next »

[JobLeonard]

Ok, somehow I hadn't subbed for this yet.

Although I don't think I'll dig through 200+ pages of devlog... Anyone with recommended highlights?

[Christian]

Ok, somehow I hadn't subbed for this yet.

Although I don't think I'll dig through 200+ pages of devlog... Anyone with recommended highlights?

Browsing the Kickstarter updates is probably the fastest way to get a summary of development over the past two years
https://www.kickstarter.com/projects/rain-world/project-rain-world/updates

Edit: pretty awesome that in a year's time, they went from getting pathfinding to work in the Unity build
http://i.imgur.com/BcAJ6xM.jpg
to giant sea monsters and implementing motion/distance-based vision

[JLJac]

Thank you everyone! Glad you like it  Still a bit of work on this one, but now it will be mostly AI stuff.

Update 453

The next waypoint is making this beast swim between rooms, but the "map" we had (colored squares, you might remember) got broken in the region connection project which would make it very hard to know if I had managed to make it roam the world, haha! So the question was, make a new map or fix the old one? As the old one basically didn't work - it was just colored squares and to check if two rooms were connected you had to search for squares of the same color hahaha - I decided to go with a new map. This new map gives soo much overview compared to anything we've had up till now, feels really good to watch the little creatures move around in abstract space!


(Note that this is not some canonical arrangement for the region - how the rooms are located relative to each other you'll have to ask James about, this is just me pulling them into some configuration where I can see them.)

On top of this I actually got the leviathan swimming between rooms ~ but it's not completely finished yet. It can use this "sea highway" as we call it to teleport between rooms if swimming off-screen, but it also needs an AI behavior that tells it to do so. I imagine this working pretty much exactly as the vulture; it will slowly roam the region, if arriving in a room where there are edible creatures it'll make a quick detour into that room.

[JLJac]

Update 454
General technical housekeeping. The leviathan needs to be able to travel and hibernate in a system completely free from the normal den-and-exit framework for abstract space, which means that some new abstract node types need to be defined. This opens up for a long awaited other change as well - no more vultures through shortcuts! With this new system vultures and leviathans alike should be able to travel between rooms completely without using the usual shortcuts.

Seeing the dev interface map, James and I had another talk about in-game maps. What we would trade away is obviously the complete no-UI purity, but we would gain some pretty nice pros as well. One is of course playability, it will just be easier to navigate this huge world if you have some sort of map system to guide you - however minimal it will be a huge difference from not having any at all. The other would be vanity, haha - these environments are HUGE and very complex, but that doesn't come through very well viewing it screen by screen. A map would give an immediate visual impression of how large these regions actually are, giving more credit to all the work James has put in.

Here is a 5 minute mockup I made using a few of the HI rooms. It's put together very haphazardly and the region depicted doesn't match the screenshot behind it haha, but it might be a starting point for a conversation:



The idea for a map like this would be that it fills in organically as you move through the environments, and that everything is drawn as a unified shape somewhat breaking up the very distinct room-by-room nature of the game and making it feel more like a united whole.

One problem that James brought up is that in games with a map, there is a tendency that the player will end up "playing the map" rather than the game. I can imagine a few solutions to this, one of which would be that you can't actually move as the map is up (the controls instead pan over the map) and/or that you can place a waypoint on the map and when you then bring the map down you instead get some very small piece of UI that only guides you to the waypoint, such as a little compass or something. That way you could bring up the map, place your waypoint, and then not have to bring it up again until you've gotten there. Basically any idea that keeps the player from having the map up all the time is game, because even if a map might be a nice aid we want the player immersed in the game, not the map.

Another question is what a map should look like if one is included. If it could look more "slugcat instinct" than "computer interface" that would be cool, but it still needs to be crisp and clean.

[Christian]

Honestly I don't think it's the sheer size that's the issue, it's remembering how to get back to your shelter with the clock ticking in that huge environment. Perhaps rather than a map of the entire world, only the rooms that lead back to your most recent shelter could be shown permanently as you explore. Other rooms could fade a while after you leave them, or something like that

1) That would visualize the innate knowledge an animal would have in regards to getting back to its den

2) But also maintain a sense of mystery in the world and exploration

[Samson]

Simplicity is key in getting across the 'instinctual' look you describe. Perhaps the Slugcat's position on the map could be suggested by emanating gradient brightness, such that far away regions are faded out. It would make it less accurate, but that might fit the feeling you are aiming for (less obstructive too).

