Plan for Thomas "Man of Ice" Lund
by Thomas \"Man of Ice\" Lund · 08/11/2004 (1:30 am) · 6 comments
We did it!! \o/
The first "Game in 2 Days" event is over, and we survived through the perrils and ecstacy of the work.
Short brief: we met 3 guys at my house for a 2 day session to get away from our usual game, and do a GID event. We wanted to have 2 days now that we physically sat down together.
Our rules: talk about and make an overall game design (1-2 pages) to set the major goals prior to meeting up. Then spend 2 days on creating the game.
Goal: primary goal is to learn and to have fun. We decided to go with gameplay a little more advanced, and where we needed to develop some engine extensions.
Game: a vehicle shooter, where the goal is to capture victory points by camping certain areas on the map. We wanted to give it "an indie twist", so we wanted to create the vehicles as bugs with weapons strapped on. The vicotry point locations are "ant hills" protected by vicious army ants guarding the areas with turrets. So the player has to race to the areas, fight the ant turrets as well as the enemy bugs. The game is played in 3rd person, and the world should be set on a lawn (with trees = grass etc)
The team:
Niels - 3d modelling
Morten - GUI and level editing, 2D art, basic scripting
Thomas - engine coding, scripting
What went well
We actually made a game in 2 days. Simple, and not containing all the nifty ideas we had, but its playable and fun.
We learned A LOT. Our "real" game is a action adventure without weapons, so we had to learn the Torque weapon system, mounting them on the player, projectiles etc. We never played with the LOD stuff in Max, so we had to learn that too. In 3rd person there is no crosshair, so we had to create a "3rd person 3d crosshair", the level designer had not really worked much with Torque yet so he ahd to learn the various mission/terrain editors as well as GUI editing.
Learning the dynamics of a GID event is fun too, because in retrospect its all about not making mistakes and knowing how to do things right the first time, where to look for info, resources and script code you made for other games and thus saving time. (This will be repeated in the "what went wrong"). I think we did quite well getting around the issues, and knowing when to drop it and attack a problem from a different angle.
One thing that worked out great was the division of work. Almost no overlap in work areas make for effective development.
Usage of subversion. Not having to send 20 MB files over emails because you changed 1 line of code. The ability to have it merge your changes etc. That saved us lots of grief.
Planning up front what to do. That saved us a lot of time, not having to brainstorm and waste development time on discussing ideas. Not that the plan survived more than a few hours, but without it we would have been lost.
Its also great fun sitting together and do team bonding, instead of doing it over Skype/Messenger!
What went wrong
First of all trying to get physically together takes time. Everyone has their own computer, it has to be set up and get onto the network. I run a pure wireless network at home......and none of the other 2 do that, duh.....and so we wasted quite some time trying to get Niels' PC running through a laptop with a wireless card. 3-4 man hours spend on that and we had to give up, sharing content on CDR instead. Better preparation would have had us running faster.
Having unrealistic planning. We tried to limit the game, but its clear that "development takes time". More time than expected and planned for. So our "evolutionary" approach to game dev with constantly playing the game, figuring out what is fun and what is not, fixing that and then go back to playing again - unrealistic. We were so busy creating "must have" code and art that we almost didnt get to play the game.
We wasted several hours doing things like:
* trying to make our own turrets. We followed Paul Dana's guide, but never got it working. It was written for Milkshape we think, and we worked in Max. The turrets got in fine after a while, but the nodes were turning in the wrong directions - shooting up in the air. Ended up using Paul's turret dts.
* messing around with the player model. We created the bug as a player model and first tried to keep is as a biped rig, attaching mount0 on the head of the biped for the turret. But the model went bananas in the engine with exploded polys all over the place. So we made it as a simple mesh instead with some simple animations for the walking etc.
