Hey, sorry I haven't replied in awhile -- love the progress you've made in fleshing out the core loop! I'm impressed by your discipline in constantly refactoring and keeping everything clean and functional.

For the breeding problem, I wonder if you could re-contextualize the dog swapper into some kind of contraption (similar to your old two-armed cupid love smasher -- although probably not that extreme )  that you could use to explain why the dog-swapping isn't 100% effective and might introduce side-effects. Then maybe you could have a system of trade-offs where if the player chooses a higher % chance of a specific bodypart mutation/swap at the cost of an increased chance of an unrelated mutation. But if the player is more willing to leave it up to fate, the bodypart being swapped becomes more random but the side-effects are less severe .... or something where the player needs to make a choice in priority.  As far as making it deterministic, yeah I think a seed per dog plus a hash-table lookup should do the trick.

The speed of mutation is a tough one. Too fast and you lose continuity and normalize it so it's less of a big deal, but too slow and you risk players not even noticing it or it being so subtle as to not serve much of a purpose. I'd bet this will be something you end up continually tuning throughout the whole development, since there isn't a simple answer. My first thought is to have a timeline for each dog where you can scrub along the timeline and see their development over the course of their lives from puppy to old age -- maybe with a collection of photos (similar to polaroids or a family photo album kind of vibe?) However, this would probably be a separate UI screen/panel which would pull you away from the main sim, which probably isn't ideal. Basically some kind of recap system, or "last time, on ___".

So far when you've been observing the simulation over time, do you notice dog cliques forming, or particular pair dynamics between specific dogs? It might be cool to see some lite pack dynamics or long term relationships emerge. For example, 2 dogs that always love to run around and play with each other over-enthusiastically, disturbing other dogs in the pen who prefer to sleep. Inter-dog dynamics and individual dog personality (doggonality?) are fun to observe in the real world, anyway, although I could see it being difficult to engineer...

Thanks! Unfortunately there'll definitely be more rebuilding and refactoring in my future (particularly for some of my AI systems), but it does at least feel like a lot of the game is starting to solidify and feel reasonable. Maintenance isn't the most fun thing in the world but it's way better than just watching my systems slowly degrade.

I've actually been working on a solution to the Dog Swapper issue and I think I've come up with something I like. I'll talk more about it below.

Absolutely agree that mutation speed is something I'll be tweaking for a while. It's gonna be tricky to get that right but it'll be crucial. I really love the idea of a scrubber for each dog you own. It'd unfortunately need to have a bit of a loading time (especially for dogs that'd undergone a huge amount of mutations) to be seamless, but I bet it'd be really fun to be able to look back on all that. I'd also have to play with compression a bit so my save files aren't out of control, haha.

As far as long term dog dynamics go, I need to do some work in that department. None of my AI systems were built when I was really able to do serious long term testing and things don't scale as interestingly as I'd hoped. There's an overhaul in my future there for sure and I'll probably try to tackle dog personality in the same swoop.

---

I did a lot of playing around with the game and thinking about the way the Dog Swapper needs to work and I think I've come up with a good solution. As I said last time, the issue with the system is that there was zero element of surprise. You got whatever you expected every time. What I've done to counter that is once again remove some element of control. Instead of being able to directly select gene categories to swap, you now have to use a relevant crystal.

Crystals are single-use objects that will act as both currency and modifiers for the Dog Swapper. Standard crystals will swap genes directly, just like the original buttons did, but there will also be crystals that modify the genes they swap and result in dogs with modified traits.

Temporary UI here, but hopefully this illustrates what I mean.

A standard body crystal will just let you swap bodies directly, but something like a "Tiny" body crystal will swap the bodies PLUS shrink the associated genes a bit.

The genetic shrinkage is new but still not necessarily surprising. It's more just another aspect of control and manipulation. That said, I also want to have crystals that are more "random". None of the crystals will be truly random, of course, but by having slightly unpredictable crystals, something pretty cool happens.

It essentially turns your dogs into ingredients as well. The results will always be deterministic and follow patterns, but you have to mix and match to know which direction things will go and it's kinda fun to play around with. I've got some ideas for more types of crystals but I can sort of do whatever I want with them so I'm not too worried about making them interesting. The bigger concern will be pacing out their delivery.

I also spent a little bit of time concepting out what the machine might ultimately look like (going for a sort of novelty bendy-straw thing with the connecting pipe) and I put in a very rough blocked out model just so I have a unique visual for it in the game.

So yeah. I'm happy with this system for now. It does everything I wanted it to do and it's functional enough that I can move on. Getting super close to the point where I can throw together a little bit of introductory pacing to test things out.

I made some updates to the cocoon flow this morning as well. It was always a bit irritating that dogs would automatically jump into their cocoons whenever the timer was up. They'd often be in the middle of something you were trying to watch, you'd often miss it, and it was super disruptive in general. To fix this I just made it so rather than automatically enter a cocoon state, dogs that are ready to mutate get a little UI pop-up you can click whenever you so desire. This also means that I was able to remove the "cancel" button from the cocoon UI, since you would've had to manually initiate it anyways. It's just generally a lot cleaner. I also re-did the hatch button so it's more clear. Still temporary I think, but it's nicer than before and should be self-explanatory now. Lastly, I added a cocoon portrait image that'll show up in place of the dog's normal portrait in the UI. Makes it much easier to tell at a glance what's going on.

That's about it for now. My departing gif this time is one of my favorite recorded moments in a while; preserved with the help of the auto-gif capture system I set up a while back.

I'm probably a bit biased but I think this is super cool. The evolved walks have a ton of personality to them and I love the look of these guys. The bendy legs, the little faces and tails... so good!