[Zorg]

Perhaps rather than a map of the entire world, only the rooms that lead back to your most recent shelter could be shown permanently as you explore.

In this case you would not need a map, only an arrow, which points in the direction of the last shelter (or closest known shelter, or more arrows for all known shelters).

I'm voting for no map.

[oldblood]

Personally, I think the map looks good. But if it was up to me and you want it to feel "organic", I'd scratch the map and include some sort of faded arrow/indicator system that simply points back to the exit in each room that leads you back to your den.

Basically the equivalent of the scent trail in nature. This keeps the mystique of the world closed to the player (they don't innately get to see the layout of the world) and regardless of where they are- they can find their way back to their den. Seems like this would be a cleaner solution that 1. Gets the player back to their den easily. 2. Gives the more "organic" feel. 3. World remains mysterious. 4. No UI.

EDIT: Looks like other have suggested similar while I was writing this...

[adge]

I'm also voting for no map. I hate HUDs and stuff. They always keep me from fully immerse into a game.

Another way would be just highlighting the shortcuts's entrances you need to enter in order to get back to the shelter. Yellow blinking or something like that. Then you enter the next room and the next shortcut is blinking to show you where your shelter is.

[jamesprimate]

im torn on this too. personally, i hate the UI aspect and think it will break immersion. but on the other hand the game world is so large and complex that there needs to be some way for the player to visualize it. also, to if we are to convey destinations and locations to the player, how do we do it without a map of some sort? its either that or literally "showing the way" (arrows or illuminated exits) which i think would be even worse, as that removes the sense of exploration altogether. its not something that is going in right now, so we'll ponder it for a while as we go.

one thought i had was to have the map develop as you explore, but have it only be viewable from certain locations, say some sort of large map terminal in rooms connected to the region gates or something. then you can have some way of planning your route region per region and seeing where is unexplored, but its also far enough away that it doesnt have you "playing the map" breaking the immersion. joar doesnt like this idea and im kinda meh on it as well.

another possibility that might work similarly would be to have the map only be visible during hibernation. this solves a couple of problems actually, as for now the hibernation is sort of just "there" without much utility. having it be the time when you can peek at the "slugcat dream map" creates an added function for it, keeps immersion and doesnt interrupt the flow of the game at all, making use of a natural pause in the gameplay. the more i think of this the more i like it.

[fall_ark]

Isn't there some "one-way-only" rooms? (e.g. a vertical drop where you have to take a detour to other rooms to go back up) Those things are frustrating enough in games with complete maps, I don't think any of the suggested "limited directions" methods will work with that. It'll only lead to false directions and frustrate the player even further.

If you want to make the map feel more immersive, it'd make sense to have landmarks and creature nests on each room instead of the exact terrain outlines. Hopefully that would also encourage the player to remember what each map around those landmarks look like, and rely on his own memory instead of constantly checking the map.

If we do go with the limited map or vague directions solutions, I hope we can at least check the explored map while in shelters.

[Saijin_Naib]

I'm a huge fan of maps, and I find that they don't cause me any issues in play or with immersion.

I look to Knytt and other Nifflas games for how maps for these types of exploration-centric games work, and their associated issues & limitations.

I understand people who do not like the idea of the map for gameplay & immersion, so perhaps you could implement an option to change how the map behaves.

Map options:
1) No map (cleanest, most immersive)
2) Simple hints (highlight shader on corner of screen that points to shelter/home)
3) Compass hints (waypoint markers that are more discrete than the shader, but still Slugcat/world appropriate)
4) Fade/Memory Map (show current room's geometry fully, and "simple" geometry of previously explored rooms that are out of sight)
5) Full Map (show full geometry of current room with creatures, shortcuts/etc, show "basic" geometry everywhere else, show all discovered shelters/homes, show vague marker for "big" creatures (last time leviathan spotted, last time vulture spotted, etc), show markers for doorways/gateways/etc)

[Christian]

Was thinking if there's going to be a bio-mechanical element to the wildlife, a map screen might be a way to hint at stuff like that. AC-style. Hazy, dream-like, with glitchy effects or something like that? Idk, could be interesting

But I like Saijin's 2) and 3). Keeps it in-game while helping the player remember where shelters are in this big environment

Or even more subtle, have the shortcut/vent indicators that lead back to your den glow a different color.