* trying to get the turret mounted on the player. We spend quite some time getting the turret onto the player, and ended up having to use a standard weapon with mount0 on the back of the player. This having the problem of the weapon not being animated (mount0 itself doesnt move with the aim - in the regular player its the arm movements that move the weapon). So we didnt have a moving weapon. We tries to see if we could get mount0 to mount on the eye node, but never found the place to hack that.
* having to code a 3rd person crosshair. The resource available on GG made the engine lock up. I never found the bug, so I had to code my own 3rd person crosshair resource. That ate unplanned 2-3 hours. In the end it never worked correctly in multiplayer due to the object being created as a server side object. If anyone can give hints on how to do client side objects please post here. Thanks
Conclusion
Its was great fun!
We hade a game, learned a lot and want to do it again sometime. Maybe even using this method to develop on our own game in intensive bursts - either live together or over the net. 3-4 hours at a time where you feel that you work towards a common goal.
I think (we need to talk this over in the team - this is purely my own ranting) that we want to finish up the game to a playable demo with a fully implemented level 1. Maybe even showcase it at IGC.
I can only recommend everyone to try this. But do something very simple!!!
Screenshots from the game
Ahhh - finally some screenshots. These are not in order of development, but just to show the features we ended up with.
Main screen with the title of the game and our player model.
Shot of the final game before closing down the event. 2 players here with the 3d crosshair. Turret is shooting a green pea at me from the ant hill.
Same area in an overhead view. The red area on top of the mountain is the scoring area. Stand there and you score points every 3 seconds.
Showing off the crosshair. It moves with the aim, and is fully usable to game with. Needs some improvement though for long distance shots, as the crosshair gets very very small.
And finally a shot of the turrets protecting the hill
The first "Game in 2 Days" event is over, and we survived through the perrils and ecstacy of the work.
Short brief: we met 3 guys at my house for a 2 day session to get away from our usual game, and do a GID event. We wanted to have 2 days now that we physically sat down together.
Our rules: talk about and make an overall game design (1-2 pages) to set the major goals prior to meeting up. Then spend 2 days on creating the game.
Goal: primary goal is to learn and to have fun. We decided to go with gameplay a little more advanced, and where we needed to develop some engine extensions.
Game: a vehicle shooter, where the goal is to capture victory points by camping certain areas on the map. We wanted to give it "an indie twist", so we wanted to create the vehicles as bugs with weapons strapped on. The vicotry point locations are "ant hills" protected by vicious army ants guarding the areas with turrets. So the player has to race to the areas, fight the ant turrets as well as the enemy bugs. The game is played in 3rd person, and the world should be set on a lawn (with trees = grass etc)
The team:
Niels - 3d modelling
Morten - GUI and level editing, 2D art, basic scripting
Thomas - engine coding, scripting
What went well
We actually made a game in 2 days. Simple, and not containing all the nifty ideas we had, but its playable and fun.
We learned A LOT. Our "real" game is a action adventure without weapons, so we had to learn the Torque weapon system, mounting them on the player, projectiles etc. We never played with the LOD stuff in Max, so we had to learn that too. In 3rd person there is no crosshair, so we had to create a "3rd person 3d crosshair", the level designer had not really worked much with Torque yet so he ahd to learn the various mission/terrain editors as well as GUI editing.
Learning the dynamics of a GID event is fun too, because in retrospect its all about not making mistakes and knowing how to do things right the first time, where to look for info, resources and script code you made for other games and thus saving time. (This will be repeated in the "what went wrong"). I think we did quite well getting around the issues, and knowing when to drop it and attack a problem from a different angle.
One thing that worked out great was the division of work. Almost no overlap in work areas make for effective development.
Usage of subversion. Not having to send 20 MB files over emails because you changed 1 line of code. The ability to have it merge your changes etc. That saved us lots of grief.
Planning up front what to do. That saved us a lot of time, not having to brainstorm and waste development time on discussing ideas. Not that the plan survived more than a few hours, but without it we would have been lost.
Its also great fun sitting together and do team bonding, instead of doing it over Skype/Messenger!
