croxis

It has been announced that there is to be only one military unit per tile. The actual definition of what a military unit is is still unknown. Some screenshots seem to indicate that units are mixed with different types, but this aspect is pure speculation at this point. The one military unit per tile means that the Stack of Doom strategy is out the door (finally)

You could also set up bittorrent to distribute the files too. Also a good way to promote legitimate use of torrent technology. On a different note: CityMania Terrain Prototype 3.1 was released today! Directions on installing and running can be found here: https://www.sc4devotion.com/forums/index.php?topic=10175.0

The fps is one of the first things i noticed. That is why I put the fps counter in after all. I can read. I know exactly what the issue is, the issue is in the collision traverser ray, which the engine default is to call once per frame. It takes 17-33 ms calculation each frame (hence the low fps). My class which manages that part of the code will need a bit of refactoring.

Optimization comes later. We need to get some sort of paged terrain implemented as well as optimize the water effect. Other things must come first!

Congrats on the upcoming release dariovillar As for us, we just (mostly) finished the core terrain engine:

We got our engine down for CityMania Programming Language: Python/C++ Rendering Engine: Full 3D, freeform. Panda3d Zoning: Yes Taxes: In discussion Civics and Satisfaction: In discussion, probably SC4 esk Utilities: Power and other big utilities to be bought at region level Customizable: Intention to be fully customizable by end user, both with code and graphics Appearance: No art direction at this time. To celebrate, here is the second prototype! https://www.sc4devotion.com/forums/index.php?topic=10078.0

i put together a very simple terrain prototype than can be run through a web browser. I have details here: https://www.sc4devotion.com/forums/index.php?topic=9873.0

Big open source projects have a initial quiet period while things start to click into place. I've already started working with the panda engine (due to my previous experience with it and I can code faster in python than c++). As soon as one of our members finishes the engine stress tests we will be able to make a decision. This is also a learning process for us, I am having to learn programming concepts that are new to me, and it takes time to understand what I need to do. But yes, we do need a project head

City Mania Programming Language: Python/C++ Rendering Engine: Full 3D, freeform. Either Panda3d or Irrlicht Zoning: Yes Taxes: In discussion Civics and Satisfaction: In discussion, probably SC4 esk Utilities: Power and other big utilities to be bought at region level Customizable: Intention to be fully customizable by end user, both with code and graphics Appearance: No art direction at this time.

As a fellow city builder developer may I offer some suggestions. * You will have a difficult time finding fellow developers for two big reasons. Darkbasic is not free and is somewhat obscure. I have never heard of it until now. I am not even sure if it is an object oriented language which is almost needed when making a game. If you would like an easy way to make a game then I suggest using a python/C engine or C#. * Telling artists which modeling program, or graphics program, to use is not looked kindly upon. All that matters is that they are able to convert it into the format the game engine uses. * Your request for help sounds like a job posting that doesn't give as much information as you think it does. From a programer's standpoint I would love to know what you have done so far, what still needs to be done, if you are using Model-View-Controller code style or some other paradigm, what you have done so far, classes, bugs, data structures, and above all else what you have done so far. The difference between saying "You need to do this" as opposed to "I am working on something where i am doing this and this and that" can work wonders!

One of our developers, Tomkeus, has made a prototype of the construction layer: http://www.youtube.com/watch?v=T_9VprsYtyg Sample program can be found here: https://www.sc4devotion.com/forums/index.php?topic=9204.0 Basic idea, you draw out plans in vector format, but it is only built when you hit apply.

That is exactly what we are doing with citymania. In CityMania the Simulation (server) and Graphics (Client) are separate. It is good design to do so. Also it lets different clients be made, say the main 3d client for most computers and a slimmer 2d client for lower end computers and netbooks. It also makes multiplayer a snap.

SimCity 4 is a 3d engine actually. They did a very impressive trick to make it perform well. It can be replicated (I'll have to ask around to see if my engine of choice can do it, but I think so). I have been told that sc4 grid tiles are finite state machines and this is how the simulation is run by cycling through each FSM. To make a non-grid system this will have to make a departure from all the FSM. A new simulation system can approximate the behaviors seen in SC4 however. S grid can sldo still be used with non-grid code. This can be done at the UI layer so the simulation underneath would not care if the user is confined to a grid or not. This will result in a very close approximation to SC4 experience with the architecture in place for moving forward. If we had the SC4 source code you would be correct in that it could not just be added in. Luckily we don't. Everything will have to be added in so it can be designed from the ground up for the opportunity to do other things.

The reason why OTTD was successful is that they started with a clone. From there it expanded into what it is today. Linux as a whole does the same exact thing. All the popular distributions are almost always based on a previous system instead of reinventing everything.