Trying to figure out a fun, accessible, and reasonably scoped way to actually package up a simulation like this is something I've been struggling with for a long time myself. It's a really tough problem to solve. I'd be curious to know how long it takes and how easy it is to actually evolve the sorts of bot movements you're showing off in the gifs. I definitely think that there's some amount of patience potential players will have but it all depends. If it takes 15 minutes, that seems pretty reasonable and I can see people really getting into it, specifically if there are actual goals and rewards. If it takes a few hours of training, that's a bit less appealing. Regardless, I think time becomes less of an issue the more player input and tinkering is required along the way. I'd be much more interested in spending a long time evolving a walk cycle if there was stuff for me to be actively doing rather than if I just kinda had to sit back and let things play out passively.

Anyways, going just from the results you're showing and from what I've seen previously of your tech (I'm assuming you're using the same base tech you've been working on for a while) I can absolutely see there being interest in this. I'm interested, at least!

I'd love to hear more about where you'd like to take this from here.

After a bit of fiddling and the usual issues, the base implementation of the Dog Swapper is in.

Obviously not the final UI but it does everything I need for now. Since getting this in I've been spending some time just playing around with the game and seeing how everything feels. Aside from design evaluation, this has led to some more bug fixing, which is nice. As far as the actual design goes, it's feeling like the right track but still isn't quite there. I like that the swapper encourages you to breed dogs for specific traits even if they have other less-desirable ones, and I like that it's imprecise and results in dogs that sort of end up as genetic dumping grounds. What I don't like is that there's nothing random about it. You know exactly what's getting passed over so there's nothing exciting about the swap itself. I think the solution might be to introduce some sort of controlled genetic drift. It has to be deterministic because I don't want the "optimal" solution to be just reloading until you get the dog you want, or something along those lines, but I think swaps need to have some sort of additional mutation associated with them. I'm not sure exactly how this needs to work but I think I could potentially give every dog some sort of genetic seed that could be used to mutate the swapped genes.

Another problem is that dogs mutate fairly slow. This works great when you're mutating a whole bunch at once, like when I was originally testing the breeding system, but it's not the best when dogs are mutating naturally through cocoons at a controlled rate. It's a tough problem because while I want pronounced changes, I also don't want dogs to change too dramatically with each cycle. If you leave things running for a bit and come back, the dogs will have started to really morph and I love the idea behind that slow burn of mutations, but I'm worried that it won't be exciting to open a cocoon if the changes aren't super evident. One idea I've been thinking about is to send the player an email "changelist" every time a dog hatches that shows before and after images and directly lists what mutations occurred. I'm not sure if that would solve things but I think it might be a nice start so I'll maybe try that out next week some time.

Anyways, that's the biggest stuff. I also spent a little time putting in a basic "attract mode" so I can leave the game running for a bit without intervention. The camera will randomly switch targets and zoom levels and do a bit of auto-rotation sometimes and cocoons will auto-hatch. There's some more I'd have to do if I was, say, gonna leave it up at a convention, but it's a nice start. I might try streaming it for a bit at some point in case anyone wants to leave some passive dog mutation up on a second monitor.

So I still haven't made it back to pill synthesis BUT it's been a really good diversion. I've been squashing bugs, making UI/usability improvements, and biggest of all I've got a new camera system set up.

I'd been hearing good things about the Cinemachine camera asset so I downloaded it to try out and was pretty much instantly converted. I'm very excited to never have to code a from-scratch camera movement system ever again. It's beautiful. That said, getting it fully integrated has been a decent amount of work because even though the base camera movement/collision logic was done for me there were still some lower level things I had to implement on my own (panning control, orbital sphere scaling, smoothly switching between targets, etc) plus I still had to actually, like, design the camera system and interface. Also, having a more polished camera had the side effect of highlighting lots of other bugs and weirdness so I used it as an excuse to tackle a bunch of that stuff as well. I can talk more about all this if anyone's interested but it was mostly just a bunch of tiny decisions along the lines of "what happens if a dog you're focusing on goes into a cocoon", "how does the user exit the follow cam", "how do I stop the dog name indicators from looking bad when you rotate the camera vertically", etc.

Anyways, since things are feeling so much better at the moment I recorded some footage of everything. Still some bugs visible in the vid and lots to do but I'm very happy with how it's all coming together.

I also went in and changed it so that when you pick up a dog with your mouse they no longer cancel their current behavior and turn off their AI routine. I'm really happy with this change so far because it means you can help them along to their goal without interrupting them completely. Plus they'll sometimes grab hold of things as you're dragging them around, which is pretty cute. Gives them a little more life. This wasn't really possible to do before all my bug fixes because there'd be too many situations where this sort of thing could result in them screwing up their targeting and walking infinitely into a wall.

Phew. So yeah I think I'm actually finally ready to get back to pill synthesis stuff for real now. That said I'm glad I held off on this because my designs for all that stuff have been marinating in the back of my head and I think I want to alter my approach. My previous idea was to slowly unlock genetic controls for you as the game progresses and you discover more mutations, but I don't think this is right. I think that if I give people direct part/slider control that people will very quickly stop caring about any of that. They'll try out the extremes and the new parts and then be bored. There's nothing interesting about trying to build a specific dog if you can just fine-tune it directly. There needs to be scarcity. The system needs to be finicky and make it tough to get things perfect. You need to have to compromise and deal with bad genes on your quest for good ones.

A big catalyst for re-evaluating this stuff was a competitive chicken breeding documentary I watched (Chicken People - 2016). You've got people dedicating their lives to raising chickens and breeding the Perfect Bird but that's only interesting because it's hard, if not impossible, to do. If perfection is available to you at the click of a few buttons, who cares? I think Petz and Creatures got this stuff right as well.

SO, all this said, what's my plan? I think that I'll no longer be giving people genetic unlocks. I think instead what I'd like to do is give people access to a device that lets them swap parts between two dogs, but on a high level. So you might be able to swap legs between dogs but you can't swap just the leg color, or just the length, or just the thickness. You get everything. I like the idea of people Frankensteining dogs together and I think it'd give a really nice balance between control and randomness without having to deal with the population and time problems that a traditional breeding system entails. You'll be raising dogs for specific properties despite other obvious deficiencies. A tail from one dog, legs from another, a pattern from a third, but oh no somehow in all the commotion your amalgamated pup's inherited an awful personality and isn't all that bright. Do you let a dog with a near-perfect pattern continue to mutate or is it too risky? Good scenarios, I think!