What went wrong
First of all trying to get physically together takes time. Everyone has their own computer, it has to be set up and get onto the network. I run a pure wireless network at home......and none of the other 2 do that, duh.....and so we wasted quite some time trying to get Niels' PC running through a laptop with a wireless card. 3-4 man hours spend on that and we had to give up, sharing content on CDR instead. Better preparation would have had us running faster.
Having unrealistic planning. We tried to limit the game, but its clear that "development takes time". More time than expected and planned for. So our "evolutionary" approach to game dev with constantly playing the game, figuring out what is fun and what is not, fixing that and then go back to playing again - unrealistic. We were so busy creating "must have" code and art that we almost didnt get to play the game.
We wasted several hours doing things like:
* trying to make our own turrets. We followed Paul Dana's guide, but never got it working. It was written for Milkshape we think, and we worked in Max. The turrets got in fine after a while, but the nodes were turning in the wrong directions - shooting up in the air. Ended up using Paul's turret dts.
* messing around with the player model. We created the bug as a player model and first tried to keep is as a biped rig, attaching mount0 on the head of the biped for the turret. But the model went bananas in the engine with exploded polys all over the place. So we made it as a simple mesh instead with some simple animations for the walking etc.
* trying to get the turret mounted on the player. We spend quite some time getting the turret onto the player, and ended up having to use a standard weapon with mount0 on the back of the player. This having the problem of the weapon not being animated (mount0 itself doesnt move with the aim - in the regular player its the arm movements that move the weapon). So we didnt have a moving weapon. We tries to see if we could get mount0 to mount on the eye node, but never found the place to hack that.
* having to code a 3rd person crosshair. The resource available on GG made the engine lock up. I never found the bug, so I had to code my own 3rd person crosshair resource. That ate unplanned 2-3 hours. In the end it never worked correctly in multiplayer due to the object being created as a server side object. If anyone can give hints on how to do client side objects please post here. Thanks
Conclusion
Its was great fun!
We hade a game, learned a lot and want to do it again sometime. Maybe even using this method to develop on our own game in intensive bursts - either live together or over the net. 3-4 hours at a time where you feel that you work towards a common goal.
I think (we need to talk this over in the team - this is purely my own ranting) that we want to finish up the game to a playable demo with a fully implemented level 1. Maybe even showcase it at IGC.
I can only recommend everyone to try this. But do something very simple!!!
Screenshots from the game
Ahhh - finally some screenshots. These are not in order of development, but just to show the features we ended up with.
Main screen with the title of the game and our player model.
Shot of the final game before closing down the event. 2 players here with the 3d crosshair. Turret is shooting a green pea at me from the ant hill.
Same area in an overhead view. The red area on top of the mountain is the scoring area. Stand there and you score points every 3 seconds.
Showing off the crosshair. It moves with the aim, and is fully usable to game with. Needs some improvement though for long distance shots, as the crosshair gets very very small.
And finally a shot of the turrets protecting the hill
#2
Welcome to the very "select" club of madmen that are gidders ;)
Won't get you into the VIP lounges in airports, or do anything for your sex appeal, but hey, you know, and we know 8p
08/11/2004 (7:01 am)
looks like a nice gi2d effort :)Welcome to the very "select" club of madmen that are gidders ;)
Won't get you into the VIP lounges in airports, or do anything for your sex appeal, but hey, you know, and we know 8p
#3
Plus, thanks for the information. I think my GID will be in something other then Torque. I still don't feel comfortable with the engine enough to be able to crank out something like what you did, even in two days. :)
08/11/2004 (8:31 am)
Despite the problems you had, it looks like a fairly well polished game there! I'd say, definately go ahead and work out the bugs, and put it up for download. I'd certainly give it a shot. :)Plus, thanks for the information. I think my GID will be in something other then Torque. I still don't feel comfortable with the engine enough to be able to crank out something like what you did, even in two days. :)
Associate Anthony Rosenbaum