I'm a huge fan of maps, and I find that they don't cause me any issues in play or with immersion.

I look to Knytt and other Nifflas games for how maps for these types of exploration-centric games work, and their associated issues & limitations.

I understand people who do not like the idea of the map for gameplay & immersion, so perhaps you could implement an option to change how the map behaves.

Map options:
1) No map (cleanest, most immersive)
2) Simple hints (highlight shader on corner of screen that points to shelter/home)
3) Compass hints (waypoint markers that are more discrete than the shader, but still Slugcat/world appropriate)
4) Fade/Memory Map (show current room's geometry fully, and "simple" geometry of previously explored rooms that are out of sight)
5) Full Map (show full geometry of current room with creatures, shortcuts/etc, show "basic" geometry everywhere else, show all discovered shelters/homes, show vague marker for "big" creatures (last time leviathan spotted, last time vulture spotted, etc), show markers for doorways/gateways/etc)

[adge]

I totally get why an arrow would be even worse, but not why illuminated exits would be.

I think a map would 10 times more crash the sense of exploration than a shortcut thats subtle lighting up to indicate that
somebody just used it. One look and everything is revealed. Even if the map shows only places where you have been recently.

I think it doesn't really make a difference if the shortcut is glowing or you stop and open the map to locate the way to the next
shelter. If you have a map the reason why you would use it most of the time to get back to the shelter. And if the way is indicated
through a light or you have to look for it on the map shouldn't really differ from each other too much. They exist both for the same
reason. Only difference is that you already are in a hurry when getting back to the shelter and a map would maybe take up
too much time and you get frustrated because you have to fiddle around with every shortcut thats near to another because the
map is simplified and does not exactly show the shortcuts different routes.
Seems I really like shortcut indicators right? Nah I just hate pop up maps, HUD's and UI's.

Illuminated exits:
If you go back to a room where you already used a shortcut in but left with two different shortcuts you would have to remember
which one shows the way back to the shelter. With this method you would have something like a helping indicator to find your
shelter. But you would also have to be very careful by using it because you could easily destroy the clear path if you left a room
through 2 different shortcuts or entered them through 2 different shortcuts.

Have to try if you are only lighting up the shortcuts exit or entrance or both. Entrance would not really make sense. So probably
the exit or exit and entrance. But exit&entrance would lead to confusion.

This for me seems to be a system that somebody in Rain World's world really could have implemented because of...different
reasons :D. It just seems more realistic to me than: "Press M for the Map". Even if you cant move during use or whatever.
It's just another layer that comes up and shouts in your face: "This is a Game!"

A problem as fall_ark said would be "one direction rooms"

Another way could be terminals that the slugcat can access but instead of a map they show the shortcut...illuminated =D

[Teod]

Another question is what a map should look like if one is included. If it could look more "slugcat instinct" than "computer interface" that would be cool, but it still needs to be crisp and clean.

It should look like camera zooming out. We already play from unrealistic perspective, seeing everything way outside of slugcat's line of sight, so this will be the least immersion-breaking way to do it.

The geometry should be simplified, of course (I know you can't technically make a true zoom-out), but not just in terms of pixel density - shapes and patterns should be made more simplistic and stylized. Focus should be made on general shapes and objects, not exact room layouts and details. That way player's memory and map will help, but not entirely replace eachother. It's player's job to remember the rooms, but map shows where those rooms are.

im torn on this too. personally, i hate the UI aspect and think it will break immersion. but on the other hand the game world is so large and complex that there needs to be some way for the player to visualize it. also, to if we are to convey destinations and locations to the player, how do we do it without a map of some sort? its either that or literally "showing the way" (arrows or illuminated exits) which i think would be even worse, as that removes the sense of exploration altogether. its not something that is going in right now, so we'll ponder it for a while as we go.