Still got some details to work out but I'm already way more excited about it than I was the previous design and the system is easier for me to picture in its entirety, so we'll see!

Got a bit sidetracked since last time with bug fixing so I haven't really touched the pill synthesis stuff much. I hadn't really taken the time to specifically fix bugs for a few months and stuff had really started to degrade. Still lots of issues I know of but after a few days of bug focus things are feeling a ton better. Nice to see my systems actually like, doing what I intended. To be honest it's been really motivating.

Most of the bugs I fixed were to do with dog behavior targeting; dogs were sometimes failing to acquire targets correctly and walking straight at point (0, 0, 0) or chain failing behaviors. I'd also screwed up some raycasting layermask code so they'd sometimes try to walk through walls or would march backwards through pipes, etc.

One pretty good bug I fixed involved dogs sleeping forever.

I also super quick added a new feature. Impact reactions.

Essentially, dogs now actually react when things whack into them. It makes them wince and has a slight negative effect on their fear/anger. This is a system ripe for consideration once I get back into dog personality implementations, but for now just this general case adds some nice life to things.

I was busy with Thanksgiving prep today and I'll probably take the rest of the week off. Hoping to resume work on the pill synthesis and genetic unlock stuff next week!

I'll leave you with this good Randy gif.

Regarding Patreon, here's an idea.  Alan Zucconi does tutorials and offers "Early Access" to $5 patrons and a few months(?) later he releases those tutorials for everyone. For example: https://www.alanzucconi.com/2017/10/27/carpaint-shader-thin-film-interference/

Yeah something like that isn't a bad idea, though I'd still feel a bit weird waiting months to show most things off. I've been getting a lot of feedback though and I've got some ideas for rewards I think. If I launch one it probably won't be until next year but I'm still thinking pretty seriously about it.

is it possible to create transparent dogs?

Not right now. I could definitely do it but transparency would cause some weird graphical issues with the way the dogs are set up so I'm not sure if I wanna go that route or not!

---

I'm still working on the pill synthesis and genetic unlock stuff. It's kinda slow going because there's a lot to puzzle out design-wise but I'll get there.

I also spent time fixing some pathing and behavior bugs (it'll never end). A few long standing issues (dogs freezing up if placed on top of an object that wasn't the stage itself, dogs trying to walk the wrong way through pipes), and a few old ones that drifted back a bit (dogs walking into walls).

Nothing too exciting, but now dogs can stand on things without getting stuck there forever and they're pathing much more nicely too.

Additionally, I played around with graphics today; experimenting with textures and normal maps. Probably not hard to guess that I've been playing the new Mario.

(Detail Crop)

I'm liking where things are going though there a few technical hurdles I need to get past to really do what I'd like. Namely I'd like to modify the existing shaders I'm using (which I didn't write) to be able to use world-space normal maps and textures (the walls and floors are made up of multiple objects). Seems doable but we'll see.

Aaaand, misc gifs.

hey! this project is really dope and heartwarming!

the physics and genetics and all the progress here really is impressive. 
through all your experimentation you seem to have an effective knack for clean, fluid functionality. the decisions to simplify certain things in order to maintain this functionality, makes for some really cool stylistic decisions (ie the rad food faucet thing, or the growth pods). i like that the interaction is turning out to be a bit funky, trippy, abstract. (love your obsession w the centipede legs too, lol)

at first in my mind i was imagining the game as some breeding or farming-focused simulator style ordeal, but especially since that epic 'failed prototypes' post, you've been showcasing some really fresh ideas. i guess with all the possibilities its hard to narrow down to what the game will essentially be, but keeping it open ended is certainly entertaining for the sake of the devlog!

keep it up!

ps please may i have loaf, please?    

Hey thanks so much! I'm trying hard blend traditional dog/pet expectations with some much weirder stuff so I'm glad that's coming through so far. Hope to take things much further once I've laid the gameplay foundation!

tom, have you considered making a patreon? a lot of people would be willing to throw money at you, and me too but i don't have any

I've thought about it before but what always holds me back is that I don't feel there are a lot of rewards I could offer people (especially in comparison to artists who give away drawings, pins, stickers, etc). I also don't love the idea of holding back info and making it patron-exclusive since I've been so open with development up until now, but I might have a few small ideas so we'll see what happens. Specifically because I feel I'm getting close to a point where I'd like to work with an artist or two to overhaul a few things and pump out some content, extra funds would be nice and I'm giving the whole idea some more serious thought now.

To everyone else, Ghost Dog is a great movie.

--

I've been working on the last big thing I wanted to do to solidify the pen flow. Directed genetics. Essentially, how do you control, or attempt to control, a specific dog's genetic drift?

I'm still figuring things out but the direction I'm exploring right now is a sort of pill synthesis machine.

Score one for system-reuse. Was able to repurpose the Dog Builder demo for this which is awesome. Essentially you can hook between 0 and 2 dogs up to this machine, muck with the genetics, and then spit out a pill that transforms whoever eats it.

I like the idea of giving direct genetic control to people, but it's not quite that simple because this is a game not a sandbox. I can't just give you total genetic access like this, there's gotta be something more. This is the part I'm still figuring out but my idea right now is that there's some sort of research mechanism at play here. Basically, genetic properties and sliders will all start off locked away and you'll gain access to them as you go through the game itself. If a dog mutates and gains a new property, like a new type of tail for example, then you'll have that unlocked for permanent use here. Otherwise, I'm currently trying to figure out a good system for researched unlocks. Playing around with the idea that you'll use dog eggs as essentially scientific gachapon tokens to gain access to new sliders and properties. Maybe some sort of internal rarity system with a mechanic where you can gamble more eggs for a better chance at rarer unlocks.

In any case, while working on this pill synthesis machine I ran into the same issue I had with auto-feeders which is that that I needed to add a new construction feature in order to place them down and save/load. For auto-feeders I added wall placement. For this machine I added floor placement.

