Messages

Hey all,

Just wanted to ask real quick, what is your experience of very large maps in late-game with many colonists ect?

I have a fairly beefy CPU (2500K @4,4Ghz) but I don't know how badly performance degrades on large maps when there are loads of entities to track, so I'd love to hear your experiences. I like big maps with lots of options and room to do stuff, but it would be a shame if I find out 10 hours into a campaign that everything slows to a crawl because the system isn't really designed to handle these sizes (Tynan does warn of performance issues in the menu after all...)

Thanks for any feedback you can offer
-Stigma

Hey all,

I'm just looking for some feedback from people who have played the "lost tribe" scenario and have more experience than me.
It seems overall pretty hard, but I can deal with the extreme research penalty and I love the "build from basics" start and the larger group of core colonists.

The thing I keep having problems with however is recruiting new colonists.
It seems like everyone I capture has a 1% chance of converting even with a good warden and comfy cells - making it a very hard and long process to turn anyone.

I guess this is because you are "tribal" and thus you suffer the same problem in reverse that colonies have of recruiting tribals. The problem is this penalty is waaaay worse on this side. I think even other tribals are 1% chance if I remember correctly..?

It just seems like this is overly punishing in the long run and it makes it very hard to expand your colony. 5 people to start helps - but not for long when getting more is extremely difficult. It seems like you have to rely almost elusively on new colonist through random events or buying slaves... is this by design? If so it seems over-tuned. Having recruitment be harder fits with the tribals slow rampup, but the penalty is just so extreme that you could easily keep prisoners for years without converting them and that seems a little too extreme.

What are your thoughts on this, and your strategies for working around it?
I really like the tribal start, but this one issue just seems like it will be crippling in the long term as you REALLY can't afford to lose anyone, plus you will get starved on manpower real fast.

-Stigma

The game likely hasn't been made to scale properly with extra large amounts of entities yet. (I haven't really hit that limit yet in my testing, but I wouldn't be surprised if it starts chugging at some point). In a pre-alpha state the point is still just make sure that the program is stable at a basic level, and to implement the basics of the gameplay features.

No doubt it will be optimized later on to scale better for a large "endgame" scenario, but that is probably something that won't be addressed until there is at least some late-game content to actually test.

Also keep in mind that if you are running 40 colonists (is that possible even in phoebe friendly AI?) then that is probably already far outside what the scope of the game is. It seems like Ty's vision is to keep the (core) gameplay around a level of colonists where you can still keep track of individuals and identify with them, rather than managing an impersonal beehive of nameless drones.

Sounds like you have been playing the game in pre-alpha too much. Don't burn yourself out. Take a break and come back when the next big patch is out and you have more content to test. By the time there is enough content to warrant that sort of late-game level of entities I have no doubt that Ty will have improved performance. No point in playing this early version so much that you get bored of the whole game before it is even finished.

-Stigma

I think the basic idea is very simple and can use pretty much existing mechanics. You just have to write up some new encounter scripts that use a little different AIs and such.

The idea could be expanded on though if need be. Though it would require a lot more work, I don't see why you could not make "away missions" that dealt with that sort of thing. let's say an event tells you that raiders ahve set up a forward camp nearby, or perhaps even a neutral colony. By going to an assigned spot on the outskirts of your own map you could choose to leave the map temporarily to go to another map where you can engage in an attack for various objectives - elimination of all foes, raiding supplies ect. Plenty of opportunity there for both good and bad roleplay. You may want to raid that neutral colony for supplies, or just send someone to set up a trading relationship.

Actually there is a lot of fun possibilities there... you could for example use actual player maps and let the AI take over from there. To raid that colony you might have to overcome the defences that some other player actually set up, which could be fun - and it saves the developer from having to code procedurally generated bases that make sense. Just an idea. The game could simply harvest some anonymized data from the clients and upload them to some central server for use in other games as attack scenarios.

This is much more involved than the simple scenarios I mentioned before obviously, but perhaps worth considering

-Stigma

Hey all,

I could write up a whole article on this suggestion (maybe later) but for the initial post I just want to get the basic idea across as it is fairly simple.

