Announcement

Collapse
No announcement yet.

[Video] Project Free Fall [FPS-Z] : Multiplayer Deathmatch

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    [Video] Project Free Fall [FPS-Z] : Multiplayer Deathmatch



    The above footage is from a recent multiplayer test. The game is inspired by Quake, Tribes, and other arena shooters. At the moment, the game uses stock assets, but they will be replaced over time. Maps are planned to span small arenas to large terrain spaces and most anything in between(or outside?). Project Free Fall will focus on arena gameplay initially, as arena comes out of the box for UDK, but will begin to incorporate more strategic(tribes like) game modes that put more focus on teamwork. The source code can be found on www.projectfreefall.com/ for download and viewing.

    I'll be posting some tutorials on net code and the movement system once I have a spare weekend.

    #2
    Continuing the first post, here is a video that covers the general idea of unrestricted speed. The video is quite old at this point, and the movement has changed significantly, but it illustrates how fun and crazy unrestricted speed can be.

    Comment


      #3
      Continuing from the previous movement video, here is a video that covers the most recent changes to movement.



      The recent changes rein in movement to a more reasonable pace. Speed is still unrestricted, but changes have balanced the jetpack in such a way that players now have to focus more on using terrain to gain speed, rather than only for changing direction.

      Comment


        #4
        For a final video covering the current state of the project, this video covers the weapons that have been added to the game. Most are pretty standard for arcade shooters, but implementation can differ in interesting ways for some.

        Comment


          #5
          good work there, you can really tell the quake/tribes inspiration
          also very respectable that you release your sources

          I know you're focusing on gameplay rather than levels for now, so I'll only talk about gameplay here. I'm a long time quake and spare tribes player and I really enjoyed the fast pace of those games. somewhat more recently (since quake3 died) I moved to counterstrike (source) and have enjoyed it as well. I haven't even played any of the other popular shooters (COD / BF / halo) but from what I've seen they all look the same to me (better graphics every iteration, same gameplay, same AI).
          that said, after playing CS:S for so long, the pure twitchy gameplay of quake seems a little to basic for me now, and I'd miss some of the real aiming and strategy of CS:S if I played quake again.

          being an indy developer pretty much forces us to do something original, else we'll end up with just "a limited and ugly version of quake/cod/skyrim/wow/starcraft/whatever - a good effort but people'd rather play the better AAA title"
          so my question here is: what are you doing here to innovate? what's your take on breaking the tendency of better-looking but overly repetitive FPS's?
          the answer to that question might be the key to make your game really interesting (besides the usual: quality, polish, concept, etc... but AAA FPS's have tons of that already and they aren't that interesting)

          also what do you mean with "emphasis on ultimate player skill"? (sounds like you want to make real aiming important, but your gameplay suggests twitch is more important) - or maybe we just think differently on what "player skill" means

          Comment


            #6
            Originally posted by Chosker
            what are you doing here to innovate? what's your take on breaking the tendency of better-looking but overly repetitive FPS's?
            My innovations will come from my take on what an FPS can be. A worrying trend of late is the lack of exploration FPS games have made since the successes of Halo and COD. Discounting a few outliers, FPS games have almost all become slow ground huggers that submit to ultimate user friendliness, rather than explore new ideas. I aim to make Project Free Fall play like a game in the classical sense; like Chess or Go. Chess and Go are games based entirely on game logic rather than real world logic; there are no arbitrary real world rules that intrude into what might hinder gameplay.

            It's my feeling that FPS games have lacked long term viability because they've overly focused on superficial eye candy. Most games that last long focus on fun player interaction, rather than passive observation. FPS games have long been like CG loaded summer movies; they can be good fun to watch once, but unlike their plot loaded, brilliantly executed, brethren, summer action flicks lack the same seemingly infinite rewatchability. I don't see why the real depth that's explored in games like StarCraft 2 and LoL can't be implemented in the FPS genre. I plan to take advantage of the FPS genre's ability to directly challenge a players mental athleticism but also introduce mechanics and balance that allow for depth in strategy and execution.