I agree. The navigation should be the players job. Map should say where you are, not where you need to go. Otherwise nobody ever will go in any other direction.

one thought i had was to have the map develop as you explore, but have it only be viewable from certain locations, say some sort of large map terminal in rooms connected to the region gates or something. then you can have some way of planning your route region per region and seeing where is unexplored, but its also far enough away that it doesnt have you "playing the map" breaking the immersion. joar doesnt like this idea and im kinda meh on it as well.

another possibility that might work similarly would be to have the map only be visible during hibernation. this solves a couple of problems actually, as for now the hibernation is sort of just "there" without much utility. having it be the time when you can peek at the "slugcat dream map" creates an added function for it, keeps immersion and doesnt interrupt the flow of the game at all, making use of a natural pause in the gameplay. the more i think of this the more i like it.

I don't think either of those are good ideas. You either have a map or you don't. Limiting its use in any way is really annoying. I remember being infuriated by a game when if didn't let me open the map during the jump. If you want to nerf map's power - keep those nerfs within the functionality of the map itself: how much it shows and how accurate it is.

[JLJac]

Wow, loooots of interesting discussion! 

#Arrow pointing back to den: This would be really easy to implement, and maybe sufficient, or at least a great step up from nothing at all. It would basically be a breadcrumb system, but where the breadcrumbs are deleted if you double back - so if you go back and take another branch or cross your previous path the trail would just be cut at that point and then continue in whatever other direction you started off in.

#One-way rooms: If we do have a bread crumb system, a one way room will become a risk and that's kind of fine right? I imagine a rat would be hesitant to drop down a pipe or something for this very good reason.

Map options:
1) No map (cleanest, most immersive)
2) Simple hints (highlight shader on corner of screen that points to shelter/home)
3) Compass hints (waypoint markers that are more discrete than the shader, but still Slugcat/world appropriate)
4) Fade/Memory Map (show current room's geometry fully, and "simple" geometry of previously explored rooms that are out of sight)
5) Full Map (show full geometry of current room with creatures, shortcuts/etc, show "basic" geometry everywhere else, show all discovered shelters/homes, show vague marker for "big" creatures (last time leviathan spotted, last time vulture spotted, etc), show markers for doorways/gateways/etc)

Nice line up! All of them except number one seems to have some UI though, so if we are going to rip that band aid, maybe try to get max value out of it and go all the way?

another possibility that might work similarly would be to have the map only be visible during hibernation. this solves a couple of problems actually, as for now the hibernation is sort of just "there" without much utility. having it be the time when you can peek at the "slugcat dream map" creates an added function for it, keeps immersion and doesnt interrupt the flow of the game at all, making use of a natural pause in the gameplay. the more i think of this the more i like it.

This is also a nice idea, as it would give hibernation a sense of purpose. Maybe if you could see what swarm rooms are active in this mode (slugcat sense!) and place a marker, and then when you go out you get a compass to that marker or something? Would make for a nice "planning mode" similar to DF's embark screen. "This is the idea, but we all know it's gonna go to chaos once we get out there."

I don't think either of those are good ideas. You either have a map or you don't. Limiting its use in any way is really annoying. I remember being infuriated by a game when if didn't let me open the map during the jump. If you want to nerf map's power - keep those nerfs within the functionality of the map itself: how much it shows and how accurate it is.

This is also a valid point. But if the hibernation mode feels very separate from the game, almost more like a between-mission menu screen, maybe you wouldn't get that frustration? Also coming home to the den and getting to see the map grow when your newly explored rooms are added could be a nice end-of-cycle reward?

[Saijin_Naib]

Nice line up! All of them except number one seems to have some UI though, so if we are going to rip that band aid, maybe try to get max value out of it and go all the way?

Well, that's certainly a difficult question to answer, and not a question that you're going to get 100% agreement upon from any one perspective. The way I had it laid out was, to my mind, a map "mode" that would hopefully agree with each of the major perspectives I've identified thus far in the thread regarding map vs no map and the various stages in-between.