The trick is to have a set game design doc. This is where Urbs Urbis fell apart. This is where most projects fall apart because there are, at best, amateur game designers all attempting to push their vision. We are lucky in we have a game design... simcity 4. Then you do FOSS style development of tickets/bug reports and developers picking up what they wanna do.

In short, not possible. The costs are astronomical. A conservative estimate for wages of a single, first year software developer is about $60,000 USD. That is a lot to fund. Solo projects like Love and Infinity took a good couple of years to complete (or are still in process). That can be upwards of over $100,000 USD. Funding is not the issue. The talent exists and is willing. The problem is, as I mentioned, there are divided. Each project has its own vision of what it wants to make. If the resources can be consolidated then there is a chance of something to lift off the ground. Right now at CityMania we're still in the prototyping phase. Thankfully we are all in general agreement about the features we want. Looking at the OpenTTD project I noticed some differences compared to all the city projects up to this point. OTTD started by exactly copying TTD. After that was completed only then did it add additional features. That is something that maybe should be considered. Looking to the future, such as gridless building, has to be kept in mind. Otherwise significant person hours will be wasted. And I responded lengthy to your most recent post.

There are a decent number of open source and closed source projects around. Check out the Gaming Talk forum.

Doofen, you have experience programming a city builder. We havent. We would love if you dropped by sometime and just shared your experience on what worked, what didn't, etc would be wonderful. Morris, we've actually implemented the essence of that feature, multiple people playing in the same region, or the same city. Every city will be its own process (to take advantage of any multicores). However the region is the upper limit for any game word. At some point a central lobby server can be implemented to those wishing to list regions open to the public. City sizes can't be unlimited though. There are limits to processing power which increase drastically with city size. The problem with doing too much verity at once means that none of it will work very well. Second Life attempted to do just as you described, but ended up not doing that great at anything.

There exists many open source 3D game engines, or 3D engine components. As you mentioned it is silly to create one from scratch. Quite frankly I don't see the need to generate a map of the entire world. The region introduced in sc4 seems to be the perfict scale for an individual player. Urban areas comprise of a very small percentage of the surface of a planet and, from a game design perspective, what is the point? I would rather have one great game, a city builder, vs 2 ok games mashed together. What I would much rather see is getting Cities of the World, City Mania (in Simtropolis 1000 thread), and the other city builder projects get over their differences and work together on a single project. THAT is worth the time and effort.

Originally posted by: Doofenshmirtz Hmm. Well if the project (most likely) chooses to use programming languages like Python rather than Game Maker i will attempt to learn them... I did try to learn C a while back, and i understand how they work, i guess i'm just not creative enough to see how 'with' 'and' and 'preprocessor directive' makes a city sim game, but i guess all things built from scratch have to be made with the trunk before the branches as such. I guess the main project would be making the core game engine. Not jumping ahead or anything but a few thoughts on how it would work... Simcity uses cellular automata, and i reckon from Game Maker experiments that this or a similar system is a good way of creating pseudo realistic development effects... etc... But this raises an issue. Tiles. The one most simple and yet most limiting aspect of any city sim game is tiles. It limits the position of buildings, allows for limited realism in road and other transport network design... The elimination of tiles would be a godsend if it could be done. But of course there simplicity and ease of use is also a factor... And in my various Gm experiments i've found no better system, aside from having no tiles but placing zones, buildings etc wherever you want.quote> CitiesXL did it, with some issues, but it worked. What we found was that a grid does help with construction, getting things to line up. AS an option the player can make a grid that they can freely rotate to snap things too. However i think of transport and terrain and other things and perhaps there could be some way to, rather than use the old 2d grid tile method, define an area, sort of like in paint with the free form cutting tool... I don't know. But i've had some vague thoughts that suggest if we want realism we should do away with tiles if it's possible. An example which i've had in mind for years is that of defining zones. In reality zones are of course often not simple rectangles of land. This brings me to another point. If you did away with tiles and replaced it with a free form development system, how would you do cellular automata type effects (finite state machines) for things like desirability, etc? What would be the... unit as such... for these operations? In Atocity i use tiles but for CA type effects such as desirability i use collision_circle, so for instance if a coal power plant is within the circle desirability drops by a given amount. I imagine most CA effects could be replicated in a free form environment... Perhaps instead of tiles there could be a much smaller array of points... quote> This part I haven't been involved with the discussion of. What I do know is that we're aiming for a resolution of 10m per "tile" I'm no expert on the matter but the finite state machine system seems a good one, but tiles seem to me positively archaic and limiting in realism. One of the main goals i'd love to see in a simcity game (one which i hoped to see in 5 actually) is free form transportation routes, realistic curving railroads and winding roads... Which would also suggest freeform terrain with a greater level of detail than the simple simcity 4 system.quote> Aye, I am just a hobbiest programmer I never had formal training. So I have giant gaps in my knowledge on how code is put together. I'm interested in seeing how things progress. Zones in reality of course are not so simple as in simcity. For instance we have height restrictions rather than 'high density residential'... And i don't think it would be too difficult to replicate this without overcomplicating the game. As zones are integral to any city sim, i think this is an issue worthy of discussing at this point, or at least having a think about. Also i remember a while back in a 'what do we want in simcity 5' thread at a forum (perhaps this one) that multi-zone developments were an idea for expansion. I.E a residential tower with shops at it's base, or a residential/commercial complex... This i wouldn't put as a high priority for discussion but i'd add it wouldn't be terribly difficult to include multi zoning. The main area imo to think about would be as i said before the limitations of the tile system with rgards to zoning. Personally atm i'd have it this way (feel free to criticize the system). Instead of tiles you have a much more compact array of points. You zone like in MS Paint by drawing lines, and the points within the defined region become 'zoned'. A point checks whether it has enough points around it to allow for development and if so a development is created by some rules, i.e if a road is near it faces the road, etc. quote> Aye, here, for residential at least, there are restrictions on the number of housing units per acre. One of our devs is working out an alternative to the rci model. Multizoning is an objective. But at this point i come to another far more important issue. Game speed. Which i think is more serious than any other issue i've mentioned. Simcity 4 lags on my 2005 compaq presario pc and whilst my pc is hardly top of the range stuff i imagine many players aren't 'l33t g4m3r5' with massively powerful machines that cost them 5 thousand dollars to get and which they'll replace in a few months when they're termed dinosaurs. As such i reckon game speeds a prime issue for discussion. In Simcity 4 of course you had graphics options like turning off shadows/cars etc... but you'd want everyone to be able to experience the game to it's full potential. In Atocity i use deactivate_all_instances and instance_activate_region to solve this issue, but my game is not as advanced as simcity 2000 let alone 4, and uses 2d graphics. For a fully 3d game like the one we're theorizing about you'd obviously need some method of cutting down gamespeed. At higher zooms for instance instance_deactivate is of little use. I've read about programming styles and various mechanisms of reducing game speed, but i also read advice not to use these methods as they will simplify the game and etc etc... So when designing the game engine i reckon game speeds a top priority for discussion. How to include high level realism without taking up too much memory... quote> We decided to have a server/client architecture right from the start, and even for single player. This opens up a lot of possibilities. First it means we can have different clients for the same game. There can be a flashy 3D GUI, or a simple 2D gui. The other option is that the simulation can be on one machine, but the gui client on another. The current implimentation is to make the simulation strongly turned based. The simulation will wait for the client(s) to send "ready for next turn" message. The simulation then moves forward one step (step currently = game month).

