next »

[Cosmocat Games]

And I'm really glad that the multi-level generation for the most part "just worked"! It's generally quite lovely when that happens, I feel. ^_^

It is lovely though I think in my excitement I under-estimated how much work is actually left Because in addition to connecting the floors, adjusting the culling, and water, there's also stuff like the mini-map, the "generation variant" states (i.e. to avoid generating two similar levels), and probably other things I'm not thinking of. So probably still at least a few days of work, there, but yeah, there is light at the end of the proverbial tunnel so I'm still excited!

Well, I've made some solid progress in the past few days, including a couple hours on today's stream.

Here's an overview of the multi-floor generation plus a (mostly) functional version of multi-floor culling:




And some initial progress on adding multi-floor support for the minimap:




And most importantly, a brief cameo from Alastair. Thanks for the help, buddy!
https://www.youtube.com/watch?v=t51IW9gvjds

I also posted my first official update on Steam in quite some time, giving a state of the union and an overview of the plans to move forward.

[Thaumaturge]

Cat-like typing (and music) detected.

I'm glad to read that you've made solid progress, including on the minimap! ^_^

(If not for your having made progress on the minimap, I might suggest simply providing one map per floor, and allowing the player to page between them at will.)

[Cosmocat Games]

Cat-like typing (and music) detected.

I'm glad to read that you've made solid progress, including on the minimap! ^_^

(If not for your having made progress on the minimap, I might suggest simply providing one map per floor, and allowing the player to page between them at will.)

Actually that is how I'm approaching it (one map per floor), minus the ability to page between them. This is partially because the map is only available in-game and a paging mechanic may be a bit too "menu-y", but also because I don't really see the value in paging between maps--at least not yet--since they mostly just show the level layout (and the path to the next floor, if the compass item is found). The objective in all cases is to reach the top.

That said, if back-tracking becomes part of the design then I'll likely need to revisit, but as it is now, I'm thinking back-tracking will be purely optional.

Some more progress since the stream:

• I actually had the minimap basically working at the end of the stream, but it was a misplaced ! operator that was causing everything to fail (classic!)
• Multi-floor water was dead easy (one line of code, since I'm already tracking which floor the player is currently on). The only limitation is that I can't have water span floors.
• I've also started doing some optimization (converting ProBuilder objects to Blender models, merging meshes, etc.). Along with this I integrated my newly unwrapped/custom textured support beam model from the other week, and downloaded some lower poly Megascans stuff to fit the new aesthetic.

(side note: I'm getting more and more comfortable doing utilitarian type stuff in Blender, like setting origins, applying transforms, quick UV edits, etc.--this is going to be huge for my productivity going forward)

So really, the main things left at this point for multi-floor generation:

• Determine optimal level size (number of modules, floors) vs number of levels
• Add some logic to increase difficulty with each floor (currently, it's global/per level)
• Implement the actual inter-floor mechanic (i.e. interacting with something to open up the vertical path)

Of course I have lots of ideas to build on top of this, but this should at least get me to a pretty solid foundation.

I've got an 8 day break coming up so I should be able to easily tackle this (and much more) during that time. If I'm really able to focus, perhaps I can even have an initial early access build ready for the new year!

[Thaumaturge]

That does sound pretty good, I do think! I hope that development has been going well during your break. ^_^

[Cosmocat Games]

That does sound pretty good, I do think! I hope that development has been going well during your break. ^_^

Yup, good progress! (interspersed with breaks to play with my new toys--heh)

I mostly stayed off internet/social media during my time off hence the lack of updates, but I'll have a big one coming soon.

[Sayrui]

God, I already wishlist on steam, good luck with it  

[Cosmocat Games]

So, the holiday sprint is complete! Plenty of distractions, mind you--new books, games, etc.--but I still managed to make some pretty solid headway and more or less achieve what I had wanted to. (and in fact, it's been a while since I've "binged" anything; probably not a bad thing to spend the holidays doing some actual relaxing!)