            Originally posted by Chosker
            also what do you mean with "emphasis on ultimate player skill"?
            Our experiences in Quake probably dictate the differences in how we infer what "ultimate player skill" is and can be. When I speak of player skill I tend to apply the same reasoning that might make a martial artist adept at their art. In order for real player skill to exist, the medium for practice must be as consistent as possible; in martial arts this means regulations that minimally effect what a practitioner can do; in games this means lack of random behavior. Coming from Counter Strike, I'm sure you've learned to love the reproducible nature of weapon recoil and grenade throws; because the weapons are so reliable, they become tools rather than toys. In COD, and the army of clones it's spawned, randomness and suppression of player skill have turned the weapons into a bunch of toys that might be well used, but lack the reliability that might allow them to become tools.

            As for player skill between Counter Strike and Quake, coming from the Quake dueling scene(as well as having played many CS scrims), I would argue that CS is the more twitchy of the two games. If the games could be martial arts, CS would be similar to kendo in that there is a large focus on pre-planning but normally a very short period of execution; competitive CS rounds tend to be very short once the teams have engaged and individual engagements are often sorted by a single headshot. If Quake dueling could be a martial art, it would likely be a mix of Judo and Jujitsu where exact planning is largely replaced by practice based on reaction and telegraphing of movement; Quake duels, professional ones at least, place huge importance on real time mental decisions and rely on constant reaction to changing situations. I wouldn't argue that either game allows for more skillful play, but they are different. The main thing is that both games get out of the way and allow player skill to be the focal point. Both focus more on spectacular player skill that will almost always remain amazing rather than pretty particles that only become dated.

            Comment


              #7
              Originally posted by gironecna View Post
              A worrying trend of late is the lack of exploration FPS games have made since the successes of Halo and COD. Discounting a few outliers, FPS games have almost all become slow ground huggers that submit to ultimate user friendliness, rather than explore new ideas.
              I plan to take advantage of the FPS genre's ability to directly challenge a players mental athleticism but also introduce mechanics and balance that allow for depth in strategy and execution.
              uh, you seem to have a clear choice in the typical 'graphics vs. gameplay' paradigm, would rather spend time adding depth instead of flashy graphics and cutscenes, etc. but besides 'exploration' and 'enforcing player skill', and maybe 'un-dumb the genre again' I don't really see much of innovation yet. you are clearly set out to do things differently but you're not going in depth with concrete ideas about how you will.
              I'm not bashing you, I'm just curious about how your game will actually work to make all your ideas truly be there.

              Originally posted by gironecna View Post
              I aim to make Project Free Fall play like a game in the classical sense; like Chess or Go. Chess and Go are games based entirely on game logic rather than real world logic; there are no arbitrary real world rules that intrude into what might hinder gameplay.
              never played Go, but Chess seems to have some very arbitrary rules.
              maybe I haven't played any recent FPS's to know, but typically an FPS's "arbitrary rules" can be reduced to physics, character movement and weapon functionality (which in theory is "design", not just arbitrary). what am I missing here?

              Originally posted by gironecna View Post
              As for player skill between Counter Strike and Quake, coming from the Quake dueling scene(as well as having played many CS scrims), I would argue that CS is the more twitchy of the two games. If the games could be martial arts, CS would be similar to kendo in that there is a large focus on pre-planning but normally a very short period of execution; competitive CS rounds tend to be very short once the teams have engaged and individual engagements are often sorted by a single headshot. If Quake dueling could be a martial art, it would likely be a mix of Judo and Jujitsu where exact planning is largely replaced by practice based on reaction and telegraphing of movement; Quake duels, professional ones at least, place huge importance on real time mental decisions and rely on constant reaction to changing situations. I wouldn't argue that either game allows for more skillful play, but they are different. The main thing is that both games get out of the way and allow player skill to be the focal point. Both focus more on spectacular player skill that will almost always remain amazing rather than pretty particles that only become dated.
              I agree with how you compare CS and Quake to martial arts. However I still think Quake is more twitchy. In Quake you can start firing and correct your aiming as you go, in CS that's usually a deadly mistake. In CS you also need to stop for a fraction of a second to aim properly before firing because of locational damage, while in Quake any shot deals the same damage so it's more a matter of who lands more shots in less time, period.

