Sburb Development Project Thread

[C.J.C.]

Good luck. If I had the chance, I'd try to help, but I am not good at all with game designing. XD good luck and I really hope this doesn't die.

[Ziddia]

On the OP, you said 'C#, unless anyone has any other ideas for the code'.

Considering that, if you use C#, this won't be available to Mac or Linux, I would suggest looking into Java, using the LWJGL and Slick-2D libraries. It's basically the same syntax, with a few minor differences, an it would allow for cross-compatibility.

[tom cruise]

Java is a pretty bad environment for scripting though. Then again this isn't exactly looking like something that'd be hugely resource demanding, so it'd probably be fine.

[greyscaleQuestioner]

I've been wanting to be a part of making a Sburb game for a while now. I can help with spriting and some creative ideas. It would be really cool.

[humbleElitist]

"Variations"? Procedurally generated variations or handmade variations? It's probably worth mentioning that attempting to procedurally generate music in any way would be an incredibly, hilariously bad idea. Like, Greek Tragedy levels of hubris.

wolfram tones.
quite a bit of it is pretty good.

[humbleElitist]

"Variations"? Procedurally generated variations or handmade variations? It's probably worth mentioning that attempting to procedurally generate music in any way would be an incredibly, hilariously bad idea. Like, Greek Tragedy levels of hubris.

wolfram tones.
quite a bit of it is pretty good/ok (to the untrained ear, such as myself).

[Frost]

wolfram tones.
quite a bit of it is pretty good/ok (to the untrained ear, such as myself).

Ehh. Had a look at it, and the results seem just bad enough to be unsuitable as ambient music. You keep unconsciously expecting the "correct" thing to happen (like a simple downwards sequence of notes), and when the computer gets it wrong (by, say, skipping a note), it's jarring as fuck.

In any case, it's sort of a moot point. The underlying system isn't open-source or embeddable, and it was created by a large well-funded corporation that's specifically in the business of programming smart AIs. Not the sort of thing a small team of amateurs could imitate.

[humbleElitist]

Ehh. Had a look at it, and the results seem just bad enough to be unsuitable as ambient music. You keep unconsciously expecting the "correct" thing to happen (like a simple downwards sequence of notes), and when the computer gets it wrong (by, say, skipping a note), it's jarring as fuck.

I should probably point out that the notes are determined by cellular automata.
I believe one could create a program capable of detecting at least most of those wrong notes.

In any case, it's sort of a moot point. The underlying system isn't open-source or embeddable, and it was created by a large well-funded corporation that's specifically in the business of programming smart AIs. Not the sort of thing a small team of amateurs could imitate.

The point was to show that it was possible.
Not necessarily feasible for a small group, but possible.

If I wanted to convince you that it would be possible for a small group, I would bring up pokemon crystal version's slowpoke well. (I think that was it)
because some of the cave music was semi procedurally generated, as storing such a long music file would have taken too much space.
thats kinda different but whatever.

Do you believe that 24 very short bittune things in 2 groups of 12 can be created such that every thing in one group can come both before and after anything in the other group and sound okish?
or not?



also in general I think that instead of there just being a bunch of sburb projects there should be a bunch of not completely related projects for each aspect.

one for alchemization

a few for different types of inputgameplay stuff. (platformer, 3d, ascii, etc.)

one for determining server connections (ehh)

two or 3 for world gen (2d/3d)

one for auto-generation of echeladder level names(there are already pun generators, just add players class and aspect to the equation)

and then there would be different groups for each, and then when two got a demo working, they could try to figure out a function to convert between the different formats that they will inavoidably develop or use for different things.

ok, actually i have no idea if that is a good idea or not
I just want to work in a group on the alchemization part, and I think a group dedicated to the alchemization part is less likely to fall apart.

one possible advantage of splitting it into a bunch of sections would be that if someone wanted to join but did not really have anything they wanted to(and thus would) do they would not know which one to join and then they wouldn't.

agree/disagree

[gumptiousCreator]

That's not a bad idea. The best thing about it is it gets people started with what they want to do straight away, if there's a specific aspect they're interested in. I doubt that a good game (or if I'm being pessimistic, any game at all) will be the final outcome, but one positive result is experience within the Homestuck community of making things from Sburb, which is a good thing in itself.

So if you want to work on alchemy, start up a project for it. People might be interested regardless of whether it's connected to a Sburb project (I, for example, am interested in alchemy for use in one of my own projects).

[ObeseCatLord]

This has been bumped I see.

So has a programming language/engine/library been chosen? Or is this all just brainstorming right now?

