and special cases like this are often the cause of bugs.

[Ben]

How about not letting builders build in front of a door? Would that be easier to implement?
Though this could cause some buildings, like iron mines and gold mines, to become very difficult to build...

[Jeronimo]

How about not letting builders build in front of a door? Would that be easier to implement?
Though this could cause some buildings, like iron mines and gold mines, to become very difficult to build...

This would be great because is notorious when you delay 20 seconds! at beginning while building school and cannot receive immediate timber.
Also happens when going to build INN (old order). These delays are the terror of perfectionists.

[Ben]

True, but Lewin has implemented serfs bringing timber first, so it won't be nearly as bad now. Still, I cringe so bad when I see my builders stop working because a serf couldn't deliver the last ware. This is especially annoying and devastating when all the labourers run off and start building something else!

[pawel95]

Ok it's implemented: "Please let the serfs bring wood first to the buildings, then stone". Thanks for reminding me.

Thank you Should make the first minutes in game way cooler

[Nissarin]

Just a little thing - can you implement a "Save the replay" button on score screen ? A lot of players save the game before they leave (autosaves are troublesome because they get overwritten), this would be far more convenient (and should be easy to do) especially if the file gets some meaningful name (like MapName-date) automatically.

[cmowla]

To Lewin and Krom,

Sorry it took so long for me to write this, but I made a small list of improvements for the AI, some of which might be "easy" to implement or modify. These are only suggestions that would improve the AI and are not requests or demands in any way.

• Incorporate storm attacks in the units that possess this option to have more effective pursuits against human player decoy troops.
• For town defense, currently if an AI troop group is moving to a specific position (for example, to replace a group which has just been lured out of the village or has died from hunger), it will not respond to any attacks on the village or villagers until it has reached its destination. Make it so that they are not bound to reach their assigned position when their village is being attacked.
• When a human player decoy is killed, make the AI more sensitive so that if there are any human player troops in their "radius line of sight", that is, the same distance they have to detect serfs being killed in their village by snipers, they pursue them if these human player troops are closer to the AI's post locations than they (the AI troops) are.
• When pursuing a decoy or fleeing army, set a time limit on how long they pursue him. The greater the distance between the human player decoy and the AI, the shorter the time limit should be dynamically adjusted to. If the distance decreases after the AI are programmed to return home (the human decoy turns around and goes back towards the AI village or the returning AI troops), make the AI again pursue until a certain distance is made and a certain amount of time has passed. Obviously the ultimate time limit for such a pursuit pattern is when the AI troops get hungry.
  
  Lastly, with this, make all AI groups independent. That is, xbows and swords should return home faster than knights, for example.
• When pursuing a human player decoy and other human player units come within their "line of sight", have them stop and attack them and forget about the decoy and attack them instead.
• This one might be something for the future, but for town defense, have the AI detect which type of units are attacking and send the rival troop type to kill them (e.g. if you attack with knights, the AI should attack you with pikes if it has any nearby). In fact, perhaps the AI should rotate its troops accordingly before you even attack their town...when you get within a certain distance, the AI will rotate its troops so that the opening which your troops are nearest will have to face the rival troop type.

And, as I mentioned in another thread, the AI villages should increase their army's size. I played TSK04 for 4.5 hours, but the number of AI troops after that much development time = the initial army of AI troops at the beginning of the game.

[Gandalf]

This is purely an idea for quite a way in the future as I know you're not keen on adding new buildings.

This is an idea taken from what you can do in other games. But in other games its considered more as an "upgrade"

The option to build a double building. E.g. A larger farmhouse that houses 2 farmers. Rather than bulding 2 seperate farms. A larger stonemason, housing 2 stonemasons. Things like that.

I can fully understand if you wouldnt want to do this. Simply just an idea.

Cheers

[Krom]

@Cmowla: Thanks for your ideas. AI is rather complex and complicated thing, we already put a lot of effort into it to discover even simplest ideas take a whole lot of coding and planning behind. It may look easy from outside, but in fact it is very hard to "teach" AI to pick a good spot for a woodcutter or make a decision about building another gold mine. These are static examples. Now army control is an order of magnitude harder. Even finding an algo to set up defense positions took me months and still you can see how imperfect that is at the moment. Said that, we are grateful for your ideas, but at the moment we don't have good enough AI operating with high enough level of consciousness to "teach" these rules to.

@Gandalf: I think KaM is fine with it's granularity level. Having double-something won't add to the game.