After getting this basically working I decided to try and add a test cube object to let me build internal pen walls. There was a problem though... the spacing is all very dynamic based on the size of the pen and placed object which meant that I wasn't guaranteed to be able to place objects directly next to each other. So I overhauled the node placement code a bit, fixed some floating point math issues, and got the spacing set up such that object bounding boxes will essentially tile.

Of course, now there are other issues. For example, this system doesn't guarantee being able to place an object directly lining up with a wall, and the nodes are very visually crowded in some cases. I have ideas for how to fix these problems but things are fine for the time being.

So yeah, next week is probably going to mostly be me trying to design and implement this research/unlock feature. Once that's in, however, I might actually have a little gameplay cycle going??? Hard to believe...

Signing off with some assorted dog moments I've recently enjoyed.

It's always a pleasure seeing every dev update here.

Thank you!

---

Happy Halloween folks.

Glad you guys like it. I'm pretty proud of how construct mode's been turning out, though I already know it'll need a lot of love to make it more user friendly once I'm ready for people to put their hands on it.
--

Tested out the wall object flow with a new object just to make sure everything is solid in the general case. Few tiny bugs but at this point the wall object pipeline seems pretty good! Could be more efficient, but I'd rather not build content creation pipeline tools until I'm ready to really pump out a ton of objects.

Phew, just got the last of the auto-feeder content in for now.

The hold up was because I still needed to implement save/load. HOWEVER, in order to implement that I needed a way to actually put the feeders into the game other than just dragging the prefabs out in the editor, otherwise they wouldn't get indexed correctly and thus couldn't really be saved or loaded.

To do that I needed to implement a feature I've been wanting to get in for a long time, wall object placement. This involved some slight refactoring and lots of new construct mode code but I'm happy to say this feature is working and should support arbitrary new wall objects without requiring any additional code from me.

Wall objects were complex to implement because there's a lot they need to do. First of all they demand a completely new placement mode, where nodes show up inside of the pens themselves. In order to support this I had to add support for wall ghosting while in the editor so you can actually see inside of the pens you're trying to place things in. This also meant that I had to intelligently update the nodes depending on which walls were ghosted. Once that all worked I had to give the nodes directional and wall type information so that the placed objects oriented themselves correctly and knew which wall they were attached to. Then I had to make sure wall objects moved with the walls they were parented to, ghosted when those walls ghosted, and got destroyed when their owning pen was destroyed. I also did some work to space out the interior nodes so that every one of them is valid for the object you're trying to place. Essentially they have offsets so that you can place things on the furthest reaching nodes without your object clipping into the wall.

Then I had to get save/load working for these things, and THEN I could finally go back and implement save/load for the auto feeders themselves.

There's still a few missing features. Wall objects don't play nice with pen expansion, for example, but they're in a good place for now and I'm happy to move on. The other positive thing about all this is that it should be relatively simple to add in floor objects now that I have wall objects working. Looking forward to getting in some decoration content to actually take advantage of this stuff!

Thank you! ActualDog, "Okay, got it." I'm glad you appreciate the humor in that. That's the default response to everything to make the speech bubble go away but it's perfect as a response to being told a "stupid" joke. (I love jokes like that )

haha, yeah I figured it was probably the result of some sort of generic response system but that's part of why I love it. Anyways, the rest of the game looks great too. The ketchup and mustard snakes are another particular fav and I'm extremely into the "yes/no" rabbit ear dialogue.

This is looking killer. I really admire how all in you're going with the aesthetic. I think my favorite detail is the "Okay, got it." response to the flamingo's joke. Laughed out loud at that one.

I. Love. The cocoons!! Also v scientifically accurate, so thats nice

I also appreciate that you took dog lifespan into account, the idea of my presh pup dying of old age is deeply, deeply upsetting. I'm sure you can come up with creative alternative retirement plan if you really need to take wobbledogs away.

Glad you like them! I'm really fond of them too and I hope they stick around.

---

Did some more work towards getting the base implementation for cocoons all set up. The two big tasks were UI and save/load.

The trick with the UI is that I wanted to have it appear in-world next to the cocoons themselves. This is easy enough but it forced me to solve a problem I'd been dealing with for dog indicators as well. I'd previously been doing my in-world UI entirely in 2D. I'd just convert the needed 3D position to screen coordinates, scale things as appropriate with the camera distance, and do a some manual depth sorting. It mostly worked great but I couldn't have those bits of UI obscured by any 3D objects (at least not without some shader trickery I'm not really up to). It wasn't a huge deal for indicators but these bits of cocoon UI really needed to exist in 3D space, specifically in the case that there were a few hanging right next to each other.

Anyways, my solution wasn't too complicated. I opted to use Unity's worldspace UI canvases with some special code that billboards the elements and allows me to specify a distance to offset the elements along the vector from their associated object's center position to the camera's position.

Aside from accidentally billboarding the dogs themselves, I got things working pretty quick.

As far as the UI's design, for now I'm just keeping it dead simple because I know it's gonna change. There's a timer and then two buttons for cancel and hatch respectively. I expect to do more iteration on this stuff in the future.

The other, larger, task was to get the cocoons working with save/load. This is one of those sorta hidden costs to everything I do in this game. I'm trying to make it as easy for myself as possible to add new things to the save/load flow but complex objects still often demand complex solutions. The specific issues with cocoons were that they have a rather complex little statemachine structure (that isn't as clean as it should be) that I have to rebuild correctly on load, and then that they also can attach themselves to any object in the game, including other cocoons. I had to do a bit of load order restructuring to get this aspect working, the cocoons specifically load in 2 steps to ensure that everything's in the world before we load their attachment data, but I'm pleased to report that they save/load in every case I could think to test at this point.

After putting cocoons on the backburner, I moved on to another large design issue I've been wanting to solve. Feeding the dogs.

