next »

[JLJac]

Update 359
Running around in the world James has been working on was amazing! I put on some rain world OST and just ran around exploring, and that's a pretty cool experience just that - no lizards or anything added yet. It's a lot of fun to for the first time play the game without having made the levels myself, the anticipation and curiosity obviously becomes much greater. 30 rooms feels big, I spent quite a lot of time just running around and trying to create a mental map of the place.

In other news, bats can now dive down and disappear in their weird "bat grass", only to pop up in the grass somewhere else. The animation isn't really gif-worthy yet, but the basic mechanic is in there. Next up would be room-to-room migration, I believe. Saving extra dijkstra maps to the level files proved to be less of a horror than feared and anticipated.

[Nilanjan]

Great man!!! We get so much enjoyment just watching the gifs so I can imagine how awesome it must feel to run around the world. Can't wait to have a go myself...

[Bones]

Have been watching this thing evolve.
It has been a pleasure.

Keep up the great work, it's already quite a world of it's own.
Looking forward to the future Rain World.

[JLJac]

Update 360
Started on a big scary thing called Flies World AI. This entity is supposed to coordinate flies on a world level, through abstract and realized rooms alike.

What it can do so far is create dijsktra maps in the abstract world, mapping the distance to each swarm room for each room. Any room (or node actually) can ask the world AI for a "migration direction", ie a neighboring room that flies in this room should migrate towards. Depending on what mode the World AI is in, it will calculate this migration direction depending on different criteria.

Right now I have two such modes set up: spread and swarm. In the spread mode, the migration direction of any room will be the neighboring room with the least flies in it. This means that the flies will spread evenly throughout the entire world. The swarm behavior has flies move towards the nearest swarm room.

Both of these seem to work nicely! I can have a lot of flies gather in a room, and then go to the next room and start the "spread" behavior. Soon flies will come out of the swarm room, and spread out in every direction. Then if I switch back, I can see them emerge from all over to gather in the swarm room again.

Next step is to create some swarming behaviors that actually make sense, and tie them to the rain cycle.

[SafetySnail]

It'd be awesome if different creatures knew the rains were coming earlier than others. Imagine running around the world and you see a huge swarm of flies all headed the other way, could be a cool way to let the player know they need to find cover.

[DarkWanderer]

I like that idea. Reminds me of how dogs/cats supposedly know an earthquake is coming before people do.

[JLJac]

Update 361
Did some work over the weekend - mostly revisiting the Quantified Creature stuff. Quantified creatures are creatures where the individuality of the creature is irrelevant, meaning the game doesn't save health and stuff like that for the critter when in abstract space. Right now bats are the only quantified creature. Basically the game does not save these creatures in the abstract creature framework, but just saves the amount of creatures per room (actually node) as an integer.

The problem was that there were some loop holes when changing between the realized and quantified states - meaning that the populations weren't consistent. Sometimes bats would disappear or appear in the transitions between the states, and there'd be situations with schrödinger's bats, bats existing and then suddenly not existing, or the other way around. After re working the system, it seems it's water tight; 50 bats is 50 bats, and they only disappear or appear when eaten or spawned.

With that down, I could actually get started on bat migration patterns.

Right now I have a pretty simple system. Bats spawn in their swarm rooms, and hang around there. Every now and then the game checks all the swarm rooms, and finds out which is #1 and #2 in amount of bats.

#1 is assigned a "spread" behavior, where bats will move out of the room and spread into neighboring areas. When they get close to another swarm room, they'll start gravitating towards that one.

#2 is assigned a "migrate" behavior, where another swarm room is picked at random and the bats start moving towards it. On their way there, they'll most likely hit another swarm room's proximity, and start gravitating towards that one.

As this goes on for a while, #1 and #2 will be more or less emptied, while other swarm rooms will gain bats. Then the same process hits again, and a new #1 and #2 will be assigned.

The effect is that bats seem to move around between swarm rooms at random. If you're out in the world, between swarm rooms, you'll sometimes see little flocks of bats passing through your room on their way to some other place. This is the key feature, as it'll allow the player to follow those stray bats and end up in an active swarm room.