There is momentum for three different projects right now (check out the Gaming talk forum). I am going to try and see if we can't come together to work on one singular project.

Aye, having a good IDE can make all the difference when developing a project. As far as I know GameMaker is a well put together IDE, but there are other excellent ones for any language. C and Python work very well together, so in our project we'll have the ease of python with the power of C.

In my hobbyist opinion Game Maker is not the best option. The amount of time spent learning game maker could be used to learn a more robust programming language (Python, C/++/#, Java) or art tool (Blender, GIMP). It is also doesn't have native cross platform compatibility. As a linux user that is important to me. I'm sure the mac peeps feel the same way. What I am in full agreement with is we need to get all these duplications of work into one project. I'm with the CityMania team, you have Atocity, and there is CityWorld which I haven't found a lot of information about. Not to mention the existing LInCity and OpenCity projects. I know I am looking for a SC5-like successor, but in order to do that we need to get over differences, hash out a decent roadmap and code-structure plan, and move forward.

I'm under NDA still but, 1) I have the same issues but I've learned to accept the premise (make the game more game and less simulation) 2) It is for the sake of balance in the online feature so that those on fast speed will get richer faster than the poor ones. The online component feels like it is made to be played for a couple hours at a time, not massive building stints 3) As mentioned, people will come back 4) I like the woodsy feel 5) It is this way with any MMO, you have a finite number of characters. Or in this case, cities. 6) I think it comes with a limited predfinied set, but there will still be terraforming options for user created content. This is just my speculation though

The problem is everyone wants something different to make it their perfect game For example I believe that the best way for a project like this to success is to be open source but lead by a handful of developers with a clear roadmap, but accepting of input and patches that meet certain standards. Also releasing early and often. On the other hand throwing support behind lincity or opencity, or forking them (on good terms with mutual code sharing) could also be very successful. To your questions (on what I think is important): I think a 3d engine is a must, even if the game is 2.5 (ie buildings are jsut textures on rectangles, SC4 did this?) 3D engines have a lot of power with shaders, particle effects, and so on and it would be silly to not take advantage of them. Client/Server architecture. I really think the city builder genera has a lot of potential with multiplayer. SC2kNE is the only one that I know of where more than one person can build a city. Advancement in gaming conventions sense that time could make the experience more fun.