Currently the combat of the game revolves around defending your base - and you defend it against pretty much the same threat each time - raiders. Amount of raiders and how good guns they have vary some, but that's about it. Hopefully variety in attacks will improve as the game development goes on, but here is a idea to add something different to the whole equation.

Rather than being all about defense (and often lots of turrets), add opportunities and incentives for going on the offensive. These would be combat objectives that require you to leave the safety of home (most likely involving elaborate chokepoints, lines of turrets and mines) to rely on just your colonists, weapons and your tactics. The idea is to add some pretty drastic variety to the combat this way by taking you out of your safety-zone. Here are some concrete scenarios that could happen:

- A small raider scout party appears and moves though the map (the AI makes sure that it does not wander into your defences for an easy kill). The objective is to eliminate the scout party before they can report back. Failure means a good chance of an especially large attack you have to deal with later, rather than this comparably smaller threat. Do you leave the safety of home to preemptively ward of the future attack?

- Local slave trader caravan appears and moves through the map (again, AI avoiding defenses as should be the case with pretty much all of these scenarios).  You can either let them pass, or attack them to potentially free slaves. (possibly add the option to trade with them also)

- non-hostile trade caravan passing through. Leave them alone, trade with them, or take on the role of the bad guys" to attack and steal whatever they are transporting - which could be anything from resources to guns.

There are dozens of variations here to work with just off the top of my head, and this opens up a lot of potential for both a of variation on combat scenarios, and also the opportunity to actually use the tactical elements of the game in a more meaningful fashion and get the feel of some squad-based combat to change the pace instead of it all feeling like "swarm wave defense". Also adds some meaningful choices both in terms of strategy but potentially also some moral choices to roleplay. if your colonists are close to starving, can you justify a "better you than me" attack on an otherwise friendly caravan?

Just something to consider for later additions to the "meat" of the gameplay
We certainly need some more different types of events than only raiders and crazy boomrats

-Stigma

And if you let players enter the seed, make it an alphanumeric seed, not just a number.

Letters+numbers are just a base 36 number that can easily be converted to decimal for internal use.

And if players want to exchange seeds (or only talk about them) then "Mud" is far more catchy than 17608.

You could play in a world created by your dog! (or it's name anyway)

Agreed - it's a very minor thing that "shouldn't matter because it's the same anyway" but it makes the whole thing a lot more human and catchy. It's easy to remember that the seed "pizza" was one that was very defensible and that you can suggest/recommend, rather than "509423990234235" which you definitely have to make an effort to record somewhere if you actually want to make use of it at a later time.

Fairly simple stuff to code the conversions too I should think. I'm only a novice coder, but even I should be able to handle that fairly easily - which probably means that a real codemonkey considers it trivial =P I would not be surprised if there exists copright-free code that does that sort of conversion too considering how generalized that sort of operation is, which would make it largely a copypasta job.

-Stigma

Hey all,

This is a really simple suggestion because it is basically the exact same as in other games. Minecraft is a good example.

I assume that the world is semi-randomly/procedurally generated. This means that there is some sort of random seed that it works from to construct the world. Let us see the random world seed so that we can recreate and share interesting worlds with other players.

This should be fairly easy to code since it just exposes something that already exists to the player, so it should just be a fairly minor change in the UI. It let's players who want to try something else but do it on the same world do so, and also creates nice community interactions in terms of "challanges" on especially hard world seeds, and you can even suggest some newbie-friendly map seeds to beginners who find the game hard to start with ect.

All in all I think this would add a lot to the game and expand player choice without requiring a lot of work. Win-win!

-Stigma

Hey all,

There is apparently a "ideal amount of colonists" number in each AI that bsically makes sure to offer you opportunities for new recruits if you are below that number (incapacitaed raiders, wanderers, slave ships ect.) but deny those same things if you are maxed. I read somewhere that this was 7 from kassandra classic and something like 14 for phoebe friendly.

I want to know - do prisoners count towards this number, yes, no, or partially in some way? It seems that if I have a lot of my total population in prison then I still sometimes get more people, although perhaps not as frequently? Getting 10-11 people on kassandra seems not too hard to do this way.

I know that doing this is a bit min/maxy but that's how I like to play the game, so I hope you can forgive me for trying to "abuse" the AI

I also want to know, what's this number for Randy Random currently?