[Frost]

Do you believe that 24 very short bittune things in 2 groups of 12 can be created such that every thing in one group can come both before and after anything in the other group and sound okish?
or not?

I know shit-all about music, but I'd guess no. Firstly because the resulting tracks wouldn't have much repetition or consistency; secondly because the instruments in each group would need to be pretty much identical to avoid sharp transitions (even chiptunes have "instruments" to some extent); and thirdly because all of the resulting tracks would have the same tempo and key, and they'd therefore "feel" the same. I'm sure a musician could give you some better reasons.

I just want to work in a group on the alchemization part, and I think a group dedicated to the alchemization part is less likely to fall apart.

Posted this way back in the thread, you might find it interesting. You might also find it laughably flawed, because I came up with it over a few hours and then never did any more work on it.

show spoiler

hide spoiler

Here's the big problem with most alchemy systems: they don't actually generate many "new" items. Almost everything interesting has to be hand-made, because the system itself is only capable of performing very simple acts of creation, like "change this shield's material from bronze to steel". This can't possibly lead to true procedural generation of billions of items; either vast swathes of these items end up being almost identical, or some cheap cop-out is introduced to heavily restrict the player's options (like duplicating items, or having many combinations result in junk items).

It follows that any "true" alchemy system, similar to the alchemy seen in the comic, will have to be capable of generating items from scratch. Impossible, right? Computers aren't capable of creativity; it would be impossible for them to convert "hammer && pogo ride" into "a hammer with a toy-like handle, and springs behind each hammerhead, that conveys specific spring-based combat abilities".

Here's the solution. The programmer creates many "patterns": very small programs or scripts which can procedurally generate a specific type of item. You might have one pattern for hammerkind weapons, one for helmets, one for hats, one for fruit gushers, one for wands, and so on. Speaking as a pretty experienced programmer, I'm convinced that creating a couple of hundred patterns would be a plausible task for a programmer or small team of programmers, although it certainly wouldn't be easy.

Every item in the game would be associated with one pattern. The CHOPKIND pattern would be associated with PORK CHOPS (a handmade item that exists in the vanilla game), and might be used to generate IRRADIATED PORK CHOPS, BEEF CHOPS, GUSHER CHOPS, and so on.

Every item also possesses a series of "traits". These traits are the parameters for the item-generation algorithm of each pattern. That is to say: [pattern + (a specific set of traits) = (a specific item)]. For technical reasons, traits would have to be very diverse and fairly abstract for each handmade item. Below are a couple of example trait lists.

PORK CHOP:

show spoiler

hide spoiler

materials: pork, pork crackling, animal fat
colors: pink, brown, white
consumption effect: restore health, enhance vim
armor effect: enhance vim
aroma: "smells like sausages"
animal: pig
icon: pork chop, pig
grists: meat
...

CALEDFWLCH:

show spoiler

hide spoiler

material: sturdy iron
color: white
weapon effect: deadly sharp, enhance damage
armor effect: enhance speed
icon: gear, hephaestus
name overrides: ____fwlch, caled____
grists: iron
...

Some of you may have guessed the punchline already. In order to alchemise two items, you select the pattern from one of the base items, and then randomly shuffle both base items' traits. The "child" item would possess a pattern from one parent and traits from both parents, with a strong bias towards giving it traits from the non-pattern parent. If the pattern was programmed in an intelligent way, it should be possible to generate a convincing item from an essentially random set of traits.

Let's look at a simple example, combining CALEDFWLCH && PORK CHOP. Let's say the result gets the pattern from Caledfwlch: SWORDKIND. This pattern generates a weapon with the following sources of variety:

• The item is named "____ SWORD", where ____ is one of the item's material traits, unless the item possesses a name override trait. Very powerful instances may be called "SWORD OF _____".
• The blade is made from a material selected from the item's material traits.
• The blade is longer if the sword is more powerful.
• The hilt may be decorated with an image selected from the item's icon traits, and is the color of one of the item's color traits.
• If the sword is powerful, it gains a special combat effect selected from the item's weapon effect traits.

So the result, generated using only cold unfeeling logic, might be a sword called PORKFWLCH, costing eight MEAT GRIST and four IRON GRIST, with a short blade made from PORK CRACKLING (with all attendant combat penalties), and an IMAGE OF A PORK CHOP on its brown and pink hilt.

