Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Locked thread
anselm_eickhoff
Mar 2, 2014

aeplay.co
New Update, again!



The Road to Alpha, Week 38 - Curve Control

Adbot
ADBOT LOVES YOU

PoizenJam
Dec 2, 2006

Damn!!!
It's PoizenJam!!!
Oh my god... That road building tool is genius and blows every other modern city sim out of the water, hands down. Good job!

ExtraNoise
Apr 11, 2007

Wow Anselm, that looks great. Can't wait to see the final product with intersections rendering.

Also, get some sleep dude. :)

Shibawanko
Feb 13, 2013


Please make a parody of the Simcity 2013 trailer where you draw some roads shaped like an ejaculating penis instead of a guitar.

anselm_eickhoff
Mar 2, 2014

aeplay.co
New Update, finally!



The Road to Alpha, Week 40 - Hyper-Active

Baronjutter
Dec 31, 2007

"Tiny Trains"

Sounds like you are having the full Russian experience!

drunkill
Sep 25, 2007

me @ ur posting
Fallen Rib
That video is amazing, while only 40 seconds of gameplay it shows you've laid the groundwork for the game. Something as simple as lines on the road and slip/turning lanes look amazing.

I hope you have a left hand drive option in the works.

Sorry about your flat :russia:

anselm_eickhoff
Mar 2, 2014

aeplay.co
Merry Christmas and a Happy New Year!



Citybound: 2014 in Review

GlassEye-Boy
Jul 12, 2001
Great review, I'm super excited to see what you do in the coming year!

Crash74
May 11, 2009
Keep up the good work man, its been cool watching you make this project. When are you planning to put hills and water in? Will this require a rewrite for the stuff you are working on now or is everything designed to mesh together? Will the roads act the same when placed on the side of a hill?

Mecca-Benghazi
Mar 31, 2012


Caught up on the thread, things are looking great so far and I always love Baronjutter's posts.

I was thinking: I have a ton of family in Saigon/HCMC city and there motorcycles are the primary mode of transportation, not cars, due to a combination of cost and space being at a premium. To tie into parking chat earlier, you could have lower working class people who work in places with limited parking yet terrible transit use motorcycles instead. Not sure how hard that would be to model though. :ohdear:

anselm_eickhoff
Mar 2, 2014

aeplay.co
First developer diary:

Developer Diary #1: Where do you think you're going?

Baronjutter
Dec 31, 2007

"Tiny Trains"

Good you are making progress! I'm trying to put together a really stupid 2d city builder in Java and it's taken me over a month and someone telling me exactly what buttons to push to just get a tile based map generating and that's nearly killed my brain.

Looking forward to more progress!

Shibawanko
Feb 13, 2013

Yeah I can't imagine actually coding a game like this, or any game really. It seems impossibly difficult although that's no doubt because I have a useless brain.

anselm_eickhoff
Mar 2, 2014

aeplay.co
New Dev Diary!

Developer Diary #2: Intersection soup

Baronjutter
Dec 31, 2007

"Tiny Trains"

Huge nightmare 6 lane intersection on a funny angle with no markings leading to a clusterfuck? You really are living in Russia now!

Nition
Feb 25, 2006

You really want to know?
The only fear I have is that you'll finally get everything working perfectly, with the traffic system flowing far more sanely than it ever did in SimCity, but then all the carefully-considered calculations will prove too intensive when the simulation is scaled up to city size.

So you'll start tweaking it, cutting features where you can, allowing a little imprecision in exchange for performance. It's still not enough. You're cutting vehicle types, simplifying trips, you realise you could massively optimise the system if vehicles would only turn left... and suddenly it's SimCity 5.

On the other hand, I really like the idea of an agent-based simulation and I really hope it works. I know a lot of people are pushing for a more tried-and-tested simulation model, but using real vehicles on the road instead of a statistical calculation just has so much more potential for interesting emergent traffic situations. That and just watching real traffic making real journeys is way more interesting to me than fake sim traffic generated just for visuals.

Baronjutter
Dec 31, 2007

"Tiny Trains"

Yeah this has always been my worry ever since you decided to model every single car real time as an agent, that there will be no more "budget" for anything other than traffic simulation and the actual city simulation will have to be bare-bones to compensate, or tiny tiny cities.