EDIT: Also, is there some fairly easy way to mod this number? I think that it might be pretty fun to play a game on hard or random (which is hard in its own ways) but with a less restrictive amount of colonists - as manpower is very often a big limiting factor.

-Stigma

An easier solution (from a micromanaging point of view) would be turning the mouse pointer into a scaly hand and allowing to *slap* a colonist that wants to interrupt an important job.

Okay, so it doesn't allow to differentiate between similar "important jobs"...

As much as I like the dungeon-keeper reference, I don't think that would be a solution, as it is basically the same as we have right now and you still have to babysit the colonists which is very tedious. (Maybe you meant it as a joke, sorry if I didn't catch it if so :3 )

I think this is inherrently just a problem with "dumb" algorithms that don't look ahead at all and thus make stupid prioritizations. A lot of this problem can go away if the algorithms just get smarter in planning the execution of tasks, but I feel that there will always be a need for some way to "directly control" when you really need to - and priority-area and groupprioritize seems like it would be a good way to handle it.

-Stigma

[- Gun progression/combat progression in general.]

I haven't read the whole thread, but I definitely agree that there is a problem of too many guns. It needs to be limited in some ways - and also there should be a way other than selling to make use of excess guns.

Just very quickly here are some ideas:

- Gun progression/combat progression in general.
It doesn't make much sense to me (given it's essentially a survival game at heart) that the game gets hyper-militarized so quick. Gun-turrets aplenty are basically essential just to survive out of the gate, and both you and your enemies gain access to advanced weapons very fast. Solution: Slow it down considerably and have a much smoother curve towards advanced guns. Add "non-gun weapons". Homemade and improvised melee weapons and bows and the like that will be the stuff you first get access to. At first guns should be somewhat rare, and these low-tech improvised weapons should have low to no sell value, at least to off-world traders. Raiders at the start should also be low-tech to start off. Gun-turrets should be much further down the eventual tech-tree, so that you actually have to use some tactics, flanking ect. to survive encounters rather than gunturret chokepoints as is the current metagame. The game seems to already have workable basics like coversystems in place and its a damn shame that you never really get to use any of that in some tactical squad-combat. Having gun-turrets later in the game and not essential in the start makes it so that they can be more expensive and more powerful and more of a supplement to defense rather than the primary factor. Taking down that one raider with a lee-enfield when all you have are bows and homemade spears could get really exciting

This can lead to many fun scenarios of asymetry, like a horde of poorly equipped raiders vs. your few survivors with basic or advanced weapons (a classic "swarm defence), or the other way around, a handful of very heavly equipped raiders vs. your slightly more more numerous survivors with hardly any real guns. This should be a lot more varied than the same old same old raider attacks where primarily the difficulty is only determined by the number of attackers.

- Broken weapons: Maybe it's a bit a of a "cheap trope" but a primary way to reduce guns should be that many of them get broken in combat when dropped. That way, you may get a gun or two in a big encounter, not 20. It makes guns that you get actually exciting and valuable, rather than "ho hum, my 200'th shotgun.... how many weapons racks do I really need to construct??"

- Recycling: Broken weapons should be recyclable for a little metal, so you feel that you get something out of the deal and that all those weapons don't just seem to "disintegrate" as that feels like an extra cheap mechanic in many games where enemy weapons just "don't exist" once the enemies die.

-Stigma

Hey all,

This will be a real short suggestion. For anyone who has played the game, it is a really common thing to experience that your colonist wants to go eat or sleep just before he finishes the last 3 blocks of your mining job, or the last few conduit pieces ect. Then you have to resort to right-clicking and "prioritize" for each and every block of work - basically babysitting the colonist through it. Sometimes this is a matter of survival or death (those missing conduits could be the thing you need to power your defense against the imminent big raider attack) so you are often compelled to do this.

What we need to fix this is very similar to the existing "prioritize X" we already have, but allow it to select a whole area of existing queued jobs. That colonist will do these tasks above all else, including sleeping and eating. This would be a pr. colonist kind of thing.

