A couple days ago, I went to play a game of Drawception for the first time in like a year-and-a-half or something, and realized that the site is in a serious state of decline due to lack of technical maintenance—there’s still an enthusiastic community, but the site has been unmaintained for years and has been continuously developing problems, to the extent that at this point no one can register new accounts, and if anyone changes their email they won’t be able to log in anymore, so the site is losing members more and more over time just due to bugs. It also takes, like, a month for a game to get finished at this point, partially because of the dwindling number of accounts and partially because the site doesn’t do the best job of matching the number of active games to the number of active accounts. Anyway, I was feeling sad about all this too, and felt for the community there, so even though I have a lot of other stuff to worry about I started working on a replacement site for it, which people on the forums there seem thrilled about so far. I’ve gotten to the point of making a drawing tool, which I was worried would be an especially difficult part, but actually I’m already quite pleased with it.
If I can kind of squee for a moment, I’m really proud of myself for managing to throw this together in two days off my meds, especially with all the other crazy shit going on in our lives right now. The code is far messier and lazier than it would be otherwise—it’s much harder like this to keep myself from taking the easy way out all the time and racking up technical debt constantly, which I’m often quite tempted to do even on my meds—but hey, even though I’ll have a lot of cleaning up to do later, at least it works!!
Well, it’s honestly far from the most important thing for me to be focused on right now, but I guess I have to take what I can get—I got a basic web application going around that interface, so now you can:
make new pictures using that tool,
save them to the database,
view all the pictures in the database,
view a single picture,
download the SVG file for a picture to your hard disk, and
delete a picture from the database.
demos
drawing and saving a new picture
downloading and deleting a picture
basic features of the art tool so far, just using a mouse
(this is with like, no styling whatsoever on the webpages, it’s just using the brower’s default stylesheet—there’ll be more styling later oc )
It now records how long a picture took to draw (tracked by the server), and also you can now change the background color.
The time tracking will ultimately be important for the game, although I think it’s kind of a nice feature. I tried to take under ten minutes here since that’s how Drawception typically works and I wanted to see how it went. (Lily asked if I do would do a picture of masks, also, so that’s what I was going for subject-wise.)
Also, just in case anyone is worried about us, the recent Crypt Worlds -GOLD- release/fundraiser has gone pretty well, we have a list of apartments to visit now, and I spent a few hours today helping Lily with her writing project she wants to do as an update for the release, making an outline from an interview I did with her and transcribed, in the form of an HTML template she can fill in—she’s been feeling kind of daunted by it so I’m trying to do what I can to make it easier on her. She said a lot of fun stuff in the interview I think will be cool to read. There’s some things I really need to do outside the house like pick up my migraine prophylaxis stuff from the pharmacy and get my driver’s license renewed so I can drive a moving truck, as well as go and look at apartments, and although I feel like I’m really kind of falling over with that stuff I can’t really put off going to the pharmacy (if you stop taking that medication abruptly it can cause heart attacks, it comes with a black box warning) so maybe that will help get me into gear? Probably wishful thinking but maybe I’ll be lucky. I wish I could get myself to think about “looking at apartments” as the big project I need to be mainly focused on right now, as opposed to something I’m doing in little bits during marginal moments, but at least we still have a decent amount of time, and hopefully, with Lily’s help (or if my meds come in!!! any day now!!!) I’ll get it together enough that together we can find some okay housing. And, if nothing else, this web app might generate some income for us in the future since the Drawception people seem really excited about it, so at least it’s not a totally pointless thing for me to obsess over probably.
Now there are prompts. I also added a saturation control to the drawing interface, because I realized it didn’t have one, and that came in handy because the prompt I took on involved drawing a wolf.
(I really like the zany atmosphere this video ended up having )
The prompt generation is still in a very early form I would say, and could use a more sophisticated representation of grammar and that sort of thing, but it’s far enough along to generate some prompts I quite like.
