• Advertisement


  • Content count

  • Joined

  • Last visited

Community Reputation

705 Good

About starbasecitadel

  • Rank
  1. Just checked out your game's home page, it looks awesome! (pun intended) :)       Is it pure UDP or do parts of your game such as lobby use TCP?   If it is pure UDP it might help me with a side networking project I'm developing.   In general, I tend to agree with hplus, where you as much as possible should assume the internet is unreliable.  Doing so usually means a slightly worse game experience for the players who have great connectivity but a far more playable experience for the substantial chunk of players who have packet loss > 0.25% etc. 
  2. MMOs and modern scaling techniques

    I'm trying to understand what problem being discussed here is.   Is it basically in a game like WoW where there is some big event and half the Alliance players on an instance all decide to meet in a single shopkeeper's room in Stormwind?  So you have to manage an exponentially (? or at least non-linear) increasing number of interactions?   eg:   1 player  =  0 interactions 2 players  = 2 interactions 3 players =  6 interactions 4 players =  12 interactions 5 players = 20 interactions 6 players = 30 interactions ....... non-linear scaling ....     My only comment here isn't from a technical point of view but from a playability point of view.  After a certain number of players, does it really matter if things aren't being perfectly computed?  For example if 20,000 players are in a single bar, you wouldn't even be able to see anything at all if it was rendered "correctly".  It might be a better experience for players if you just show a random say 30 or even 200 players maximum and it is what it is.
  3. Great posts everyone.  I am a huge fan of LoL and more broadly other kinds of real-time team multiplayer strategic/tactical games.     I am very interested in this topic, and want to do something sort of similar (was thinking more along the lines of a King of the Hill style game for Tablet, for more of the casual player / very quick games) but other more hardcore versions too.  I have too much on my plate right now so am not working on it currently, but hopefully one day would love to explore this much more.   Just to chime in on some random points:     - in terms of permanent death as opposed to resurrection timers: my initial gut feeling was this would be unappealing.  A simple mistake and you would have to sit out potentially a long time until the next game spawns.   That said, I do remember playing I think it was Ghost Recon or something like that 10 years ago and the hardcore death mode was a great part of the fun.  Even if I got killed in the 2nd minute and waiting 15 min for the next game, at least I could observe the other players which was still fun and it was good motivation.  So I guess my point is perma-death can work effectively, but probably would not be popular for casual players.   - as far as other kinds of income etc, I don't see anything wrong with Starcraft / LoL where you buy upgrades over time.  Then it is just a matter (once again) of how casual versus hardcore your game is.  On the casual side, simpler is better, on the more hardcore side, complex can be better (but even then, ideally with a simple learning curve in terms of at least getting into the basic action).      - the commercial breaks concept is pretty interesting.   Doesn't make any sense for a game with a small number of players, but for something around LoL scale that could be an improvement.  Bathroom / beer breaks are good as well as much as commercials are unpopular, they do provide an important function by helping fun the game developers, professional teams etc.   - love the coaching integration idea.   The game developer could even take a split (say 20%) of the coach's income but in turn provide excellent tools for real-time coaching.      - for gold/income generation this is a great game design topic.  I actually like LoL style jungles, but you could even go for a mix of Starcraft and LoL ideas by putting in a mini-version of mineral + gas mining.   So any player could try to set up their own resource gathering bases, and as well they could be destroyed by enemy players.  Or it could even be simpler than that, it doesn't have to be a full on "Command Center" style tech tree Starcraft style.  It could even be as simple as TF2 style Sentry Guns etc that also provide gold income over time (and give gold to the enemy team if destroyed).   
  4. High Speed Gaming Network