It will take some playtesting before we know the system really works. But I'm happy with the fact that the system can potentially lead the player towards swarm rooms without necessarily catering too much; the bats are not trying to help you actively, they're just going about their own business and if you're clever you can take advantage of that behavior.

Today I'll fine tune it a bit further, and see how it plays out. Maybe if stray bats become too sparse, so you very seldom have any to follow, I'll add a phase at the very start of the cycle where migration is more common.

[jamesprimate]

^ super excited about this. we'll have to do a bit of testing how it works in-game, but having such an elegant, natural solution to the "how do we unobtrusively push players toward swarm rooms" problem would feel like a real victory for the mood were trying to create.

[JLJac]

Update 362
Huuuuge step. Got the entire world up and running, with migrating bats and lizards spawning in their dens, later emerging to hunt, and then returning as the rain approaches.

Good news and bad news. Good news are that YYYAAAYYYY it's up and running! Everything basically seems to work - the creatures spawn, move about, you can encounter them, they don't disappear or randomly appear, the ai seems to be able to somewhat handle most situations.

Bad news are not really anything unexpected, it's just that in this new - vastly more complex - context every little glitch and lack of polish becomes more prominent. The lizards getting stuck, the AI throwing missing reference exceptions, the AI not always making the best decisions, all of that is now way more visible. In my previous test environment there was just a single room, or a few interconnected ones, and I'd hit restart over and over while testing stuff. Now we have way longer sessions of gameplay, making every error stick around for a longer time and becoming more visible. This, in combination with the environments and situations being generally more complex and error-prone, makes for a generally more glitchy experience.

But, as I said, this was pretty much expected. Now the glitchy stuff is there, all visible, and I'll just have to hunt the issues down one by one. Everything essential and important is working.

It's probably too early to say anything about the game design implications of this new environment, I just haven't had enough time to play the game. I've noticed a few things, such as despite sprinkling the world with what seemed like quite a lot of lizards, lizard encounters seem sparse while moving through the world. Maybe this has to do with the amount of lizards, but I rather think its about their mobility. I might want them to move more between rooms.

Gonna play the game a bit more! Brb!

[Christian]

So you're going to need testers soon, right? To iron out those kinks?
 

[ofx360]

It's super exciting to see and hear about these day to day changes. It's amazing how quickly you guys find and resolve your issues. Keep it up guys, really inspiring stuff!

[Caspar2000]

So I missed the Kickstarter campaign for alpha access. Playtesting seems like a cool way to catch up. I could give it 2 full weekends.
=D

Would love to be part of the process.

[DarkWanderer]

So I take it the challenge rooms you can play in single player or multiplayer(after having had discovered them) will be tied to the "swarm rooms" then? I remember James I think saying the challenge levels are based around rooms with bats in them. 

[JLJac]

Yeah that was the original idea, but I think James have had some other thoughts since then. James?

Update 363
Another significant step - the room loader. This is the thing that loads rooms that are near to you and unloads rooms left behind. Right now it'll light up all neighbors to where you are (there is a limit, so not infinitely many rooms), and if too many rooms are active it'll start shutting them down. Then it goes by stuff like how long ago the room was viewed, and how many screen transitions ago.

Neighboring rooms that are activated and then never visited are unloaded on the next room transition. The system seems to run well! Performance wise it's smooth enough so far, and having all neighboring rooms loaded creates the nice effect that whatever room you're entering has been loaded a little while. Earlier the creatures would display a few frames of "I've just been loaded and will now try to act like I was here all along", but now when you first encounter them their animation states are already looking natural.

Other stuff of today's progress include:

* James requested being able to define some rooms and room connections as off-limits to bat migration, in order to be able to exercise a little more control over how bats move in the world. So I got some functionality for that in.

* The dripping water from the ceilings would be much more intense in small rooms, because it worked by picking a random ceiling tile every now and then and creating a water drip at it. Smaller ceiling, more concentrated dripping. I fixed this so the ceiling dripping appears a bit more even.

