Pimpage: Tank Universal
Community Forums/Showcase/Pimpage: Tank Universal
| ||
| Hi all - another round of pimpage Work continues on a single level demo of Tank Universal. Have added new units - laser tanks, beam tanks, mobile bomb units and a battleship that circles overhead and fires down. Becoming quite graphically intensive - looks like a 1ghz'll be the minimum and that's what's really holding up the demo release - need to upgrade so can fully test all the fx features. Madjack New Zealand     |
| ||
| rahh lovely tron-like |
| ||
| I just bought the 20th anniversary Tron-DVD... GREAT stuff in it and a great movie! And TU is looking mighty nice too. Very Retro-ish. |
| ||
| Mustang I'd buy the DVD as well but I've only seen a zone 1 so far and my player takes zone4.. save us from dumbass copy protection schemes.. I'm writing TU as the kind of tank game I'd like to play - strategic depth but quick to pick up and play. As I don't want to have to create megabytes of graphics, I've set it in an abstract world where I have the greatest freedom over the FX and models. Some people 'just happen' to think looks like Tron ;-). It might be difficult to see, but I've found that wrapping a flag 64 (spherical environment) texture onto my previously untextured walls, has helped give them extra definition. |
| ||
| Grr, grr, DVD zoning; don't get me started... looks fabulous, as I've said before. 1ghz minimum is a little worrying though - it doesn't look particularly intensive... |
| ||
| Anthony I'm finding it a little difficult to gauge a final spec as my system is pretty low spec to begin with(Matrox G400, p3660mhz), and I think the vid card is a big bottleneck. I have a long view distance and to avoid polygon tearing, I have to set screen depth to 32bit and the card slows down for that. I wouldn't be surprised with a card upgrade I'd get a big boost in speed and could lower my minimum spec accordingly. The other thing the screens don't really show is that the level contains about 80 computer controlled actors (tankbases/turrets + bullets and explosions) and there's a lot of AI/pathfinding going on in the background. I've been pretty stringent in not 'cheating' with the AI and that adds to the loads.. |
| ||
| I'd buy the DVD as well but I've only seen a zone 1 so far and my player takes zone4.. save us from dumbass copy protection schemes.. Thank god for players that can be modded region-free with it's own remote (three digit code and wham - no worries!) :) [Finland is region 2] My player also eats NTSC as well as PAL and plays MP3s and shows .jpgs from CDRs - not bad for a DVD-player that cost only 199€ and even looks stylish! |
| ||
| Looks great. A little like a PlayStation game I used to play. Except this looks like it could be fun. :o> |
| ||
| Holy shiznit....thats some incredible scenes. How long until a demo? |
| ||
| Looks really great MadJack, can't wait to see a demo running. |
| ||
| That looks very good and playable (already). Mustang mentioned it looks "retro-ish", I agree; it looks like a cross between Encounter, Beamrider and Fort Apocalypse! :) I've been pretty stringent in not 'cheating' with the AI I think that as long as the player does not notice the AI is cheating, it's ok. Of course, if the enemy tanks just know where you are without seeing you and actually *act* on this knowledge, that would be bad cheating. Often much (quadratic) time is lost with every object checking the location of every other object, but again often it is enough to have the enemies check only the location of the human player (however, to have team-play they would have to do more). And a shortcut I used in some experiments (otherwise known as unfinished games) is to check which objects could be spotted by the player (potentially), and to mark these objects right away as being able to spot the player (if they were to look in a certain direction). Well this is probably not relevant to your game, the AI you are talking about may already be a bit more complex! edit; indeed, your game looks like it has teams fighting each other, so just taking care of "the" human player will not work, in that case... |
| ||
| Foppy Yep, and there can be up to 7 teams - 2xred,2xblue,2xgreen, 1xneutral, fighting it out over 4sq km. I don't want the player to be the 'centre of the world' in that there can be events going on that are completely separate to the player. However, there is a fair amount of cross checking going on - in fact too much and come to think of it, a not inconsiderable amount of duplication. Need to look at perhaps keeping some sort of visibility check table. The other possibility is doing some terrain height preprocessing to create a height table (which includes walls), and use that for coarse entity visibility checking (I don't have any arches or covered areas in the game). Finally, can someone give me an idea of upgrading options. I have a P3 660mhz, Matrox G400 32mb, 128mb mem, running windows 95 (yes, W95). If I were to keep the CPU but swap out the vid card for say, a geforce 3 and upgrade the op to w98, what kind of an fps increase might I expect? Small, big, medium? |
| ||
| what kind of an fps increase might I expect? Small, big, medium? Huge? :) Seriously, Matrox is no good. I had one years ago, although can't remeber what G-model is was... but it couldn't do even multitexturing right! It had good picture quality though. Even measly GF1 has T&L which means that GPU handles transforming and lightning calculations freeing CPU to do other things. This is a BIG advantage, even for Blitz3D apps which is only DX7. And upgrade to W2000PRO if you want to upgrade the OS - but anything is better than W95! |
| ||
| Is it going to be multiplayer online? This game looks like it could be a lot of fun with a lot of players! |
| ||
| RexRhino The short answer is no - I'm finding it enough work just to implement my one player goals! And once the one level demo is out, although I'm expecting that'll be final code for the bulk of the game, I can then begin developing the 14+ levels and I'm tentatively budgeting another year for that.. Neuro/Mustang I really need to upgrade before I can release. So from what you've written (Mustang), it sounds like a vid card and OS upgrade would be a good temp measure before a full machine replace. |
| ||
| The short answer is no - I'm finding it enough work just to implement my one player goals! And once the one level demo is out, although I'm expecting that'll be final code for the bulk of the game, I can then begin developing the 14+ levels and I'm tentatively budgeting another year for that.. Oh, shame... maybe some time in the future? Neuro/Mustang I really need to upgrade before I can release. So from what you've written (Mustang), it sounds like a vid card and OS upgrade would be a good temp measure before a full machine replace. Oh yeah. A half-decent vidcard should make all the difference (probably much more than a faster processor in your case) - my game runs fine on a 400mhz processor (or even less?) but NOT with a crappy gfx card. When you do upgrade your whole system though, you should keep the old box - very handy to check downwards-compatibility on! |
| ||
| Looks great!!! Can't wait to play a demo of this! |
| ||
| FlameDuck - Assault Rigs was ACE! But this looks far nicer. :) |
| ||
| Looks brilliant, can't wait to play a demo :) I've always liked Tron-Like gfx, the VR missions in metal gear solid reminded me of that, so did assault rigs on PSX |
| ||
| Thanks for the comments guys - on a long term project like this a bit o' encouragement goes a long way. |
| ||
| What about multiplayer features? =) |
| ||
| The coloring and look of objects also remind me of that Terminator game, Terminator Future Shock. In that game the enemies really gave the impression of being AI controlled enemies (well obviously, they were, but you see what I mean). Giant grey and red killing machines trying to find a way to walk round a wall behind which our tiny human hero was hiding. I can imagine situations in your game were your CPU controlled team members have all been destroyed and you are the only one left, and enemy scouting planes are circling overhead, and you try to shoot them down lest they give away your location. Or is that not the way it works! :) |
| ||
| Foppy Heh, well I have programmed a small scout with a spotlight that can follow the player around, reporting his position to the enemy. The issue with aerial targets in a tank game is that they can become frustrating (and I don't want to introduce 'fire and forget' kinds of weapons). In terms of biggish enemies, I'm thinking of virus type enemies that on contact with tanks, suck shield energy and grow larger, or perhaps can combine to create larger enemies. Currently extending the one level demo to a two level demo while I get the money together to upgrade... |
   |