On the whole the application has gotten relatively far along. Once you can respond to drawings with new prompts, it can have a hard timer you can enable for drawing on the client side that auto-submits the drawing past a certain point, there is some kind of abstraction for games built around that (barely needs to be anything extra at that point), and there are user accounts, the basic structure of the game as a whole will be there.
Things have been really busy over here and I’m not doing the best job of managing my time, but I did manage to add a basic representation of user accounts.
I feel like it’s getting into the home stretch now. There’s still a fair amount of work to do but it’s really close to actually being able to support the game.
It’s neat to be using Rails again after so long. I wish there was a game engine that felt more like it. It’s been thought-provoking in that regard.
still playing around with Godot, got the trenchbroom connection working so have been having fun trying to doodle quick levels to bounce around. i’m using some assets from my big heap of scavenged cd-rom era texture packs but idk if i’ll continue in this mode… it’s fun and new to me but also feels a bit too selfconsciously retro, it turns out.
@vodselbt i’ve only been figuring out trenchbroom a little over the last few days, it’s interesting but quite a change from any way i used to set up levels before!! esp each object being a discrete and fairly simple polygon. i never even played quake but if there’s any resources or levels you think are good for like getting a sense of what the editor can do, or even your own map, i’d love to see…
indulgent 3d platformer notes below:
having trouble deciding what’s enough in this context. i like really floaty jumps and even god help me a double jump - my favourite 3d platformer is jumping flash, where part of the fun isn’t so much the very technical plaformer thing of knowing exactly what each tool does… it’s more like the jumps you have are strong enough that they always tempt you to overreach, and then if you do fall down it’s always possible to come up with a different way to where you’re going rather than just having to redo the same approach.
i do also like games with more limited, castlevania / worms style jumps, and actually tested these out in game, but it’s kind of a rough fit with 3d as you’d expect hah.
one thing i do want to try putting in is a very basic “grab ledge” thing, and maybe a little automatic hop to clear small ledges so you don’t have to always do the big jump. the first tomb raider is also on my mind… i dont really know what to do beyond that, but it’s all good practice i guess even if i dont end up going much further with this.
For you at this point I bet dumptruck ds’ tutorials for Quake mapping will be pretty helpful with getting a handle on Trenchbroom as a whole. And honestly maybe getting a Quake environment set up just to be able to have a playable thing to hop around your levels in could be handy. But maybe you’ve got that covered in your godot project already?
Either way, these will intro you to bsp based level editors well enough I think
They’re a bit old and Trenchbroom has changed. So if you ever have questions, feel free to message me and I may have an answer.
Edit: Andrew Yoder also has a brief (maybe too brief) intro to Trenchbroom via an intro to Quake Mapping, which could be useful and take up less time than watching videos. I hate video tutorials personally. But the dumptruck vids are definitely more robust than this is Making Your First Map in Quake (Part 1) | Andrew Yoder
There is also the Qodot discord, which I’m not in, but could be a good resource or have resources in it. Qodot
I am getting my next Quake map ready for publication tonight. It’s totally done. I think it’s been seven months of work? I have been through so many compilations to make the stupidest smallest adjustments today. I’m exhausted!!!
Kitbashing various Godot Library thingies together to make a videogame. There’s a pretty decent amount of ocean shaders but this one is good enough, and there’s a good enough flightsim model that I can use as a starting point, complete with an already made UI. so…
I am going to have to figure out how to get the flight model to work in the way I want to but this is a much better starting point than yesterday, when I had basically the worlds worst flightsim I copied from a place called “Kids can code”
I am gonna submit something to an upcoming Quake mapping jam themed around brutalist architecture. This should press against all the friction spots that are part of my usual process: a desire to fiddle, to deal in big scope, to be blaze about presentation. I’m excited. I already have an idea for a theme that I’ll try to design some levels around…
You see, a lot of the maps from the previous brutalist jams have this same monolothic epic ambiance, and playing through the 35+ maps in all the sets my favorites ended up being the ones that broke from the expected in some way. I want to make a map that is dense and small, just as an exercise against my usual habits. But I was thinking… how do you do epic monoliths and brutalist architecture at a small scale, and how do I do it also without being so boring?