Fully integrated multi-floor levels
For this I:

• Added two more floors for a total of four per level (and four levels total, at present)
• Implemented multi-floor difficulty logic: in addition to the global "difficulty" that's incremented with each level, there's now a separate "adjusted" difficulty used for the generation of each floor (adjusted difficulty = base level difficulty + floor number). This ensure the levels gets gradually more challenging, while still retaining a reasonable difficulty curve for each.
• Integrated the multi-floor scene into main game flow
• Re-integrated the old exit modules for the top-most floors (i.e. interact to proceed to next level)

Progressing between floors
To facilitate progressing between floors I added the ability to interact with the ramps that connect the floors (using the rope item, as previously mentioned). I ended up removing the climbing/sliding mechanics I had been playing around with (for now), as they weren't really needed (yet). Instead I'm just using a collider to block access until the rope is placed.



I had to do some work on the interaction system to get this going. It was previously based on the angle of view relative to the center of the interaction candidate object. This was problematic for large or oddly shaped objects though, as the "center" didn't necessary reflect the actual shape or size, and as a result there were edge cases where the angle of view and/or raycast distance could never actually reach the center of the object. To remedy this, I'm now using the center of interaction candidate object's collider instead.

There are still some quirks with the exit interaction, like if you get too close you lose the interaction prompt, but this should be an easy fix. (and besides, this is all just temporary: I plan to have something like a "rope anchor point" that you'll interact with instead of just an arbitrary collider at the base)

The rope is currently a single-carry item (one rope per floor, one exit per floor) but I do plan to look at opening this up some more during early access (i.e. extra hidden ropes, secret areas that can only be accessed by using a rope).

Optimization
I'm taking every opportunity to optimize things, and it's timely since the multi-floor generation has effectively quadrupled the size of the scenes. In addition to merging meshes where possible, I've recently done some enhancements to dynamic occlusion culling system to reduce the number of raycasts. Essentially, I don't bother with doing a raycast for objects that are beyond the view distance of the camera.

Side note: the dynamic occlusion culling is based on a third-party asset, but I've basically made it my own by fixing a number of issues, adding additional performance optimizations, etc. The author has basically stopped maintaining it and since made it free (annoying, because I paid for it!), but he's talking about open-sourcing it, in which case I will either see about contributing my changes to his repo, or forking and making them available there. Stay tuned...

Unplanned Unity upgrade
A couple nasty issues (builds completely freezing my PC, skinned mesh collider bounds being calculated incorrectly) forced me to upgrade to the latest Unity 2018.4 LTS version. It's entirely possible that it was the asset reimport that fixed either or both issues, but hey, it's good to be on the latest 2018 LTS version, and it's something I had been planning on doing anyway.

You've reached the end of the mine...
Since there's no real ending yet, I've added a note that can be found on the final floor of the final level. Kind of half journal entry/half breaking obliterating the fourth wall:

June 3, 2001

We can see daylight but can't find any way out. I guess we'll need to wait for a new early access build...

Putting it all together
Here's a video of me running through the first level. You can see the difficulty progressing with each level (blind mutants -> normal mutants -> toxic), and the multi-floor stuff in all its glory.





The path to early access
So it's basically playable from start to finish in its current form. There are a few odds and ends I wouldn't mind dealing with before uploading the first early access build to Itch, but I would say it's all pretty minor stuff in the grand scheme of things so I may just give myself a deadline of a week or so and say "whatever's done is done". It is early access after all!

[Thaumaturge]

Very cool! It really does look like this is coming along very nicely indeed. ^_^

The rope is currently a single-carry item (one rope per floor, one exit per floor) but I do plan to look at opening this up some more during early access (i.e. extra hidden ropes, secret areas that can only be accessed by using a rope).

A thought: it might be interesting to build some player-decisions on top of this. Perhaps there are more side-paths/alternate-paths than there are ropes, leaving the player to decide which way they want to try. Perhaps different side-paths provide different benefits--I imagine for example the player having one spare rope on hand, faced with the decision of whether to use it to climb one slope leading (potentially) to health-items, or another leading (potentially) to weapons.