    For various reasons (mostly relating to being unhappy with the main UI -- the basic feel of piloting starships and firing weapons was not satisfying) I stopped working on Starbase Citadel months ago, the project is on-hold. The last few months, I've been working on a new project which is to create a High Speed Gaming Network, fueled by personally facing severe lag when playing League of Legends. The goal is to create a commercial service that will substantially reduce lag and disconnects for League of Legends players, and ultimately other games and possibly apply to general internet usage as well. The application is only for Windows Desktop. It reduces packet loss by sending redundant UDP packets-- very similar to how Halo sends 2 redundant packets for each "real" data packet. In addition, just by placing the proxy/tunnel servers on high quality hosts, it allows the user to upgrade the routes to more premium providers. Further route optimization techniques are planned in Phase 2. At this point I believe I have the UDP side working, at least it passes my internal tests. However, League of Legends requires the TCP IP Address to be the same as the one UDP traffic uses, so I need to tunnel TCP as well. I'm making steady progress on the TCP aspect, but it is more complex than the UDP tunnel as it requires low level packet injection that simulates TCP handshaking.
  5. Unity 2d game engines/sdks

    Since Starling is just a library for Air/Flex SDK, it has access to everything those platforms from Adobe can offer. Including http://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/flash/net/DatagramSocket.html and much much more.       Thanks!  
  6. Unity 2d game engines/sdks

      This looks great, I'm leaning on going this route.   A friend recommended it 3 years or so ago but at the time, it didn't support nearly as many platforms or have any real multiplayer (UDP sockets) support.  Now it looks much more polished.  The only downside is the IDE is Windows only, ideally the IDEA could work on Mac OS X too but that is ok (my laptop is Mac, home desktop is Windows).    But it looks like a great platform.. probably I'll go this route.       This one looks really good too, but I didn't see any networking (UDP Sockets) support in a quick look through the API.     Thanks!
  7. Unity 2d game engines/sdks

      That does look really good, thanks!
  8. I'm pretty familiar with the Game Engines out there for 2d games, for mobile.  (Personally I love Corona for that).   But, for the Desktop PC side for 2d action games with multiplayer capability, I'm out of the loop on what indie devs are using these days.  This is just prospective, kind of curious, not a pressing need but wanted to look into it and maybe start learning one in my off-cycles.   It should be Steam and SteamOS compatible (is that Linux?), 2d graphics (3d not needed) as well as multi-player ( UDP sockets).  I program in many languages but prefer a high-level language like Lua or Javascript, not C++.     Export to Windows desktop, Mac desktop, and SteamOS.  Ideally it has a very rapid dev/testing cycle where your code changes can be tested within seconds, not a 2 minute compile.   Unity3d (in 2d-mode) and Love2d look the best so far.   Love2d looks like a much smaller learning curve, but it has 1/100th the user-base which is a big disadvantage.  Anything else?
  9. This is a continuation of a 1.5 year old thread so it wouldn't let me reply to the original:   http://www.gamedev.net/topic/633066-large-geographical-area-network/   But, for anyone interested, thought I'd give an update on this.  My current project is to try to implement some of the concepts described in that thread.   The motivation for this was I play League of Legends a lot, and starting about 1 year ago, ping and packet loss increased substantially even though my provider was the same.  This was due to a standoff between Cogent and Comcast, two major internet backbone providers that resulted in frequently congested routes for those on the East Coast.     I will say things have got much, much better the last 2 weeks as Riot (and possibly Cogent+Comcast) has made network improvements.  Even so, lag is still a huge problem for some players, and a minor/occasional nuisance for many players.   My current project is a high-speed gaming network that's goal is to mitigate a lot of the network congestion.  It will do this by several methods:  1)  by proxying through server(s) on premium networks, it can optimize the network path in many cases  2) There are techniques to optimize the network performance further in terms of connectivity to/from the proxy servers that can reduce packet loss and jitter of the overall network connection.   In terms of development, I have a very early version that works for UDP-only and just 1 proxy server on a simple UDP echo program.  I need to test that with an actual UDP only-game.  Next I'll work on implementing the TCP-tunnelling side, as most games use a combination of UDP + TCP, where the lobby/match-making/authentication uses TCP and the actual gameplay is UDP.  The IP's have to match, so you have to tunnel/proxy both.          
  10. My current game project is creating a high-speed gaming network, specifically optimized for League of Legends but also extendable to other games down the road.   I have a very early version ready for testing, but it only supports UDP tunneling, not TCP.  I certainly need to add TCP tunneling as well but that is still a few weeks away even if it goes smoothly.  In the meantime, I'd like to find a free-to-play Windows 7 game that does not use TCP, only UDP, to test my tunnel/proxy with.   Anyone know any?   I've tried several games (League of Legends, TF 2, Dota 2, Netrek) but none of them work since they all use a combination of TCP and UDP and expect the client IP address from TCP to match the client IP address from UDP.           
  11. multiplayer map / team composition