It probably looks like I've skewed the numbers to create a deliberately interesting item, but I honestly haven't. Almost every combination I punch into this system ends up generating something at least passably interesting. It wouldn't of course end up being as creative as the comic (that swordkind pattern could never generate caledscratch, for example), but I can see this being just functional enough to keep players interested. The main drawbacks would be the programming manpower and experience required, some degree of repetitiveness in the generated items, and the tedium of specifying dozens of traits for each handmade item.

tl;dr: caledpork

[humbleElitist]

That sounds like a pretty good system.
One small part that I don't particularly agree with is the randomization aspect.
I think that making it deterministic would not be too difficult, and would more closely match the comic.
making it deterministic also allows for actually using the codes.

short thought section were i had an idea and determined it wold not work:

show spoiler

hide spoiler

I guess one would have to come up with a bunch of types of grist first, and store whether it needs a certain type of grist (I really want to type girst when I type grist...) is a set of bits.
one problem with that is that it would need upwards of 2*types of grist bits just to store which types it would need.
Not even how many, just what types.
and also that leaves the problem of modifying the code to make items free.
so a better solution is required there then one bit for each type of grist.
so it probably should be included in the type of item info.
which is probably what you were thinking.

so grist would depend on item type.
or item type would depend on grist. that's an idea. possibly not a very good one but still. (lets discard it temporarily)

so I think it would make sense to split things between all possible weapons and armor/clothing and everything else.
so that would be the first bit.
provided that the first bit indicated weapons/clothing/armor
the second bit would specify whether it was weapons or clothing/armor
otherwise it would specify... whether it is, some type of object... or some other type of object... (little help with that?)
alternatively pairs of bits could be used for both of those so as to allow for more control over the results of && and ||
(although that control would decay after repeated combinations(which I don't think is that big a deal))
possibly a bit to specify whether the object has any powers would be good for the third bit. maybe.

I suppose then one would have to determine what qualities/types of things (like meat(that says meat, not heat), or magic) are mutually exclusive, and then use that to choose what bits would correspond...

[Frost]

making it deterministic also allows for actually using the codes.

You could pretend to use the codes (and have them match up properly when you perform an &&-combination, for example), but truly generating an item from just its code would be mathematically impossible. There's simply no way you can store enough information in 48 bits, at least if you're using the system that I described above as a starting point.

I guess it just might be possible to have the captchalogue codes work if you were to design an alchemy system from scratch with them in mind, but it'd be difficult as all hell and the resulting system would be crippled. You'd be sacrificing a lot for very little gain.

[thottobeextinct]

I read the first few pages and then the last page and got scared by the complete and utter contrast of mood and many of the reponses.(just wanted to point that out)


plus: it would probably be too difficult and overcomplex to add, but what about being able to choose either troll or human.

That's probably one of the stupidest things suggested, and i apologise because i know less than a rock about game creation or programming beyond basic html, but i just wanted to put that idea out there. Now please take that idea and trash it, mock it, disregard it, like it, or ignore it as you please.

[darkDetective]

Alright, so I just wanted to give everyone a heads up. We had most of the information, including the 2 or 3 pages of a design doc I had written, contained in the group forums. I did not have control over the group forums, someone else had set it up and thus had ownership. When I decided to put the project on hiatus to deal with real life things like money and a job, the owner deleted said forum.

Needless to say I haven't been keen on working much on it at all since all of my, and the rest of the teams, work was basically deleted over night. All I have right now are some old notes and vague memories about various in game systems. Oh and maybe some old contacts.

But, Anime Central is in little over a week now. I was planning on pitching a more localized project at ACEN before our work got deleted. I will be handing out flyers there, just in case. I will probably try to rewrite the design doc, too. Hopefully my poor memory won't be an issue.

I miss this project. :/

it would probably be too difficult and overcomplex to add, but what about being able to choose either troll or human.

That was originally going to be an expansion pack if we felt like it was worth it after the original release.

[Milksweet]

Hey! Don't sweat! Just remain calm, and

[aeonsleuth]

I have a (possibly already mentioned?) idea for dream bubbles.
It should function somewhat like Spore, every player who perma-dies/dreams in a dream bubble has all their memories, lands and stuff submitted into a behind-the-scenes 'dream cloud'. Locations and characters from this are randomly placed in other players dream bubbles, even those who aren't in the session.
This means that dream bubbles sort of function like the multiplayer in Journey, with a random land and player being connected. You can still solve your land's puzzles and sidequests in the dream bubble, meaning you still have an incentive to play, but now other players' lands and memories appear from time to time. If another player isn't playing but you stumble upon their land, they will just be controlled by an AI. You might be encouraged to choose/design a quirk in Pesterchum/Trollian early in the game. If you talked to a NPC version of a PC, it would be a dumb chatbot that used that quirk. (Maybe).