Jamfrost
Jul 20, 2013

I'm too busy thinkin' about my baby. Oh I ain't got time for nothin' else.
Slime TrainerS

Nition posted:

So you'll start tweaking it, cutting features where you can, allowing a little imprecision in exchange for performance. It's still not enough. You're cutting vehicle types, simplifying trips, you realise you could massively optimise the system if vehicles would only turn left... and suddenly it's SimCity 5.

I did not expect to read such a terrifying story in this thread.

anselm_eickhoff
Mar 2, 2014

aeplay.co

Nition posted:

I know a lot of people are pushing for a more tried-and-tested simulation model, but using real vehicles on the road instead of a statistical calculation just has so much more potential for interesting emergent traffic situations. That and just watching real traffic making real journeys is way more interesting to me than fake sim traffic generated just for visuals.

This is exactly my motivation. Let performance be my worry.
Under no circumstances will I dumb down the game to SC5 levels due to performance issues.
There are always much more intelligent compromises to be made. Maybe I don't even need to make any.

Here is today's dev diary:

Developer Diary #3: The Struggle

Microplastics
Jul 6, 2007

:discourse:
It's what's for dinner.
I'm all for an agent-driven modern city simulator and am fully aware it'll probably mean compromising on a lot of other things, but dammit I want to see it attempted and I'm not counting Simcity 5 among the attempts because that game's problems go far beyond the mere use of agents :colbert:

For me, I like to stop building or managing my city and instead just peruse it. I would love to be able to follow a citizen around as he commutes, shops, moves house and switches jobs, good to school, starts a family, falls ill, retires, etc etc. But I would be dissatisfied if the engine just made it up on the fly as soon as I looked up close, with everyone around him still abstracted (I think Simcity 4 did this with imported Sims). I want an engine where agents persist and really drive the patterns in the simulation.

Poil
Mar 17, 2007

Well if the game can utilize 64bit and multiple cores properly you'll get a ton more power to work with. No, I'm certainly not bitter about anything. :argh:

Supraluminal
Feb 17, 2012
My concern, as someone with slim qualifications for giving advice to a professional application developer so please feel free to ignore this, is that too much time is going into polishing the road and traffic system while the other systems are still extremely skeletal. It's the one of the motivations behind the agile development philosophy I guess - features can influence each other as they take shape, so it's best to do lots of iterations and keep pushing everything along to some extent instead of focusing on Feature A until it's "done." Because then when you dig into Feature B, you discover that it requires making some fundamental change in Feature A that invalidates a bunch of work you've already done.

That's just what I see as a web developer and casual observer of Citybound, though. Maybe from Anselm's perspective this is a non-issue.

nielsm
Jun 1, 2009



Nah, making the roads work is fundamental. Since the simulation is intended to be based around things being able to move between locations through the roads, if the roads don't work the foundation of everything else is missing.

Kylra
Dec 1, 2006

Not a cute boy, just a boring girl.
City sims tend to live and die by their traffic models. Not only is it an important issue in real life, pathfinding is often computationally intense. It will probably end up being where the vast majority of non-graphical processing goes if it is going to be agent-based, and will be by far the main factor of things like maximum city size.

Kylra fucked around with this message at 16:49 on Jan 12, 2015

Supraluminal
Feb 17, 2012

nielsm posted:

Nah, making the roads work is fundamental. Since the simulation is intended to be based around things being able to move between locations through the roads, if the roads don't work the foundation of everything else is missing.

Yeah, it's a balancing act. There needs to be something in place for other parts of the game to work, but arguably it only has to be good enough at each stage of development relative to the other features. That might provide more opportunity for the other features to inform the development and refinement of the road system. Of course, defining "good enough" may be difficult.

I'm not trying to second-guess Anselm too much here, it's his project after all. Just talking about what comes to mind looking in from the outside.

e: I guess my point is that, if Citybound is to be more than a pure traffic simulator, whether or not the roads "work" is ultimately defined by the other pieces of the game and how they all work together. Anselm could end up with a perfect self-contained traffic system that ends up being completely unsuitable for other features as they evolve if it's developed in a vacuum.