[edit] You know, it occurs to me that while I've seen plenty of "toxic", I have yet to see any "ant"s! I am very disappointed. ;P

[Cosmocat Games]

You know, it occurs to me that while I've seen plenty of "toxic", I have yet to see any "ant"s! I am very disappointed. ;P

Hopefully you know that's not what I'm going for, but now you've got me paranoid that other people are going to think it's a game about toxic ants! That said, I do want to incorporate more insect-like creatures, and I'm thinking toxic ants would actually make a lot of sense, but then that makes me even more paranoid...

In other news, I've been comparing GIF capture and editing tools lately. LICEcap has generally been my workhorse in the past, but it has its limits (no ability to convert from video, no ability to edit, no control over encoding) and often this results in me having to do several takes before I get something that works.

My most recent search has been for a tool that excels at video conversion. So far, the (apparently lesser known) Easy GIF Animator seems to be the winner. It converts from video without issue, the GIFs are of decent quality, reasonable size, and there's enough control over encoding (not quite as much as I'd like, but likely enough for my needs).

Here are a few of my test GIFs with different export settings:







(I didn't end up using this particular GIF for anything--the footage is too dark--but it should give you an overall idea)

It's not free, mind you, but given how integral GIFs are to my marketing strategy, I have no issue shelling out $30 for something like this.

Other options I evaluated: ScreenbToGif, Giffing Tool, and of course LICEcap. I won't go into detail about why these didn't work for me, but if anyone is interested let me know and I'll gladly share my notes.

Secondarily, I'm looking for a GIF editing tool that allows me to edit without re-encoding. I could have sworn I came across one at some point, but none of the tools I've recently evaluated seemed to support this. (if anyone knows of one, please let me know)

[Cosmocat Games]

Really good progress yesterday! I spent a fair bit of time trying to make the multi-floor connectors look like an actual thing. To that end I:

• Updated the wall texture to match the rest of the environment
• Added some support beams
• Added some additional wooden platforms and ramps higher up
• Replaced the placeholder rope texture with something slightly better (based on the rope pickup)
• Added some lanterns

Still a heavy WIP but in the end I hope to really tap into some DS inspiration (Blighttown/Valley of Defilement)--rickety wooden structures and the like:



The above was captured with "Unity Recorder", something else I'm trying out. It allows me to capture footage straight out of the editor. You can't tell from the recording but it slows the editor WAY TF down--slightly distracting--but the actual footage comes out nice.

In this case, I recorded the video at the exact resolution and framerate that I wanted for the GIF (640x360 @ 12 FPS), and then converted with Easy GIF Animator with no downscaling or frame reduction. I think it turned out pretty well.

Separately, I've been toying with the idea of flashing some random images at certain points in the game (when the "ghost" enemies hit you and/or when I trigger random spooky noises). Originally I wanted to have animated ghostly visions of miners working, but this would likely going be time and/or cost prohibitive, so I was thinking maybe old still images of miners. The only thing is that the game is set in ~2001, so I worry it might be a bit out of place.



I welcome thoughts on this.

Phew, k, back to work!

[Thaumaturge]

Hopefully you know that's not what I'm going for, but now you've got me paranoid that other people are going to think it's a game about toxic ants! That said, I do want to incorporate more insect-like creatures, and I'm thinking toxic ants would actually make a lot of sense, but then that makes me even more paranoid...

Oh dear, I am sorry! I didn't mean to have that effect--I was just playing with the words. ^^;

Honestly, I doubt that many people will seriously expect there to be ants--the "-ant" suffix isn't exactly an unknown one, after all. I would expect more jokes about it from others, however!

Secondarily, I'm looking for a GIF editing tool that allows me to edit without re-encoding. I could have sworn I came across one at some point, but none of the tools I've recently evaluated seemed to support this. (if anyone knows of one, please let me know)

I see that you've tried out using a general video recorder, then converting to gif; have you considered using a third-party video-editor to then edit before the final conversion? My workflow is something like that (specifically, OBS->Blender->ffmpeg), and I do like that it allows me to cut, crop, scale, and otherwise edit.