From a technical point of view, I was thinking that you could implement a class system in the geometry that would determine what objects got mapped, how they got mapped, and when they would show up on the map. In this way, you simply set the classes, and then the various Map Modes are simply choosing what classes you include.

For instance, if you give the home/shelters a special class "home" and "shelter" and use Map Mode 2 (Edge Highlight), you simply have a small shader on the edge of the screen showing the general direction (in X and Y axes) where home lies (bottom left corner, upper left corner, left center, etc). That same class would work for Map Mode 3 (Compass/Primitives), and would just use those overlays to communicate the direction instead of the more general edge highlight.

If you give the "major" geometry of the room a class, say "geometry_major", on Map Modes 4/5, these would be the structures that would show up on the rooms outside of the view of Slugcat. So, instead of getting a detailed picture of every feature in the room, every platform, every exit/entrance, you only get the boundaries of it. Think of the "geometry_major" class as defining the outside of a container only.

If you give various other elements of the scenery a class, "geometry_minor", on Map Mode 4 these would show up ONLY where Slugcat is. "geometry_minor" would encompass things like small platforms, outcrops, rails, etc.

If you gave the "big" creatures a class, "creature_major", this class could place a large/obscure dot on the map in Map Mode 5 that would indicate that Slugcat last saw this big thing in this room, nothing more. This seems lore/world appropriate to me, as one would assume that a creature that has a strong survival instinct would take note of where it last saw something that was a danger to it.

If you give the other creatures the "creature_minor" class, on Map Mode 5, they'd show up "live" on the Map in the room Slugcat is in, because he can see them.

In terms of whether or not this is all feasible and sensible, just understand that living creatures do have an innate ability to navigate and remember locations and layouts. These maps we create are derived in part from every sense we have (sight, sound, smell, touch, taste) active use of, and passive senses we have (grid and place cells). With grid and place cells, these cells fire off when we are at or a near a specific place that we remember. It is not inconceivable that Slugcat could have more active control of these cells, as many animals today rely upon them (with great accuracy) to navigate them large distances repeatedly and reliably.

I hope this helps you understand what I was going for Joar, and if not, please let me know and I will do my best to do some mockups and otherwise explain myself better.

Thanks for your time.

[Zorg]

#Arrow pointing back to den: [...]

I like this!

[Crispy75]

another possibility that might work similarly would be to have the map only be visible during hibernation. this solves a couple of problems actually, as for now the hibernation is sort of just "there" without much utility. having it be the time when you can peek at the "slugcat dream map" creates an added function for it, keeps immersion and doesnt interrupt the flow of the game at all, making use of a natural pause in the gameplay. the more i think of this the more i like it.

This is also a nice idea, as it would give hibernation a sense of purpose. Maybe if you could see what swarm rooms are active in this mode (slugcat sense!) and place a marker, and then when you go out you get a compass to that marker or something? Would make for a nice "planning mode" similar to DF's embark screen. "This is the idea, but we all know it's gonna go to chaos once we get out there."
...
Also coming home to the den and getting to see the map grow when your newly explored rooms are added could be a nice end-of-cycle reward?

I was going to make these exact two suggestions. When you hibernate, the map comes up (lo-pass filter on rain noise, soothing music plays...) and your previous expedition is plotted out, painting in new parts of the world. Major landmarks get highlighted (shelters, active feeding rooms, last known locations of Megabeasts). Maybe the "new information" is painted bright, but the "old" information is faded to indicate that you don't know if the situation there has changed (bats moved on, vulture hunting elsewhere etc).

Then the player is given control and can place a waypoint on the map, which will light up the "outgoing" breadcumbs.

Maybe progression through the game could be rewarded by allowing multiple waypoints with different breadcrumb colours.

Overall, I think a map would have a positive effect on the game. It makes the true scale of the world apparent, can be useful without being overpowered, and just looks cool - even the preview is impressive (look where I've been! Look where I could go next! Options!)

[Crispy75]

I know it's just a quick mockup, but I'm finding it hard to tell what's solid geomtery and what's navigable space. Obvious in places, but you have to solve a maze to figure it out on others...