Supraluminal fucked around with this message at 16:51 on Jan 12, 2015

PoizenJam
Dec 2, 2006

Damn!!!
It's PoizenJam!!!
I agree with Supraliminal insofar as it relates to PR releases surrounding the game. For die-hard city sim fans like myself all of Anselm's updates and fine tuning of the traffic system are like sim-porn to me. I love it, the depth of detail, the level of modularity... As someone who's played these games to death I can REALLY appreciate the consequences a well implemented traffic system will have.

That being said, for the more casual observer, their enthusiasm for this game may run out of steam if they just keep seeing road updates over and over again. After so many months as its been, I can see fatigue setting in with this and the Reddit thread- there's not as much to talk or get excited about for the casual city sim fan. Now as far as I'm concerned, this has more to do with flawed expectations of your average observer, especially for a 2 man team to complete a game independently. The game was revealed extremely early in its development cycle too. In fact, to some observers it may look like progress has slowed or reversed, since the earliest teasers of the project showed incredibly large cityscapes, now we're mostly getting updates about roads since the redesign. If you look at Cities: Skylines by the time we were exposed to any media surrounding it there was a functional game underneath. It's not Anselm's fault, it's just the nature of indie development; one needs to generate funds and interest and without the initial capital from working with a corporation you'll be limited to revealing your project early and hoping for donations.

Nition
Feb 25, 2006

You really want to know?
Also, game dev just slows down as you go. Even if Anselm is a machine and keeps working at full-speed forever, it's always way faster to add feature 1 than to add feature 10. New stuff has to work together with more and more old stuff as you go, which often means changing the old stuff around a bit as well. And finishing a feature properly - ironing out all the little issues - can often take longer than actually getting the basics working did.

So er, my point is that roads might seem to be taking a while compared to the initial basic city work but the other features will probably take a while too. Of course obviously roading and traffic simulation is a really major part of the game too.

PoizenJam
Dec 2, 2006

Damn!!!
It's PoizenJam!!!
I predict roads and procedural building generation to take up the bulk of his time from the point of view of gameplay mechanics. The rest should be gravy- even sea/air travel are usually more abstracted and conceptually easier to perform, especially without dedicated lanes.

I'm wondering if traffic accidents/jams will occur in the final product. It would be nice to have scenarios/challenges to work out grounded in reality rather than SimCity's monsters/aliens/meteors. I never personally played too much of them, but the old SC and SC2K scenarios were a neat idea- Say you start with Detroit at the beginning of its decline and have to retool the city's economy to prevent it from becoming a deserted hellscape.

Baronjutter
Dec 31, 2007

"Tiny Trains"

I just hope the traffic/road system will be robust enough to correctly handle peds and bikes and things that aren't cars, rather than them being more of an afterthought. All this car stuff is good to have, but as someone wanting to build mostly car-free cities I hope the tools will be there to have big dense european style car-free cores serviced by transit and bikes.

It's really all wayyyy too early to judge. I though can't imagine having really large cities with this level of agent simulation. Simcity couldn't do it and now we're learning skylines is going to have a fairly small population cap as well as abstracting the gently caress out of larger building's populations because the agent system can't handle it. Maybe those companies are stupid or Citybound will be crazy optimized in ways no one has ever thought.

Baronjutter fucked around with this message at 23:13 on Jan 12, 2015

The Deadly Hume
May 26, 2004

Let's get a little crazy. Let's have some fun.
Or maybe pathfinding a vast number of vehicles is HARD.

Part of the issue is that in reality peds have a LOT more freedom to move, particularly across parks and plazas, to the point where you pretty much have to have a completely different approach to motor traffic, where you can rely on a system using discrete lanes, with a limited flexibility to allow for lane changes etc. (Not that cars strictly stick to lanes in real life)

Whereas bikes are sort of in between.

I don't think you're going to get the game you want here, either, Baronjutter.

NatasDog
Feb 9, 2009
I'm in the camp that enjoys the traffic sim as much as, if not more, than the city builder itself. Being able to handle lane assignments to a somewhat granular detail would be awesome; bike, ped, auto, etc. Even dedicated emergency/mass transit lanes would be nice to have. Picking the number of left/right/thru traffic lanes would be nice to have too; I can't tell you the number of times I've been pissed off at SC4 because I couldn't dedicate more turn lanes on an avenue that has a lot of left/right turning traffic at certain intersections.