Really good progress yesterday! I spent a fair bit of time trying to make the multi-floor connectors look like an actual thing

I do like that new connecting shaft! Nicely done, I think. ^_^

Separately, I've been toying with the idea of flashing some random images at certain points in the game (when the "ghost" enemies hit you and/or when I trigger random spooky noises). Originally I wanted to have animated ghostly visions of miners working, but this would likely going be time and/or cost prohibitive, so I was thinking maybe old still images of miners. The only thing is that the game is set in ~2001, so I worry it might be a bit out of place.

Hmm... I'm not sure of how I feel about using actual photographs of flesh-and-blood people in this way. If you were using whole-cloth art, then it might not be a bad idea, perhaps--echoes of the old days, before things went horribly pear-shaped...

[Cosmocat Games]

Oh dear, I am sorry! I didn't mean to have that effect--I was just playing with the words. ^^;

All good

I see that you've tried out using a general video recorder, then converting to gif; have you considered using a third-party video-editor to then edit before the final conversion? My workflow is something like that (specifically, OBS->Blender->ffmpeg), and I do like that it allows me to cut, crop, scale, and otherwise edit.

Yup, I do that from time to time, especially for like "hero GIFs" (with multiple shots), but it would also be nice to be able to capture something quickly with LICEcap and trim it without having to re-encode.

Hmm... I'm not sure of how I feel about using actual photographs of flesh-and-blood people in this way. If you were using whole-cloth art, then it might not be a bad idea, perhaps--echoes of the old days, before things went horribly pear-shaped...

Not familiar with the term "whole-cloth" art". Do you mean like drawn or painted? Stylizing may be an option here, and/or if I were to downscale and point-sample them they'd get a gritty pixel-art-esque aesthetic which might fit a bit better.

[Cosmocat Games]

I spent some time last night adjusting the enemy spawning progression to work a bit better with the new multi-floor layouts, including:

• Not spawning any enemies (except for ghosts) in the very first floor; I think it sets the mood quite nicely
• Allowing enemies to spawn in the "spawn" room for floors > 1, i.e. after moving to the 2nd floor or beyond, you may encounter enemies right away
• Not spawning any rats in the first level at all (in some ways these guys are a bigger threat than the mutant miners, since they can follow you through tunnels)

Overall I'm feeling pretty good about this. There's a nice gentle curve, and no "cheap" early deaths. The current iteration (4 levels x 4 floors) feels a bit too long, but I think it will really start to take shape as I add more content and variety.

I also took stock of what "luck" affects (currently you gather luck by collecting gems), as I had somewhat forgotten, so I'm going to put it here for posterity:

• Mutant miner and rat spawn probability
• Secret room probability
• Item spawn chance
• Upgrade room spawn chance

In other news, I'm debating if (or for how long) I bother with early access on Itch before moving to Steam. My experience with Super Rocket Ride, and chatting with a few folks, had led me to believe that it's not a particularly viable platform for commercial projects. Case in point: there is no elegant way to handle key distribution, and I've heard (and also observed) that some streamers and review sites won't even bother with Itch keys.

So if anything, I might just throw it up there as an easy way to grant my patrons access (since that integration does work well) and then jump right to Steam. (side note: I may also fast-track getting SRR on Steam, for the same reasons)

[Thaumaturge]

Yup, I do that from time to time, especially for like "hero GIFs" (with multiple shots), but it would also be nice to be able to capture something quickly with LICEcap and trim it without having to re-encode.

Ahh, fair enough! ^_^

Not familiar with the term "whole-cloth" art". Do you mean like drawn or painted?

I mean created from nothing, not using actual people. Sorry, I do fear that my wording was poor there; I'm a little tired. ^^;

[Cosmocat Games]

Here's a recap of the progress on the multi-floor generation as of late, including the new and improved multi-floor connection module, and the difficulty adjustments:



Bonus: beacon launcher trajectory visualization is now integrated!