[Captain Combusken]

I really want this project to happen! I desperately feel that we need an actual SBURB game. So this project is awesome in my eyes. That said, I would like to help.

I'm mainly good in three areas.
-Ideas. I can churn these out with ease. I'm quite creative.
-Organization. I have a compulsion to organize and categorise things. Collating various aspects of this into lists and folders will be important. I'm happy and willing to do that.
-Spriting. Not any good at large scale art/3D, but I am fairly decent at small-scale pixel art.

If you need me for anything, feel free to ask.

Has anyone thought in detail about characterisation? I think for a game like SBURB, we need detailed character features:

Appearance:
-Hair (Styles and Colours)
-Eyes (Colours, shapes, expressions [Eyebrows are included here]
-Mouths (shapes, expressions, colour [lipstick])
-Glasses (various shapes, styles and colours).
-Clothes (we will need a huge amont of styles and colours for top and bottom half.)
-Clothing Acessories (Scarves, Capes etc)
Basically, for the appearance, we need to create. It won't take as long as other aspects, granted, but a group of people is needed just to make a large catalogue of things for the sprites of our characters. I'm willing to run that.

Personality:
A Similar process. I think a Pokemon Mystery Dungeon style personality test before the game starts is a good idea. The personality would also include likes and interests. The people doing personality would need to work with the alchemy team to a degree, and to the plot outcomes team (how would your character's story work based on your personality?)

SBURB details: These would probably originate from personality, but a list of what personality gives what title is important.
I also feel we need more classes and aspects in pur version of SBURB, just to individualise ourselves more.

House and Relatives:
We need, similar to Appearance, a list of different houses, different rooms, different furniture. Also a different Guardian each is important.

Pesterchum: Chumhandle and Quirk, mainly. We need to find a way to get the Pesterchum we already have in the game as well.

I have probably forgotten a lot, but I suppose it's a start, right?

[Blastoking]

This really looks like it's going to be epic, I've already got some friends who are going to try this out with me, can't wait until this is released.

[UranianUmbra2]

Will we be able to choose an associated item and possibly an exile? Just an idea. If you add the ability to choose an associated item, here is how it should work (Based off the homestuck kids' associated items):

Seeing how John's associated item is oil, which is a liquid, perhaps being able to choose an associated item that, like oil, is a liquid? Choices could include mercury, bromine, water, hydrogen peroxide, etc.

Jade's associated item is Uranium, so possibly being able to choose associated items that are elements on the periodic table? However, the only ones that could not work are the artifical elements that don't last more than a few seconds. Also, perhaps certain compounds, such as Capsaicin (The chemical responsible for the heat in peppers)?

Rose's associated item is chalk, so possibly being able to choose associated items that are common objects?

And Dave's is amber, so possibly associated items that are fossils?

For the exiles, perhaps being able to make them in a creator similiar to your kid?

And finally, will we be able to use a troll as a character, and be able to have up to possibly 24 - 48 players in a session?

Edit: Adding trolls to the game would be worth it, seeing how alot of the fanbase is only reading it for the trolls, it'd probably boost popularity of your project.

Edit: And what about sprites and first guardians?

Edit: You should make it so one cannot use grist in the title of a planet.

[Blastoking]

Is this project still alive anyways? Cause the last time the Op was updated was quite some time ago.

[darkDetective]

Is this project still alive anyways? Cause the last time the Op was updated was quite some time ago.

The topic starter hasn't had anything to do with the project since about the third page.

[Blastoking]

The topic starter hasn't had anything to do with the project since about the third page.

Ah ok, I just saw the first post really, I hadn't searched through the entire thread yet.

[clr10]

I would love to be a part of the development of this game. i am a artist who has basic concepts of shading, lighting and third dimensional perspective. if you need artist i will be happy to help.

[theMiles]

i hope this game is still being worked on

anyways as i was reading this threads you have to put into prospective of how complicated and diverse sburb is, i mean andrew hussie created two UNIVERSES! that takes a true master to do...as for the game the whole idea of time travel seems amazing but it whould take a team of countless programmers,artists,a hell of a lotta other stuff, and a whole team effort to make. it whould be really easy to not understand time travel as a player because where not all stephen hawkings. we whould all most likely create horrible in game paradoxes and crash our games.......if you guys know how to pull that off well...i dont think you should be wasteing your time with these small games and go build a masterpice for some large company...XD best of luck though