* You could wait at a shortcut entrance and eat an infinity of bats as they popped out of it. This obviously kind of broke the game. The hack I put in place is that bats now shoot out of shortcuts pretty fast, and have a few frames of "uncatchability" - the in canon idea being that they're swishing out of there too fast to catch or something.

* There was an error with effect colors (colors on plants, signs etc) where the last room's effect color would carry over when moving between rooms. Fixed that.

* Some minor animation quirks with the player character ~ there are more of those, but they're not really game breaking so they don't have the absolutely highest priority.

* Made the debug stuff (ugly map and ugly mouse cursor) hideable with a button. Long time coming, this one!

* Made it so that you can start the game in any room in the world, hopefully making level design a bit less painful for James as he won't have to actually walk all the way to the level he's working with in order to try it out. 

So, that's one chunk of quirks down! The most scary stuff though, the lizard related glitches, I haven't gotten around to yet. Maybe tomorrow!

[fylth]

Many of the rooms seem to be open (as in, no ceiling), are bats capable of leaving one room this way and entering through another?

[jamesprimate]

So I take it the challenge rooms you can play in single player or multiplayer (after having had discovered them) will be tied to the "swarm rooms" then? I remember James I think saying the challenge levels are based around rooms with bats in them. 

still planning on that! though it looks like we are having some multi-screen swarm rooms, which would be impossible to do 4-player multi on, so im thinking rather than "every swarm room is also a multiplayer room" we maybe just choose the best ones (we will probably have QUITE A FEW to choose from) and make a bunch of multi-player specific ones down the road as well.

Many of the rooms seem to be open (as in, no ceiling), are bats capable of leaving one room this way and entering through another?

we just talked about this actually! i think it would be really cool for mood, but Joar rightly points out that it might break the usefulness of migration as a "guide". So we'll see? Maybe there is a clever way to do both. But I think we'll aim to get the build fun and functional first. (it is getting there though!)

[JLJac]

Update 364
Messing around with glitchy lizard behaviors. Introduced a few countermeasures against different getting stuck scenarios, and they do seem a little bit less prone to end up wiggling in some corner. Still a lot of those to hunt down though.

When I made levels, I'd keep the lizard AI in the back of my head and adapt the layouts quite a lot to cater to lizards. James, who hasn't programmed the lizards, made a bunch of levels that were more free-form, and he naturally gravitated to adapting them for slugcat platforming rather than lizard movement. This meant that the lizards would oftentimes end up isolated in sub sections of the rooms, because the intended connection to get out of there was designed for slugcat jumping and not accessible to the lizards.

Yesterday we had a little talk about this issue, and decided to work on it from both ends. So James has been adapting the level layouts a bit, and I have been doing some additions to the lizards movement. Among those additions are a few new ways for lizards to move between tiles that are available to them. For example, earlier a lizard could stretch across from one tile it was allowed in to another, across a third empty but non-allowed tile. This was very blocky though, as it would only be allowed in perfectly vertical or horizontal cases. Today I added a connection that goes 2 tiles vertical and 1 tile horizontal, or vice versa (like the knight in chess). These kind of additions enable the lizards to traverse more different kinds of terrain, and opens up the maps to them.

With James' level design adaptions and these changes, I think the lizards are going to become quite mobile throughout the world. However, we noticed that the isolated lizards had some benefits as well, as they gave us more control over lizard encounters. So now we'll try to find the perfect balance between penned-in and free roaming lizards.

[JLJac]

Update 365
Not been many gifs lately, so here's a little bonus. After digging in the lizard AI all day I decided to do something more fun, and implemented this hanging animation for the lower end of vertical poles:

[Wlad]

Update 365
Not been many gifs lately, so here's a little bonus. After digging in the lizard AI all day I decided to do something more fun, and implemented this hanging animation for the lower end of vertical poles:

It looks a bit weird, shouldn't the slugcat face sideways during the animation? Also arms have strange curve.

I am not animator and I don't have drawing talent, so it's quite possible that I am completely wrong.

[Citsua]

Best animation ever or something. I'd spend hours looking at that emotionless slugcat hanging like a boss.