[Thaumaturge]

The new interval for the footsteps does sound pretty good indeed, to me! (And indeed, I quite like the footstep sounds, as well as I'm hearing them in my current headphones, I feel.) ^_^

When the rope is on top of those barrels, how would a new player discover it?

Regarding the photos appearing on the ghost hitting the player, if it was me that you mentioned as feeling that they would be out of place, then that wasn't quite what I was going for, I think. (Although I do admit that I was somewhat vague, I fear. ^^; ) Instead, I just feel like it's... disrespectful, perhaps? To use such photographs in that way. Perhaps I'm being oversensitive or overcautious, but... I find myself a little bit uncomfortable at the idea of using such pictures of people as "scary images".

I've just realised on watching this: Are the ghost basically Boos, from Mario? ^_^

Overall, this is looking pretty cool, I think! ^_^

[Cosmocat Games]

The new interval for the footsteps does sound pretty good indeed, to me! (And indeed, I quite like the footstep sounds, as well as I'm hearing them in my current headphones, I feel.) ^_^

Thanks! At some point I need to do a proper multi-surface foot step implementation, especially since I've added some additional surfaces (mainly, wood), but yeah I'm pretty happy with what's there now as the base dirt/gravel SFX.

When the rope is on top of those barrels, how would a new player discover it?

There's a subtle outline that appears when you're near interactable objects, and a fairly clear pattern for each floor (find exit, find rope), but if I get any feedback that people get stuck then I'll do something to make it a little more obvious for sure.

Regarding the photos appearing on the ghost hitting the player, if it was me that you mentioned as feeling that they would be out of place, then that wasn't quite what I was going for, I think. (Although I do admit that I was somewhat vague, I fear. ^^; ) Instead, I just feel like it's... disrespectful, perhaps? To use such photographs in that way. Perhaps I'm being oversensitive or overcautious, but... I find myself a little bit uncomfortable at the idea of using such pictures of people as "scary images".

That's a fair point. I was seeing them less as "scary" and perhaps more as "haunting"; like "echoes from the past" or something like that. Ideally they would be original content, agreed, but failing that I'm still thinking there's a way they could be stylized that would obscure them pretty well.

I've just realised on watching this: Are the ghost basically Boos, from Mario? ^_^

Maybe...

[Thaumaturge]

There's a subtle outline that appears when you're near interactable objects, and a fairly clear pattern for each floor (find exit, find rope), but if I get any feedback that people get stuck then I'll do something to make it a little more obvious for sure.

Fair enough! Indeed, player-feedback is likely the way to find out whether or not it's really a problem.

I will say that with my own playtests, it can be surprising what players sometimes don't see, especially when it comes to verticality.

That's a fair point. I was seeing them less as "scary" and perhaps more as "haunting"; like "echoes from the past" or something like that. Ideally they would be original content, agreed, but failing that I'm still thinking there's a way they could be stylized that would obscure them pretty well.

That makes sense; and indeed, obscuring their faces could perhaps add to that "haunting" feeling--perhaps even suggest not just echoes, but faded, half-forgotten ones.

[Cosmocat Games]

That makes sense; and indeed, obscuring their faces could perhaps add to that "haunting" feeling--perhaps even suggest not just echoes, but faded, half-forgotten ones.

That gives me the idea of maybe crossing out faces, like with pen scratches or something like that. Not only would that anonymize the photos but also maybe give them a slightly darker context. More respectful? Less respectful?

[Cosmocat Games]

Proof of concept for "miner flashbacks" when struck by ghosts:



I'm thinking the "obscuring" may not even be necessary given how brief and pixelated they are. Thoughts?

While I was at it I expanded on the "damage type" system that I had started a while ago. It now supports a "terror" type, which is used in the above. Eventually I'll expand it to include a "toxic" damage type as well which, at the very least, will change the visual feedback, and may eventually form the basis of a toxic status effect.