              Originally posted by gironecna View Post
              Our experiences in Quake probably dictate the differences in how we infer what "ultimate player skill" is and can be. When I speak of player skill I tend to apply the same reasoning that might make a martial artist adept at their art. In order for real player skill to exist, the medium for practice must be as consistent as possible; in martial arts this means regulations that minimally effect what a practitioner can do; in games this means lack of random behavior.
              I agree that "player skill" is something that must be built upon, and "ultimate player skill" is what differentiates good players from great ones. But to me, proper aiming is an important part of what we call "player skill".
              It just seems our definitions of "player skill" are somewhat different.

              Comment


                #8
                The levels in your game appear to be quite huge! I am trying to get such huge landscapes to work for days now. However if I use dynamic lighting only the shadows either have shadow acne or eat A LOT of performance. If I use lightmass for the lighting the build times take ages if they complete at all (depending on landscape size). So could you explain a bit how you got those big levels to work well please? I'll download your source code now to see if I can learn something about it from that.

                EDIT:I checked the content of the source files. hmm unfortunately only the depriciated terrain seemed to be included?!? So how much of the playable space in a UDK map do you use? And how did you create the landscape/lighting for the game?

                Comment


                  #9
                  Originally posted by XilenceX
                  Could you explain a bit how you got those big levels to work well please?
                  Crank the light map resolution down and bake using light mass. They can take quite a while to cook, but I just make sure to only cook when I feel that more changes don't need to be made. If you want to make huge maps, I suggest breaking up what the user can see at any given moment. One issue I've found is that having a massive landscape that can be completely seen is that it can be pretty hard on framerates.

                  The source code is meant to be paired with a fresh install of UDK. And the maps were made with the landscape system. I used the terrain system a couple times, but found the landscape system to work better. If you want to take a look at the map itself, you might try pulling it from the demo and dropping it into the UDK that's running the PFF source code.
                  Originally posted by Chosker
                  I don't really see much of innovation yet. you are clearly set out to do things differently but you're not going in depth with concrete ideas about how you will.
                  I'm not bashing you, I'm just curious about how your game will actually work to make all your ideas truly be there.
                  Yea, the game is very standard at the moment. I've been focusing on the fundamental movement and interaction systems for a while now, and they are finally feeling good enough where I think I can begin to explore some interesting ideas. Two ideas I plan to explore are flexible gravity and match progression. I also have been long at work brainstorming better ways to implement standard gametypes such as CTF and Clan Arena. CTF especially is a gametype that has remained unchanged for a long time; a few games have tried minor changes to flag behavior, but nothing has really broken away from the idea of constant capture points or bases. I remain a bit vague on the subject because I would rather not risk having the ideas ganked; I have them committed to paper, but not to code at the moment.
                  Originally posted by Chosker
                  I haven't played any recent FPS's to know, but typically an FPS's "arbitrary rules" can be reduced to physics, character movement and weapon functionality (which in theory is "design", not just arbitrary). what am I missing here?
                  Air strikes, tactical nukes, character progression, weapon unlocks, bullet time, and generally pointless fluff that has been added to FPS games lately. Most FPS games haven't been designed as games so much as they've been designed in a way someone can pretend to be something else. For single player, lot's of fluff can be harmless fun, but in multiplayer vs scenarios, single player campaign mechanics tend to break game balance. Call of Duty is a great example of a game that is more role playing rather than actual game; it has basic movement, but ruins gameplay by artificially granting players an advantage based on time played, and introduces gimmicky mechanics such as nukes which are awesome in coop and sp, but are infuriating in vs.
                  Originally posted by Chosker
                  In Quake you can start firing and correct your aiming as you go, in CS that's usually a deadly mistake. In CS you also need to stop for a fraction of a second to aim properly before firing because of locational damage, while in Quake any shot deals the same damage so it's more a matter of who lands more shots in less time, period.
                  In the lower levels of play, you can certainly fire and correct, but at higher levels, everyone shoots more or less perfectly. CS is definitely different in that at both high and low levels of play aim is a critical skill. However, Quake dueling ceases to be aim heavy at higher levels of play and instead becomes more about control. Because aim is so strait forward in Quake, it just becomes another tool. In Counter Strike, aim is more an artful skill. I would agree that the aiming skill of top CS players is probably better than that of Quake players, but that's because CS is so much about aim. Quake players could focus on better aim, but because the skill becomes a tool, a player normally will find gains in other areas to be more important and productive.

