Finally you can download and test the game.
http://www.redrobotinc.com
If you have the time, try it out and send me some feedback. I am looking forward to your comments. (except please don't flame the sound, i know it sucks *g* searched a sound database for one day and thats it :p)
9084 - Red Robot Inc.
I played it through : ) I give it this is probably one of the best stuff (along with for example TimeWalker) I've seen done with Irrlicht. But, as the project clearly aims high, I also had high expectations and the critique will also be quite strict. Just keep it in mind : )
Okay, first of all the idea for the game is very nice. The looks are very polished but unfortunately, the gameplay is not so :/ First of all it would have been great to have some kind of on-screen GUI about health and ammo that I have left. Secondly I managed to to get the character into endless die loop for falling in the wrong place. On other note, sometimes the character dies when falling and sometimes it just respawns (I don't really know if there is a difference).
But the most major flaw I noticed was that you really can't trust the jumps. When it is the most critical the character just might make a smaller jump for no reason And then you die, of course...
Also one small thing I noticed, I don't know if the fps is capped or something, but it runs only 30fps on my Quad Core + Nvidia gtx460. But I don't really want to complain, as I really know how hard it might be to optimize this kind of game : )
There's some pointers, I'm a bit in of a rush so I don't have time to write more in-depth analysis : )
Okay, first of all the idea for the game is very nice. The looks are very polished but unfortunately, the gameplay is not so :/ First of all it would have been great to have some kind of on-screen GUI about health and ammo that I have left. Secondly I managed to to get the character into endless die loop for falling in the wrong place. On other note, sometimes the character dies when falling and sometimes it just respawns (I don't really know if there is a difference).
But the most major flaw I noticed was that you really can't trust the jumps. When it is the most critical the character just might make a smaller jump for no reason And then you die, of course...
Also one small thing I noticed, I don't know if the fps is capped or something, but it runs only 30fps on my Quad Core + Nvidia gtx460. But I don't really want to complain, as I really know how hard it might be to optimize this kind of game : )
There's some pointers, I'm a bit in of a rush so I don't have time to write more in-depth analysis : )
Played it too. It would be very awesome game, the best one I have played from irrlicht projects, BUT:
*FPS is very low on my 7600GT (17 FPS on 1280x1024 windowed, and funny but 25 fps on 2048x1152(native) fullscreen).
*Sometimes jump won't work as expected.
*No control over the view sometimes is pretty annoying.
*No life indicator.
*Surviving requires real skills(with that fps).
Overall it's good but REALLY slow.
Ah, and the game pops an error everytime I shut it down. Who knows why.
*FPS is very low on my 7600GT (17 FPS on 1280x1024 windowed, and funny but 25 fps on 2048x1152(native) fullscreen).
*Sometimes jump won't work as expected.
*No control over the view sometimes is pretty annoying.
*No life indicator.
*Surviving requires real skills(with that fps).
Overall it's good but REALLY slow.
Ah, and the game pops an error everytime I shut it down. Who knows why.
Thanks for your comments. Let's go through them
As you can see, sadly we are aware of all your critique points. Our main flaw was to start playtesting way too far into the project where most of these issues couldn't be adressed anymore without risking to miss our end of semester deadline.
It's a student project after all and thats one huge thing that I learned from this project. Test early on, test often and test on as many different hardware setups as possible. Oh yeah, and listen to your testers Sometimes we just overlooked the problems because we didn't have much time left.
Additionally, as you mentioned, the project is quite ambitious. I'd even say way too ambitious for the relativly little development time we had. For the next project I'll make sure to check thoroughly if the scope is managable otherwise you'll sth like this. An okay looking game with little flaws all over the place which diminish your game experience.
So thank you very much for your comments and I hope next semester I can present you with a game that is actually fun to play
Since this game was required to run in stereo 3d (and a 2d hud can mess with stereo 3d easily), we decided not to use a hud but instead have all the hud elements transfered to the main character. There are 3 white diodes on the back (health) and the arm shows your ammunition. We are however aware that playtesters often do not recognize these since they are not very visible.First of all it would have been great to have some kind of on-screen GUI about health and ammo that I have left.
That is unfortunate but yeah, it can happen in rare cases.Secondly I managed to to get the character into endless die loop for falling in the wrong place.
This is the same as question 1. Since you did not recognize the meters you of course cannot differentiate the two behaviours. One happens when all 3 health diodes are broken and the other happens if you still have some lives left.sometimes the character dies when falling and sometimes it just respawns (I don't really know if there is a difference)
This bug sadly was there since the beginning and however often I try to adress it I never could fix it. The whole physics simulation is run by Havok and I am not an expert in Havok. What I think happens here is that the character collides with the floor immediatly after the jump and therefore reducing the jump force.But the most major flaw I noticed was that you really can't trust the jumps.
Yeah the fps is capped because the physics simulation is run continuously after every frame. Not the best method but the easier one to implement. Next game I will try the more advanced method of asynchronous physics If you want to check your performance you can open a profiler by pressing 'P'Also one small thing I noticed, I don't know if the fps is capped or something, but it runs only 30fps
The game got several performance optimization cycles but in the end, the rendering part takes the most ms per frame. (See profiler by pressing 'P') At max there are 130k polygons on screen which, in my oppinion, isn't that many but I don't know what else I can do to optimize. I hope this isn't the fault of irrlicht because 130k polys isn't all that impressive, imo.FPS is very low on my 7600GT
We got that from playtesters aswell. We chose to do so because with our alternative play controls (WiiMote) it's hard to make aiming and turning work well together. And since we didn't have much time to develop this game we simply fixed the view. But you are correct, the level design sometimes forces you to go sideways which wants you to be able to turn. We should have created a better level design where you only want to move forward anyways.No control over the view sometimes is pretty annoying.
It's true, since the physics run each frame, the simulation gets quite ugly when you drop below 30fps. Since the computer at our university we had to present our game on is a power horse, we initially didn't care as much. Later when we got complaints from playtesters, we were screwedSurviving requires real skills(with that fps)
As you can see, sadly we are aware of all your critique points. Our main flaw was to start playtesting way too far into the project where most of these issues couldn't be adressed anymore without risking to miss our end of semester deadline.
It's a student project after all and thats one huge thing that I learned from this project. Test early on, test often and test on as many different hardware setups as possible. Oh yeah, and listen to your testers Sometimes we just overlooked the problems because we didn't have much time left.
Additionally, as you mentioned, the project is quite ambitious. I'd even say way too ambitious for the relativly little development time we had. For the next project I'll make sure to check thoroughly if the scope is managable otherwise you'll sth like this. An okay looking game with little flaws all over the place which diminish your game experience.
So thank you very much for your comments and I hope next semester I can present you with a game that is actually fun to play
Hey fmx, thanks for playing.
You are right, we were kind of overwhelmed with the scope of the project so actually making it fun kind of fell by the wayside That is definitly something I will look out for on the next project.
But, while answering to Tihtinen, specifically regarding the "jumps are not reliable" part, I came up with a simple idea to fix the jumping bug. I just disabeled collisions after a jump for 100ms. Not the most elegant solution but it works. So everyone who wants to give the game a shot now can at least rely on the jumps working
You are right, we were kind of overwhelmed with the scope of the project so actually making it fun kind of fell by the wayside That is definitly something I will look out for on the next project.
But, while answering to Tihtinen, specifically regarding the "jumps are not reliable" part, I came up with a simple idea to fix the jumping bug. I just disabeled collisions after a jump for 100ms. Not the most elegant solution but it works. So everyone who wants to give the game a shot now can at least rely on the jumps working