Basically, how do I make that fun? Previously you'd just have to keep watch on the pen and go into the inventory to drag out more food whenever dogs seemed like they needed it. This is fine, I guess, but it's not that interesting and I can only imagine it'd get kinda old after a while, especially if you have a lot of hungry dogs mowing through food at a fast rate. Also, how do I deal with food quantity?

For now I'm operating under the assumption that there won't really be money in the game. I don't think this game needs an economy (not to mention I definitely don't want to design one), and I think it'd be more fun anyways if you have unlimited quantities of most things once you've unlocked them so that you can experiment and set up whatever types of scenarios you want.

What I've got for now is an auto-feeder object. Basically a big pipe in the wall that spits out specific quantities of food and keeps track of those objects so it can spit out more whenever they've been consumed or otherwise destroyed.

I quickly mocked this guy up with prefab shapes because that's way faster and easier for me than using blender still, but it'll get replaced with something more legit at some point. You can click on the feeder to change the quantity and type of food it dispenses, and as the game goes on I want you to be able to unlock lots of different types. The idea behind the selection GUI is that it'll show you previews of all the types of food you haven't gotten access to yet, which'll hopefully be a nice little progress indicator (personally, I'm a sucker for this stuff).

The feeder still needs save/load support (which'll be way way easier than the cocoons), and I need to figure out how it interfaces with pen placement/building, but other than that it's pretty much good to go for the time being.

I also fixed a bunch of bugs over the past week, and this morning I was able to play through several cocoon cycles without any error or warning spew, which was pretty awesome to see. Obviously things still have a long long way to go and there are some large obstacles I have to clear, but the pens seem like they're really starting to come together and this is the best I've felt about the project in a while.

You could always pipe data to FFMPEG te create a video

Oh that's not a bad idea actually. Sounds like a good fallback option if I can't find an existing solution.

---

As I work on solidifying the base game flow I've been having to finally start addressing design problems I'd previously just kind of glossed over. First up is breeding and mutation. Basically, I have systems in place for these things but I don't have any solid way to interact with them. Eggs have obviously been in the game for a while but there are still issues there. How do dogs breed with each other? How do you deal with the population boom that laying eggs introduces? How do you meaningfully control or direct breeding? What's the lifespan of a dog? If the lifespan is too short will people get attached to them? If the lifespan is too long will people have any reason to play with breeding? How do I shorten the amount of time that breeding would normally take to experiment with? etc.

I don't have answers to all these questions but I do have a direction I'm exploring for now and that direction is cocoons.

The idea behind cocoons is to provide a way for dogs to continuously mutate as the game goes on without you having to churn through hundreds of individuals. I still think eggs will have a place in the game and I still think I'd like some sort of more traditional breeding but I really really want the ability for you to play long-term with specific dogs and I don't want that to lock you out of discovering new mutations. I know I just added mutation pills but I wanted a more natural and integrated solution too. What I'm thinking is that throughout the game, dogs (assuming they're in healthy shape) will continuously retreat into cocoons and mutate. I think I'll provide a way for the user to manually cancel the mutation in case they'd like the dog to stay just the way it is, but I also want to let you inject genetic mixtures into the cocoons to direct the mutations. Maybe you can even take DNA directly from another dog and crossover the genes that way.

Still exploring specifics but even if I don't end up using the cocoons as a recurring mutation vehicle I have other ideas for ways to get them in and I really love them. I think they just generally lend themselves to fun situations.

Quick technical details in case anyone's curious. The cocoons do a raycast straight up from their attachment point and attach to the first thing they come in contact with. I get them to hoist themselves up by slowly decreasing the joint's linear limit. The strings themselves are drawn after the fact. They're just a thin, white, prefab cylinder that I scale to the desired length and then rotate to intersect both the joint anchor and the joint attachment point. It was pretty much the dumbest way I could think of to do them and it works great. It'd be fun to have them actually collide with things but that creates other issues and sorta highlights the fact that they don't bend so I'm happy with leaving them alone for now.

Other than cocoons I've also been fixing some old bugs and addressing some tech debt. Biggest thing I've tackled is my object location validation and good location finder system. I've known for a while that it needed to be shored up but while working on cocoons my dogs kept spawning outside of the pen and it was really frustrating to test so I spent some time improving things.

Slowdown here is from the debug lines not the location system btw. The system now lets me specify a specific pen to contain the given object inside of (or to just keep the object inside of whatever pen it started in), and it way more accurately and efficiently searches the world for a valid location to move to. I essentially do an incremental 3D flood fill using the object's bounding box and stop searching in any given direction once it results in the object leaving the needed pen (assuming one is specified). It also obviously stops things from moving to places that'd clip with other objects. There are ways I could optimize things even more but until it becomes a problem I'm pretty happy with this current solution. Did some stress testing this afternoon and it's feeling pretty solid. Every one of the balls in this gif spawned without intersecting with an existing one.

And that's about it for now. I have a bunch of cocoon-related tasks left still to finish (UI, tuning, game flow and design, save/load) but I'm very pleased with them so far. Looking forward to getting them better integrated.

Thank you so much Tom! It is always super inspiring seeing wobble dogs and the new creature - creature interactions between dogs. We really want to focus on creature led puzzle design and consistent behaviours that players can begin to learn. Will you be at IndieCade this year? Any tips for someone going for their first time?

Aw, thanks. I'm not sure yet if I'll be at the actual event (definitely not the conference but maybe the arcade), but if glitch city hosts a party again this year I'll probably be at that. As far as tips, I'm probably not the best person to ask, tbh! I don't have a ton of experience exhibiting and I've never exhibited at indiecade before. It's a cool festival though and I'm sure you guys'll do great.

This is looking really charming and wonderful. I love that there's lots of different critters instead of the levels just being purely environmental. Good luck at indiecade!

i love the idea of dog commercials, i haven't seen anything like it before!
if you add a sandbox mode i hope we can make our own custom commercials

Thanks! Dunno if I'll be able to support that but I'd like to try and make it as easy as possible to set up new variations without touching much (or any) code, so we'll see.

Looks like the final idea is (in a nutshell) a good framework to be able to create "mission" style levels.