My idea is to make a brutalist museum-like space, or a library, that has been shot through with holes like someone has taken a sledghammer and created openings in unconventional spaces. I guess maybe something has invaded the space (scifi) or advanced construction is taking place (realistic) or time has eroded the archive (still scifi but a little more normal). I think it would be cool if players proceed through this library but never use an actual door to enter or exit any of the rooms.
playing with some old painted textures i had, i like the effect more than the cdrom ones, for environments at least. i also now have two different jump buttons - one a normal but generous floaty jump / doublejump. the other button is contextually either a high but finicky backflip or a low but far longjump - they’re based off the two types of jump in Worms games but i also was thinking abt the mario 64 longjump and the banjo kazooie backflip. you can also stitch on a regular jump at the end of them for control, which hopefully makes them a little less fussy. they’re differentiated - a backflip is higher than two double jumps, a longjump is longer etc - but i’m happier to keep them all as loosely overlapping things to switch between for now than treating them as keys to specific problems. i did tone down the regular jumpheight just a little after this vid though… maybe one should not bonk their head on the ceiling so easily
still thinking about whether i want a second thing, aside from just jumping around. on one hand right now it’s hard to think of anything to do in these levels rather than get from A to B in two seconds or, god help us, collecting things. look i know the guy from jumping flash collects things, i’m not better than robbit, but it feels horrible to commit to heaping bricabrac around as like a structural necessity rather than patchy and occasional level cement… on the other hand most of the immediate possibilities feel sorta like design sinks i don’t wanna get bogged into just for lack of anything better to do… idk i don’t want to use my 10 Free Hours of coding brain per year up without knowing what for just yet. so every so often i just make a new little zone and hop around it, dreaming…
it is fun to try things out in the level editor and see what works though - so far it definitely feels like the most abstract and asymmetrical test spaces work a lot better for me than more worked ones. i don’t think i’ll ever be a level design guy and most of the level design stuff i tried reading for this was profoundly depressing (“If the designer calls a map a warehouse, then there better be some crates lying around, because players will be looking for them.”)but the process of quickly spinning up areas in this odd tool is interesting. i do like as well how godot’s skybox equivalent seems to automatically tint the game the colour of the sky…
it’s trenchbroom! so maybe not odd but new to me (no real experience with dedicated 3d level tools, probuilder in unity was the first time i didn’t just make everything in some modelling software and then connect the bits in-engine)
i guess it feels odd to me because it has a mix of like much more obviously limited ability than a standard 3d modeller in some respects - every polygon has to be convex, doesnt seem to be built in functionality for things like rotating or scaling individual faces, every extrusion is a new object - but also has some very powerful tools as a result, like cutting or the CSG subtract button etc
ooh okay yeah. using qodot for the importer i assume?
yeah trenchbroom owns. it’s a totally different modelling paradigm from most stuff but i find that it’s baroqueness cultivates simple solutions to problems which could otherwise be a whole afternoon of learning arcane blender secrets.
yeah! or at least a qodot rewrite called “func_godot” which i ended up getting to work faster. it works well although sometimes it seems like not all the colliders are trustworthy, which is probably just due to my own laziness in sticking things together without making sure they don’t intersect. but once i got past the initial learning curve it’s been good, it’s funny how trying to get trenchbroom to quickly do the kinds of room you’d find straightforward in wings3d or probuilder feels very laborious but once you accept what it’s “for” it’s amazingly quick at things which would be a pain in other programs. the texture mapping is basically all i’ve wanted from a 3d engine, hah.
i worried that doing everything in a seperate level editor would make it a pain to playtest but it’s been surprisingly fine so far. i’ll probably want to start chunking things up a little more for real tests though since i always like to be able to try things out in engine.
)