    That is a good point, it does hurt creativity in the meta-game.  Though it also prevents possible troll-type situations (all Engineers, all Spy's in TF2? A server I was on put a class limit for this various reason)   Btw, on this subject, LoL just announced they are adding in support for strict team comp (I think as an option, not a requirement):   http://forums.na.leagueoflegends.com/board/showthread.php?t=3951169       edit: Looks like this is more flexible than I was thinking, they will even support things like double jungle etc.  So it is more along the lines of something to make Matchmaking easier (eg, make it a better user experience for you to get the roll you want, without really limiting the types of meta too much).  
  12. multiplayer map / team composition

    I'm not dead-set on doing the new map / composition as the next thing.  Some other possible big features to do instead could be (in no particular order):   - Admirals (an entire new kind of player that each team has 1 of, I'd create a new UI for them that is more strategic and based on pings and coordination)   - Economy system.  Allow ability to mine resources and purchase weapons, armor etc upgrades   - Experience system.  In-game leveling system.   - More starship types (Assassin-class starships, Spy-class starships etc, more differentiation for battle ships and support-class ships)   - Carriers / carrier groups.  (A new kind of starship that grants Aura bonuses and refueling to Starships of a compatible class that dock with it)   - Chain explosions.  Explosions themselves can do damage, often times causing chain explosions (domino effect).     - Starship Collisions cause "bouncing" ( a friend suggested this one, if you hit another Starship you bounce of them )   - Pilots + Droid Units.  You can think of these as the LoL equivalents of Summoner Spells.  The idea here is each Starship player would chose 1 Pilot and 1 Droid Unit, each of which grant probably 1 passive bonus and 1 Active ability on a cooldown.   I still have hopefully less than a month or 2 to finish off what's needed to get a bare minimum alpha/demo.  But looking after that I'm pretty conflicted on where to go.  I see a lot of things that could really enhance the game but it's hard to determine the best sequence of anything here, whether along the lines of my original post, some of these ideas, or a new direction entirely.    If anyone has further feedback please chime in  
  13. multiplayer map / team composition

      I am somewhat worried about in game communication too, but I really wanted to go higher than the standard 5 per team MOBA's for several reasons.  The first is it gives the game a more epic feel, but also simultaneously I think it can appeal to more casual mobile players more.  When there are just 3 or 5 people on your team, leaving the game makes a much bigger impact than when there are 9.   Normal (unranked) games will make it very easy for players to join within seconds, and then leave say 5 minutes later, without hurting their team.   That said, I plan on adding several kinds of pings in terms of which planets to attack, to retreat etc.  That should be much faster than typing (though less good than Bluetooth voice communication).     Additionally, (haven't decided if this feature will make it to launch yet), but at least down the road, 1 player on each team will be the Admiral.  The Admiral's primary responsibility is dealing with exactly the problem as you describe-- coordination.  They will not be playing tactically, but strategically, giving instructions (pings essentially) to their team to help time attacks and retreats.         This is something I've been torn about.  I am really tempted by such approach, but my fear is launching with the game as is (or even in a month or two) it would just be so unpolished and lacking features it would not attract many (any?) players.   In such a scenario, even if the game slowly gets better, there might just be so few players that new releases aren't able to generate sufficient buzz to increase the player base rapidly.   I think the chances of success would be higher coming out of the gate with a more polished product, hopefully with a publisher helping on the marketing side, even if it takes much longer to launch.    Maybe I'll split the difference here and open it up sooner than I was expecting as an early alpha work in progress, though that limits it to Android (iOS requires games to be "complete" / not marketed as alpha or beta).
  14. multiplayer map / team composition