Yeah, exactly!

Add the possibility to export the resulting commercial and you have a great starting point for encouraging a community of players to share their work - as you probably know better than anyone here, people sharing their mods with furniture and whatnot were one of the things that made the original Sims so special.

Make the game automatically record the commercials as a ready-to-share video/GIF. YOU WILL MAKE SO MUCH MONEY TRUST ME.

There's a lot of other things that are higher priority for me right now but I'd love to try and get some sort of in-game sharing working! I don't know very much about in-game footage capture but hopefully it's not too wild to set up or at least there's maybe some sort of asset store plugin I could use. I'd looked for gif solutions earlier and didn't find anything great but maybe video would be more reasonable? Definitely gonna look into it more at some point.

I absolutely love this concept! It is extremely original and seems to be a good counterpart to the dog breeding/raising. I must ask, are hats confirmed? Can dogs wear hats? Looking back I see you mentioned the accessories system, that is extremely exciting! Also I like the idea that you could involve some of your earlier gameplay loop concepts in the commercials one. A dog racing commercial may require the dog to come first. A... capsule laying commercial... could require the dog to lay a capsule. I love the idea that there are a world of strange dog commercials to go along with these strange dogs.

I'm glad! Hopefully I can strike a nice balance between having very distinct-feeling commercials and being able to make them relatively quick. There's a lot of ideas I wanna explore. As far as hats go I don't wanna say hats are "confirmed" because there are some issues I still need to sort out but I'd really like to have them in the game so we'll see!

If you release on early access I think you will  

Ah, I've thought about early access a lot. I think the game could benefit a lot from it but there's a part of me that really wants a full initial release with all the resultant fanfare, etc. We'll see... there are benefits to both sides for sure.

--

Small update. I may have a direction I'm shooting for now but there are still a lot of base systems and features I need to implement or shore up before I can really get into making content. The past few days I tackled persistent dog ownership.

Previously I'd only save dogs that were actually in the world. If they got deleted for some reason, that was it. Now there's a temporary UI window where you can take your dogs in and out of storage and look at some of their stats. If dogs get deleted they go in the box for you to take out later. This sounds simple, and maybe it should've been, but it ended up touching a ton of systems. For example, loading a game with a bunch of dogs in storage means I need a way to get portrait sprites for all those stored dogs. Generating portrait sprites isn't the quickest thing so doing it for potentially tens or hundreds of dogs would be horrible. I added support for sprite serialization and now I serialize and reconstruct portraits rather than generating them from scratch every time and it's much faster. Of course this meant I needed to update my save/load structures and actually build a solution for sprite serialization (which I still need to add compression to). I also had to improve my portrait caching system, which was trickier than I thought it'd be, and I took the opportunity to add timeslicing to dog and portrait creation so that it doesn't give huge frame spikes. Things are in a pretty good place now, though they'll definitely need more love in the future.

Anyways, next week I'll probably try and tackle some slightly more interesting systems, which'll be nice. Like I said last time, I've got a lot of work ahead of me!

Anyways, have this gif.

[Failed Prototype #1 - Dog Racing]

I think the similarities with The Sims' UI could actually play in your favour, because Wobbledogs looks like the kind of game that can also appeal non-gamers. And The Sims is, by definition, the game even my mother tried to play once. So by looking similar it could actually feel less "menacing" and complicated, and help you reach a bigger audience  

Yeah agreed, it's something that's worth leaning into to a certain extent.

this game looks good

Thanks!

I plan to eventually have multiple head types.

I misread this and pictured dogs with multiple heads for a moment. It was glorious.

We need Cerberus the Wobbledog!

a system where identical twins could be born out of one egg and in rarer cases conjoint twins would be really interesting

lol you guys are figuring out all my secrets... I've had this on my project tracker for a while now to try out at some point (we'll see what I can make work). If I do end up putting it in I probably won't show it off before release though. Gotta keep some mystery going.

---

Ok, I think it's time to unpack some stuff. Anyone who's kept up with this thread is probably at least somewhat aware of how much of a tough time I've had trying to figure out what specifically you do in this game. I've tried a bunch of stuff, none of which has really stuck, and at more than one point I even said I was just gonna ignore gameplay all together in favor of a sandbox-first approach. I think I've finally found an approach that works so I wanna go over everything I've tried before, explain why I gave up on those ideas, and then go over what I'm gonna try to do instead.

So first of all, the failed prototypes.

Failed Prototype #1 - Dog Racing

Concept: Breed and train dogs for their racing prowess and watch them compete.

Positives:

• Fun to watch
• Takes advantage of physics
• Once the systems are in place, courses should be quick to make
• Could add a track editor for user-generated content
• Casual gameplay

Negatives:

• Only fun to watch the first few times or in short bursts. If your goal is to win rather than just observe it becomes even more tedious
• Very hard to tie the pen gameplay together with the race gameplay. How do you meaningfully raise a dog to do better at certain courses when their performance comes down to their base physics?
• My physical movement makes for very random outcomes even for more optimized dogs

Let me just start by saying that this prototype is still the closest to my heart. I honestly loved watching the dog races. The main problem, however, was still that the moment you needed a specific dog to win things stopped being entertaining. Races could take a while and retrying them was annoying. They were only really fun to watch if you didn't have much of a stake in who won. With my inconsistent movement physics I just don't believe I could've made this mode stay interesting for a whole game.

Failed Prototype #2 - Economy

Concept: Build and optimize a self-sustaining economy around your hive of dogs and their unique assets.

Positives:

• Complex automation gameplay
• No additional game modes needed
• Near infinite gameplay

Negatives:

• No moment-to-moment goals
• Potentially conflicting audiences
• Moves focus away from the dogs and their personalities

The idea here was to use the dogs to build a little self-sustaining economical system. Stuff like having the dogs lay eggs which auto-collection pipes would suck up and move to incubation chambers which you could then hatch or sell immediately. The main issue was that it took the focus away from the individual dogs and turned it to a larger overarching system. I also never came up with great ideas for more interlocking systems beyond the eggs. There was some stuff I floated around like the idea of breeding dogs for extra legs or fur, etc, but nothing ever really clicked. I think ultimately there's a disconnect between how long it takes the dogs to do things (because of their physical movement) and the efficiency you actually want for stuff like this to feel right (sort of a recurring theme with my prototypes). Anyways, I scrapped this as a core gameplay concept but there are elements I might still re-use, like the eggs themselves.