I think it could also be useful to have a priority category for all colonists combined that works a little different. Let's say there is lots of stuff to haul, but right now I need to grab a load of metal that fell down from orbit. I don't want to let them haul all the other stuff first because I need the metal right now. i don't want to select priority on a single colonist because I need the job done faster than a single man can do it (and it may be too much work to do in one go without sleeping and eating too).

Instead the solution is to set a priority area. Simply put, anything in that area is classed as higher priority, but ONLY higher than other tasks in the same class. Example: a hauling task inside super-priority area is more important than hauling task outside the area, but not more important than mining task outside the area.

Result: I can simply mark the minerals area with super priority and then go do something else. As soon as my colonists have availability to do hauling jobs, they will do that hauling job before any other hauling. Problem solved!

-Stigma

Just wanted to comment on the "maintenance" portion of the post.
Couldn't agree more.  I was thinking the same thing.  Repair and cleaning would be great to have as "maintenance" so that I could set one or two colonists to run around and fix things and tidy up.  Let the construction guys work on new projects.  Depending on my base I would have enough janitors to cover a realistic section of the base.

Which brings up my next wish.  It would be great to assign colonists to only service a certain area.  They spend too much time running from one job to the next all over the map.  Hauling that important ore?  Oops, a conduit on the other side of the map needs repaired.

Cleaning and repairing is much more the same sort of job yes, so it could make sense to have that in the same category. At the same time though, you usually want repairs to be a higher priority task. I think I'd prefer it as a separate category.

Being able to set a work-zone for each colonist would be AWESOME yes!

The next best thing would be a simpler toggle of "only work at home" - which makes the colonist never try to do tasks that are outside the homezone. That is "almost as good" while keeping the system fairly simple. Personally I always prefer advanced controls and maximum control, but that is not always the best design as a game designer =P

EDIT: Thanks Semmy, yes this category probably makes more sense.

-Stigma

[- Customizable prioritization]

Hey all,

I've noticed a couple of problems that I want to draw some attention to when it comes to colonist task-handelig and priorities. I know it is early alpha still

This list is a little haphazard and not really organized, sorry. It is just a pile of issues in roughly the same sort of category.

-----------------------------
- Customizable prioritization is really needed. Current system is workable, but there are a lot of situations where you have to do less efficient distribution of tasks just because the priority line is fixed. A really simple example is this: miners and constructors often have skill-overlap (colonists good at one skill is usually decent at the other too). it would make a lot of sense to set one primary miner, and one primary constructor, but still have both of them do both things so that if there is no more mining then the miner helps out with the constructing and vice versa. Instead you now have to either toggle on and off skills all the time, or "bite the bullet" and set your miner to never construct - so that he spends his time when he is not mining not using his skills well and only doing something nonessential like cleaning.... A customizable priority list for each colonist would fix this and is probably not that hard to change coding-wise. For the UI, consider using color-coding? It could be a good way to still be able to see at a glance what things are high priorities.
-----------------------------

-----------------------------
- Repair/maintainance needs to be its own job. It really really really needs to. Why? Because repairing is logistically a completely different job than constructing is, and it makes sense to assign very different people to it.

Currently it is part of constructing which seems to make sense, but when a base grows big and complex the repair job becomes bigger than the constructing job and mostly involves running around more than needing a good constructor. In short - new construction projects (lots of work in a small area) needs a good constructor to finish the job fast. Repair jobs need lots of people to do the job so that someone close to the job can do it, and they don't really need to be highly skilled because it is running to the jobs that takes the most time usually, not the job itself. Repairs would make sense to assign to the same types of people that you have set as cleaners, haulers, maybe even wardens.

If repair is not separated out into its own skill then there will be big big efficiency problems once bases become larger than small-medium because you become forced to enable most of your people to do construction just to be able to handle repairs.
-----------------------------

-----------------------------
- Smarter task handling in general is needed. Everything works fine using "dumb" algorithms as long as everything is small-scale, but as complexity and size of the system grows these sorts of inefficiencies will cripple the system until the majority of the colonists time is spent not actually getting any work done.