                  Also, first to fire rules only apply for the hitscan weapons of Quake. The rocket launcher, probably the most used weapon in the game, is not hitscan and requires a fair amount of prediction beyond simple telegraphing of velocity. The railgun and LG are both very powerful, but are far from the only weapons used frequently. CS aiming is a bit more demanding when compared to the hitscan weapons of Quake, but it also prevents the game from being athletic due to how slow players have to move to be able to function. In the end, CS style aiming is just not suited for a game that's meant for speed. The chaotic nature of damage prediction introduced by complex hitboxes is compounded by the fact that projectiles aren't reliable enough over the internet to justify their inclusion. Range unlimited hitscan also doesn't lend well to games that are expected to have open maps with little cover; hitscan is better suited towards games that allow for easy breaking of line of sight.
                  Originally posted by Chosker
                  to me, proper aiming is an important part of what we call "player skill". It just seems our definitions of "player skill" are somewhat different.
                  For me, proper aiming, or rather the ability to shoot a head, should only be implemented in games where it makes sense to do so. When shots are simplified to a binary hit or miss check, damage becomes perfectly reliable. Perfect reliability is something I desire with Project Free Fall because I hope that players will become more involved with overall strategy, rather than moment to moment strategy. Using Quake and CS to illustrate what I mean, in CS there is virtually never a situation where killing an opponent might not be a good idea. In Quake dueling, killing a player is firmly an option; there are plenty of situations where killing an opponent may be the worst option even when behind in frags. Lack of respawning dramatically changes how players approach similar gameplay.

                  Comment


                    #10
                    Originally posted by gironecna View Post
                    Crank the light map resolution down and bake using light mass. They can take quite a while to cook, but I just make sure to only cook when I feel that more changes don't need to be made. If you want to make huge maps, I suggest breaking up what the user can see at any given moment. One issue I've found is that having a massive landscape that can be completely seen is that it can be pretty hard on framerates.
                    Hi thanks for the hint about that! However what I was more curious about is how you have set up your light(s). Currently I always use one dominant directional light and one skylight. So you do the same? And if so what settings for the dominant light? Also my terrain has a 1024 sized lightmap atm is that about the same for your level? Since you talked about spliting the terrain up I assume you are using static meshes for the terrain too?

                    Comment


                      #11
                      I am just using the defaults found in the template maps for my lighting setup. As for the light map, I just have set the resolution to 1.

                      When I mentioned breaking up terrain I meant visually. The maps in the first videos can be seen in their entirety when the player jets above the hilltops. I've found that introducing cliffs, mountains, and other features, which prevent the player from seeing literally all the map at once, has helped with frame rates. I'm running a GTX 460 at the moment and can struggle to pull ~45 fps when I can see the entire map. Plenty have video cards that are worse than mine, and they often struggle to pull more than 30. I've also found that aggressive LOD switching for the terrain does very little for performance. I'll also have to admit that I'm not very well studied on the aesthetic side of UDK. I come from a programming background and I have learned enough to get by visually. So I might not be the best person for advice.

                      I am not using static meshes for the terrain, I am just using the UDK Landscape system. The snow map was made by someone on my forums that generated a height map, mirrored it, then applied it to a landscape along with some smoothing.

                      Comment


                        #12
                        Swapped the video of the first post. It better shows the state of the game...

                        Comment


                          #13
                          I'm having an online test Friday, April 5th, from 10am-12pm & 4pm-6pm PST(Saturday morning for Australians). For anyone interested, in the test or just having some fun offline, the tech demo can be download from www.projectfreefall.com For anyone that's read through the thread before, whats new now is vastly better net code and minor mechanics changes.

                          Comment


                            #14
                            Session 2 has started. http://www.twitch.tv/saccaed for my stream. I'll be answering Q & A along with any gameplay happenning.

                            Comment


                              #15


                              Recently had another multiplayer test. The latest net and movement code is functioning much better than previous builds. Planning for some more multiplayer tests the weekend of Quake Con.

                              Comment

                              Working...
                              X