Failed Prototype #3 - Wilderness

Concept: Explore hand-crafted puzzle environments with your dogs.

Positives:

• Puzzles and exploration
• Opportunities for deep discovery and lore

Negatives:

• Very time consuming to set up levels
• Trying to directly control or issue commands to these dogs is aggressively not fun

The last point says it all, I think. Watching a physics dog try to do something on its own is entertaining even if (or especially if) they have a tough time doing it. Trying to get a physics dog to do a particular thing is extremely not entertaining if they have even the slightest bit of trouble.

Failed Prototype #4 - Arena

Concept: Dogs gain abilities based on their genes and then you send them to an arena to duke it out with other dogs.

Positives:

• Character and story opportunities
• Wonky physics aren't a huge issue because abilities can be hand-crafted and magical and separate from the dogs' movements.
• Fun opportunities to import and battle against other people's dogs.

Negatives:

• Violence, even if cartoony, felt a little at odds with the rest of this game's mood
• Prototyping was extremely time-consuming and slow. I didn't believe I'd have a good idea of whether or not this concept would work without at least a month of dedicated time.

This just didn't feel right. I couldn't get over the fact that forcing you to fight your dogs felt wrong for this project and I didn't want to spend an extended period of time on a prototype that might not even pan out.

Failed Prototype #5 - Tricks

Concept: Teach your dogs tricks.

Positives:

• Fun moment to moment possibilities
• Player-Dog bonding

Negatives:

• Couldn't come up with a meaningful way to implement this system
• Even if implemented perfectly it wouldn't be enough to drive an entire game

Thinking simultaneously too large and too small. Procedural physics-based tricks would be a huge huge complex system but even if I made it work it probably wouldn't be enough to drive the entire game.

Failed Prototype #6 - Dog Streaming

Concept: Each dog's life is being livestreamed to an in-fiction twitch platform. Gameplay revolves around keeping your viewers happy and collecting donations.

Positives:

• Funny concept
• Potential for funny choices. (Your viewers love it when you put squeaky shoes on your dog but it drives you crazy. Your viewers love a super fat dog but it's too fat to move so you have to painstakingly take care of it. etc.)
• Good high level goals

Negatives:

• Hard to define the system feedback loop
• Moment to moment gameplay still lacking
• Wasn't super excited about building it

Getting closer to something, but this didn't quite pan out. Ultimately I couldn't reconcile the actual goals with the player's desire. If the stream was always running you felt no pressure to do specific things. You just let the dogs do what you'd always let them do only now you don't even get to try specific things you might want to because you're beholden to "the viewers". If the stream only goes live at certain intervals then those specific moments would require boring setup from you (make sure the dogs aren't sleepy, put them in the right clothes, change your pen's wallpaper, etc) and you'd probably just end up dreading it.

Failed Prototype #7 - Capsule Laying

Concept: Dogs lay capsules periodically. Depending on a dog's genetic makeup a capsule will have a different pool of items it could potentially contain with corresponding rarities.

Positives:

• Simple
• Slot machine feedback loop
• Good high level goals (breed certain dogs to get certain items)

Negatives:

• Zero moment to moment engagement
• Hard to define which genes should map to which items/rarities

Not much to say about this one. It just ultimately wasn't super interesting. Good UI could've helped that a bit but it would've only wallpapered over the boring interior.

Failed Prototype #8 - Custom Scenarios

Concept: Click through a list of hand-crafted scenarios and poke around with them as much as you'd like before moving on. After viewing a scenario you unlock all the items used within it.

Positives:

• This is what I do anyways for setting up a lot of my gifs
• Can show the player the limits of all my systems

Negatives:

• How do you know how long to engage with something before moving on?
• What's the incentive to actually engaging with anything?

So close, but not quite. I had a bit of an epiphany at one point and thought to myself why not try to merge the game and the gifs I've already been making? Why not just set up a bunch of scenes that are literally just the scenarios I already create (sometimes on purpose and sometimes on accident) and share with everyone? Ultimately, it's still a little aimless and not quite worth going forward with, but this'll lead me to where I am now.

Current Concept - Commercials

(FYI: the graphics for the systems in this vid are almost entirely placeholder)

Concept: Raise your dogs and periodically employ them as actors in various commercials.

Positives:

• Funny concept
• Potential for a cast of characters and an actual story
• Show the player the limits of all my systems
• Drip-feed new items and unlocks as payment for commercial participation
• Casual, scaleable, and played out in very short bursts
• Gives breeding a point (certain commercials could have requirements)
• Space for fun and meaningful choices (e.g. we want your dog for this commercial but have to dye its fur first)

Negatives:

• Lots of work!

This is where I am now. You can probably see that this concept borrows a bunch of things from previous concepts I've explored, though it took me a while to get here specifically. After a few weeks of thought and prototyping I'm still feeling good about it too. It hits all of the major points I wanted my gameplay to hit and it just feels like the right direction to take things. The idea popped into my head late one night a few weeks ago and it immediately felt right to me. With this structure you'll always know your next goal, it'll slowly give you new objects over time, and you'll get a steady stream of funny scenarios that your dogs can literally star in.

Like, really, these types of things should just be in the actual game IMO, and with this structure they can be.

There's more I can talk about in terms of specifics but I think this post has gone on long enough. Basically, this feels like a game I know how to structure and make. It's gonna need a lot of work but there are some ways I can make things easier on myself, so I'm not overly terrified. I'm sure this concept will warp a bit over time, and it's possible I'll throw it out entirely, but it's absolutely the closest I've come to something legit so far and for now I'm gonna forge on ahead and see where things end up!