There are a LOT of issues that need to be covered under this, and I'm not sure i can list them all here, but here is an example of what I mean:

Something far away needs to be mined, but the colonist is about half-hungry already. The dumb algorithm does not care and sends the colonist on the long walk to the site... and then he mines ONE block and starts walking home again because he is hungry. You yell curses at your colonist but it dosn't help Then he goes to eat and has to go back again to mine. When he finally gets back to the mining location he mines ONE block and then starts walking back because he is now tired. You slam your face into your keyboard until your eyes start to bleed...

Instead of course, the algoritm should do some simple checks before heading to a new task to see how far away it is and see if it needs to take care of hunger/sleep and other such needs beforehand, so that if you have to mine or construct something far away then just make sure that you are well rested and fed before you go (like a normal sensible human would).

IF the next task on the colonists priority list is not far to walk AND his bed is also close (typical base work scenario) then go to bed a little earlier than usual so you can wake up earlier. It makes more sense to be well rested and have good energy reserves for later if something important happens. Basically, when you can keep well resed and ready without sacrificing efficiency.

On the other hand, IF the colonist is close to a cluster of work AND that cluster is far from the bed THEN keep working until you are a little more tired than usual. On work projects far from the main base it makes sense to maximize the on-site work time so you don't have to make lots of extra un-needed travels (again, more common sense).

Let colonists know what the others are doing and make smart choices based on that. Let's say that there is a fairly small mining project far from the base, only 10 blocks, and you have 4 miners allocated. Currently, if priorities allow, then all 4 miners will get sent across the map to do the work. very often the first miner will finish before the others even arrive, and because the algorithm doesn't even check again until it arrives at the job the others don't even "realize" that the job is done ages ago until they arrive and try to start working. The correct response would be to analyze the task-cluster, see that it is a small job, and just send one guy to do the job - preferably the one that has the best combo of surplus sleep/hunger and is also highly skilled.
------------------------

I will draw a line here for now. I think this topic probably needs its own thread where people can make more spesific suggestions using pseudo-code for Tynan to tae sugggestions from.

Again, I realize its pre-alpha. I don't expect this stuf to be perfect yet, but in this sort of a game this is really important stuff. Half the fun is to watch your little guys be busy little worker bees after all, and if they become stupider and stupider the larger and more complex the pool of tasks become then that fun quickly turns into frustration instead.

-Stigma

Raufgar: that's how I read the suggestion too, and it seemed to make sense to me at least. I think Tynan might have misunderstood what he said.

That scenario works fine to explain a remote isolated rimworld with a surprising amount of survivors on it, but orbital traders wouldn't make much sense in that case, as the transcendentals would not be making it known that interference with the moon is forbidden (unlike a regular old government of some sort could easily do).

If you remove orbital traders and stick with on-planet trading only then that angle works though.

-Stigma

Here's an idea:

Instead of having a "stranded, robinson crusoe-style" scenario, have the planet be an exile/prison planet. The sort of place where people get banished and stranded as punishment - without the sort of support structures needed to ever get off the planet on their own. (perhaps a possible end-goal can be to defy the odds and create a craft from scratch capable of reaching orbit and thus escaping Rimworld).

I think this could make a lot of sense.

- Your colonists could be there because they are/were really terrible criminals - or you could still crash there. Either works. Pick one for the story, or leave it open to interpretation/roleplay by the player.

- It explains how there can be orbital traders that are willing to trade, but unwilling to otherwise help in any way (they know that no one is allowed to interfere, under extremely harsh penalties). Local traders make sense to have in the game too though. I don't see why you should not have both. Orbital traders can be much more rare, but have larger stocks and access to technology levels that are nearly impossible to get your hands on planetside. Of course... the orbital traders would know this very well - and would squeeze you for every dime they could much more so than local traders. They know how desperate people are down there and don't care one bit about exploiting "criminal scum".

- It explains well how you seem to be on a planet with seemingly endless hordes violent raiders and assorted other criminal scum (which doesn't make sense at all otherwise especially if it's supposed to be some random uninhabited rock)

- Plenty of opportunities for all sorts of stories and characters and "quests". There are bound to be some good guys there too, not just crazy criminals. Political prisoners, children of criminals ect. The place isn't a "prison" so much as it is a place to dump people you want to make disappear from the "real world". Rimworld can be basically however you need it to be under this setting - but with a perfectly logical reason for why it is almost completely isolated from the rest of the galaxy.

-Stigma