Of course, all this is probably more than your average sim lover would feel like micromanaging and therefore won't make the cut, but a man can dream.

MikeJF
Dec 20, 2003




I'm kinda reminded of how SimTower was often described as an elevator sim with some extra bits attached.

carticket
Jun 28, 2005

white and gold.

MikeJF posted:

I'm kinda reminded of how SimTower was often described as an elevator sim with some extra bits attached.

You did have to put some thought into what to put where so you could find elevator expansion. I wasn't patient so I'd build lots of condos for quick payout and then have no recurring revenue.

Poil
Mar 17, 2007

Mr. Powers posted:

You did have to put some thought into what to put where so you could find elevator expansion. I wasn't patient so I'd build lots of condos for quick payout and then have no recurring revenue.
Condos were such a trap option. Especially when people started moving out after you had spent the money because your elevators weren't good enough.

Baronjutter
Dec 31, 2007

"Tiny Trains"

The Deadly Hume posted:

I don't think you're going to get the game you want here, either, Baronjutter.

It's still too early to say. He said he wants the game to be able to model his hometown, which has trams, wall to wall buildings, and great pedestrian zones. I still hold out hope that all this car focused stuff will just be for cars and peds and transit and everything that isn't a car will also get the attention they deserve (rather than being badly crammed into the same system cars use)

Baronjutter fucked around with this message at 18:04 on Jan 13, 2015

Shibawanko
Feb 13, 2013

Maybe pedestrians should be mostly on rails, like cars, when they follow a sidewalk or alley or something but can enter a kind of bumper car mode where their movements are sort of randomized if theyre on a plaza or shopping street?

crabrock
Aug 2, 2002

I

AM

MAGNIFICENT






sim tower was easy, but you had to set all your hotel prices down to the lowest price option. then they'd get hella full and you'd make a shitload of money. I worked out a great pattern, with a maid service on every floor so the whole hotel could get cleaned by the next day.

were condos the ones you couldn't destruct later?

miss my yellow VIP :(

Baronjutter
Dec 31, 2007

"Tiny Trains"

Shibawanko posted:

Maybe pedestrians should be mostly on rails, like cars, when they follow a sidewalk or alley or something but can enter a kind of bumper car mode where their movements are sort of randomized if theyre on a plaza or shopping street?

I think Cities in motion 1 handled this the best I've seen in any game. Cars drive on roads as you'd expect, trains and trams obviously follow tracks, but pedestrians were much more free-floating. They'd follow sidewalks and use crosswalks when available, but they'd absolutely take short-cuts if that was a significantly faster route. They'd cut across plazas, through archways in buildings, down ally ways, or even just walk through the woods if it was the fastest way, yet their routes generally always felt natural. They'd use a crosswalk if one was nearby, but if not they'd just cross where ever. They'd say on paths and sidewalks and pavement when available, but they'd cut across dirt and grass if they had to. They basically behaved exactly like actual pedestrians. I'm not sure if they had too much in terms of collision detection with each other, and there was no measure for sidewalk capacity (something actually important) but any time you'd zoom in everything would look right.

Bikes would work the same way, just with a much higher favoring of pavement and bike lanes (when available) but having no qualms about riding on the sidewalk or between buildings if that route offers significant time savings. They could also dismount and walk their bikes, transforming into a pedestrian, as needed.

Then of course model bike parking too, so we could construct massive bike parking structures!


Also anyone who liked sim tower should grab yoot tower, it's the game simtower was meant to be.

Adbot
ADBOT LOVES YOU

Yaoi Gagarin
Feb 20, 2014

Shibawanko posted:

Maybe pedestrians should be mostly on rails, like cars, when they follow a sidewalk or alley or something but can enter a kind of bumper car mode where their movements are sort of randomized if theyre on a plaza or shopping street?

Maybe pedestrians could have two modes, "get to destination" and "relax". In the former, they just try to walk to wherever they need to go, and in the latter, they meander through parks and plazas. Then allocate an hour or two to everyone's day schedule for leisure, where they are switched to "relax" mode. If the leisure time itself is assigned randomly, it will seem like people are constantly enjoying the public spaces in the city. You could tweak the random distribution so that certain times are more crowded.

  • Locked thread