Hi, I'm in the same boat re: making an account just to comment here. I've been following Wobbledogs for a long time and love all of what's going on!

I was curious how in-depth (if at all) any genealogy might be? Some way to keep track of all of a dog's ancestors on a family tree or something? Or a way to commemorate them (pictures or something) without having to make your own family tree in a separate program? I have no idea how hard that'd be to code, and would definitely understand if that's not a possibility. Tracking family trees was just always one of my favorite features in the Sims and I was curious if it'd show up here, too!

By extension, will related dogs able to breed? Will there any penalty for this? Or will I just have to keep super close watch on my dogs if I don't want that happening?

Thanks a bunch for all the work you've done so far, I love every one of these goofy dogs I see!

Glad you like them!

I haven't given a ton of thought to the nuances of genealogy yet but some sort of family tree could be pretty cool. I don't want to promise any features at this point but I'll definitely keep that in mind cuz I can totally see why people might want something like that.

As far as breeding goes, I probably won't stop related dogs from being able to breed but I dunno yet if there'll be specific inbreeding penalties.

The specifics behind breeding and genealogy in this game will need a very in-depth pass at some point later. The systems are baseline functional at the moment but they're not very well integrated into the full experience. All these points are great things for me to keep in mind when I get deeper into this stuff though!

What about giving accessories different sizes, so a dog wouldn't be able to equip them if, say, their head is bigger than a certain number?

Maybe, but it's not quite that simple. Head size isn't actually too big of an issue; the bigger deal is snout size/rotation plus the fact that I plan to eventually have multiple head types. It's just hard to define in-code what will and won't clip with all these variables. I have a few ideas but ultimately it might just be easier to live with some clipping and let the players decide if they want their dog to have a wonky hat or not.

Yaaay! Thats a relief : D Glad you're keeping it in mind!! I'd suggest something like just having them getting bloated and flopping over but then you can't have dog vomit-eating-mechanic I suppose. Good point with the sound too, the sound is usually what triggers the WORST bit of empathetic-nausea, the like, gushing glorby liquid is also really bleghblegh, could do Cotton-Candy-Fluffy-Puke-Mode :D

SO EXCITED to show this game to my friends who dont normally play games!! I know theyre gonna be hyped to take care of weird-puppers : >

Glad to hear it!

The current UI, including the font, is such a throwback to the original Sims game—which is fitting, since IIRC you said you've worked on one of those?

The current UI, including the font, is such a throwback to the original Sims game—which is fitting, since IIRC you said you've worked on one of those?

Oh interesting, I hadn't realized that until now but you're totally right

@Prinsessa: I noticed the same thing a few pages back. That means there's two fans of the style by now ActualDog, you can't go back

lol, well I'm TRYING not to make it too similar but yeah, there's definitely some inspiration. To some extent I think the similarity is kind of a good thing because it gives people a specific initial impression of how things function but to that end I also think things will start to drift a bit more as time goes on and I'll end up adapting the UI to be better suited towards this game's specifics. Part of the similarities come from the fact that the game doesn't have a super defined loop right now to inform the functionality.

You're mad dog. Keep this project going!

It's going! It's tough to work on sometimes but it'd be even tougher to shelve. Gotta finish this thing...

--

No huge update today but I have been working on some new stuff. I know I JUST said that I was gonna double down on the sandbox elements of the game going forward but I had an idea literally pop into my head out of nowhere late one night last week and I've been working on defining and prototyping it since. It's, unfortunately, a lot of UI work (like everything I try to do in this game) so it's not too exciting yet but we'll see where it ends up. At the very least it's absolutely the closest I've gotten so far to something legit and it checks off all the boxes I've been trying to hit with the gameplay loop. I know I'm being cryptic but hopefully I'll have more to show soon.

In the mean time, a collection of loose-leaf gifs I hadn't previously posted here.

the pills are a nice idea. and they made me think about that i don't know whether i want to take care of the dogs as pets or experiment on them. that could be a gameplay thing, like getting rewards for doing tests and dog experiments, but the rewards shouldn't be big important things so you can play without having to do all that if you don't want to. and breeding specific kinds of dogs could be an optional goal too.
also, i like how the hat just says DOG on it.

haha, thanks. I've actually got some stuff cooking that might hit some of those points you brought up. We'll see how it shakes out in the next week or two!

I made an account just to comment here... I've been following this thread FOREVER and I absolutely love your doggos!! Every minute this game gets closer to being a game I can play at home the more excited I get : D

Your more recent update has me a bit worried though, I've always been really excited to share this game with friends of mine who aren't into shooters/beat em ups etc, but the addition of dog vomit has me concerned.. I am the sort of person who has a like, reaction to hearing/seeing vomit that also makes me SUPER sick, and I know a bunch of other people have had the same problem with Rick & Morty (Which is why they've toned it down)... so I worry dogs puking will make me unable to enjoy the game as much : ( 

Just food for thought, as its something you may not have considered! (Not here to tell you what to do/not to do, just in case you hadn't considered this viewpoint/that this is something some people have)

SO HYPED FOR PETTABLE DOGS!! GONNA PET EM!

Glad you like the dogs so much! You bring up some good points re: dog vomit. Although real life dogs are gross (they're great but they barf on your carpets and eat poop etc etc) and I want to capture some of that, my intention is DEFINITELY not to make people sick and what I'll say baseline is that I really don't expect dogs to be barfing all over the place. The idea behind the feature is that it provides a bit of pushback towards dogs overeating. One more thing to keep an eye on with some real systemic feedback if you don't notice. THOUGH I do also want dog sickness at some point so it could make an appearance there as well, I'm not sure.

I want it to be a little bit gross, but I also don't want it to ever be a reason for someone to not play. One thing I'd been thinking about is making the barf the color of whatever the dog had just been eating. That might help lessen the gross out factor a bit, especially if I mixed that with some sparkle particles or something. Also, I'm sure the eventual sound design will have a big impact on how things come off so I'll be sure to be wary of that when the time comes!