        You will be able to pick your role.  There will be different queues for each role, so if you queue as "First Available" (the default), you are always instantly guaranteed to be place in a game instance (unless every single game instance is full).  For some of the more in-demand roles there will be a queue where you have to wait until a slot is open to get in in some cases       I'm planning to pitch to publishers sometime next year, and if I find a good fit with a publisher who can help with polish and marketing, I am confident the 9v9 maps will indeed be constantly full or nearly full.   Whether that ends up being 2 full instances or 100's I won't venture to guess, but in general when you join, the chances will hopefully be that your instance is full/near-full.      During periods of slower activity (3am), game instances will automatically be shut down so that players are assigned to a smaller number to keep them full.   If there isn't a full stack of players, the game does degrade in the sense the other Starship slots etc are played by AI Bots (at least in "normal/unranked" mode), which to most people is less interesting.  But the goal is most instances will be say 80%+ filled with players, and with AI Bots taking over as placeholders just during transition as players leave and new ones take their place.
  15. This is for a 2d MOBA for Tablets (initially Android only, iPad too eventually) game, more details available at starbasecitadel.com.  I’m reasonably close to being done with my first major milestone which is an early, fully playable (though very buggy and lacking much features) prototype.  For the rest of this discussion I will make use of LoL terminology just as a point of reference.   As currently stands, there are 2 teams.  Each team has 1 starbase and the rest are all starships (currently each team just has a single all-round class of starship).  Planets are placed randomly around the map upon initialization, with about 1/3 assigned to each team or neutral.  One of each team’s planets is randomly designated as the Citadel.  The first team to capture the opposing team’s citadel wins the game.     Planets you own produce armies on occasion, and to capture another planet, you go to one of your planets with an army, pick it up, and drop it on an enemy planet (making it “neutral”) or onto a neutral planet (capturing it for your side).  The benefit of owning planets is they can be upgraded with defense grids, giving your team a tactic advantage in fights near them as well as fuel and repair stations, which help your team project a deeper presence.  I haven’t implemented minions yet, but each planet will produce 1-2 minions periodically.   Currently, you join the game and are randomly assigned to any open slot on either team, possibly including the Starbase if it is available.   So this is all well and good :)  But I think the game just needs more depth and strategy to it, even out the gate.  So here is my new proposal (see attached image too), that adds a lot of structure to team composition and the map:   There will still be 2 teams.  However, instead of each team being 1 race as it is now, instead each team will consist of 3 races.     Each Race will consist of 3 players, making a total of 9 players per team.      Each team will have 1 Race designated as the team’s Primary race that consists of the following:   Starbase   Explorer Starship x2   Each team will also have 2 other races, in this configuration:   Cruiser Starship x2     Support Starship    Cruiser Starships are like all round offensive ships.   You can think of Explorer Starships like Junglers, they have high sustain and are the best for soloing “jungle” camps, ganking, and mobility. Support Starships are a cross between “support” champions in LoL (healers etc), and the Engineer in TF2.   At the game start, each race will have a Capital planet, with the Primary race’s capital being the Citadel.     Capital planets, Citadels, and Starbases each will give a Racial proximity bonus to nearby starships of their race.    Citadels and Capital Planets will have a higher tier of defense grid than other planets can achieve.   Onto the proposed map:   Team A will consist of Maroon, Red, and Orange.  Team B will consist of Purple, Blue, and Cyan.  Each color represents a distinct Race.  Each planet that your team controls will periodically produce minions.  The arrows show the direction these minions move.  I put arrows for Team A’s left side, but it is symmetric, so the right side of the map is the same thing.  Team B will have minions too of course, again symmetric to what you see, create 2 “lanes” that go from the Citadels’s of each team roughly towards the left and right sides of the map and then back in again toward’s the other team’s Citadel.   There are Jungle areas which are encounters that can be fought over to get experience and buffs, and a “Super Jungle” which is a particularly tough one that isn’t winnable until later in the game when starships are leveled up enough. 
  • Advertisement