• Advertisement


  • Content count

  • Joined

  • Last visited

Community Reputation

242 Neutral

About Madster

  • Rank
  1. Very Complex Combat

    Quote:Original post by AngleWyrm I like the idea of defensive moves. Martial arts have them. But I'm not so sure about making a player choose between dealing damage and damage reduction, because the second option is generally just postponing an inevitable death. That may have to do with using hitpoints. You'd end up always attacking and never defending. I'd like to see something I saw in DOA polished: counters. in DOA you just have a counter button (button combo actually) which counters just about anything. I'd rather have counter moves, with different reception zones. For example, with one counter move you'd counter attacks from above, with another you'd counter low moves (leg sweeps and such), etc etc. This makes defensive moves actually usable. ALSO, on throws: I liked Samurai Spirits throws. Not just a simple block-breaking 1-hit move that drives the opponent to the ground, but a shove that throws them off-balance and opens up a combo opportunity, so that combat flow is not broken. Question. In 5-way system, as stated: Quote:A beats B and C B beats C and D C beats D and E D beats E and A E beats A and B player1 uses: A player2 uses: B Who wins? :s About that long off-topic on optimal strategy: this is real-time fighting, so keep in mind that all the strategy must be considered in split-second. In this case, surprise is a big factor. It reminded me of watching a korean pro Starcraft finals game, where one of the players decided to do a zergling rush, and won. In case you don't know, zerg rushing was so common that most people knew how to defend aganist it... but this pro-player wasn't expecting such a cheap shot, so he fell for it. Skill, speed, psychology. Real-time. Edit: idea-mangling typos corrected
  2. Hmm I'm doing my undergrad thesis on displacement mapping. Thanks for the tip! I'll be looking for ways to make my proposal "free" then, using tesselation chip resources (at least in theory, since there's no PC API for it yet) Can you say which 3 LOD approaches were tried and discarded?
  3. On second reading.... Are you looking for obstacle avoidance? if so, Sonar's a much better idea, as is interferometry. Subs use it. Sonar would drive the surronding fauna nuts tho... bats specially! With it, you should get a map of angle/distance to obstacle all around your emitter. Of course, this is only a cone, as the plane itself would block a 360º emitter, but should be enough to correlate with a camera image. If obstacle avoidance is not the issue, go for gps/gyro as those won't show you trees/buildings with proper accuracy, since satellite mapping is designed for large scale terrain mostly. you *could* combine both, too. you'd get large-scale navigation aids and local obstacles. woop!
  4. Cloud Issues, suggestions?

    I see only small sprites on a big volume. If I remember correctly, that paper showed examples where some big sprites would suggest the main shape of the cloud. Those were really blurred because of the stretching, but the small ones added textures. Try mixing sizes. About lighting, I have no idea.
  5. I'm guessing your plane won't have a huge autonomy range. You could download the datasets for the zone you'll be flying over and use the GPS for the position. Then render a z-buffered textured mesh (as in: showing edges but opaque faces) over the video. Interpolate GPS positions in time if using a very low-res GPS. This shouldn't be a problem unless you fly real low or close to the floor, buildings or vegetation, as scenery won't change too fast depending on position since it'll be really far away. You'll need a mounted gyroscope to align the 3D-view with the camera view, this *is* really important. Otherwise, you could go top-down or render separately from the camera image. I wouldn't recommend going the Image Recognition way, as it would be too much work and not give good enough results.
  6. I've been trying to find out what current (and recent) games use for rendering their terrain, but info is minimal (just marketing). Does anyone have any examples? I'm fearing the most high-tech mainstream games have gone is chunked LOD.
  7. Render to VBO on 2008 cards

    the in-band part lost me, would I need to do something special for this to be in-band? I already looked up the functions needed and I got the idea, except for that. Thanks! EDIT: Quote:How far back does PBO support go for hardware? Not far enough apparently :-/ My target hardware changed to OpenGL ES 1.1, and that doesn't seem to have PBOs It does have FBO if it has the OpenGL ES 1.1 Extension pack (which I'm not quite sure if it's installed by the client or the driver). anyway, seems I could do the copy anyway with readpixels, but I'm not sure how fast it would be.
  8. Render to VBO on 2008 cards

    Thanks, I'll give the PBO technique a try. I remember reading it was very slow, but when I read this it was about modifying texture data trough the CPU, so the transfer was needed.
  9. Have things changed since 2006? http://www.gamedev.net/community/forums/topic.asp?topic_id=434618 I'd like to try an algorithm that requires vertex shader to look up textures, but when searching I keep bumping into info from 2005-2006 saying that ATI requires using render-to-vertex-array extension in a pixel shader instead, while NVidia uses a very picky texture sampler in vertex shaders. Has any of this changed? is there now a standard (and fast) method to calculate new vertex positions using texture data? I'm thinking about terrain rendering/displacement mapping, btw. Edit: I'm running an ATI 9600XT, Should I look into an upgrade? [Edited by - Madster on August 25, 2008 9:30:27 PM]
  10. Thanks, will carry on with Irrlicht and GL 2.x shaders
  11. The Angry Vendor Idea

    Adding on to previously mentioned concepts: -Instead of having each NPC have a "personality" (valuation for goods), make roles and have NPCs fit to those roles, so the player has an easier time guessing who is who while meeting bunches of them. -Trade items usable by yourself (had this in a game called "transcendence"), so you'll be on the lookout also for specific or goods which boost your whatever. -Giving rare items as gifts to obtain favors. This post is small. It ends here.
  12. Flexible map making

    Quote:Original post by Kest Picture a complex 3D character walking around the corner, then turning into an Atari blip. Both can achieve similar cause/effect, but the simplified version is much easier on the CPU. This will allow me to have actual background combat, where your allies are engaging battle in a building across the city from you. But the whole system will also be a pain in the ass to implement. Every character action or feature will need to consider a dumbed down version. You mean LOD? you'd still have to do the simulation though. It's standard procedure, done in many RTS. Impostors, billboards, etc... all just LOD in some way. Be careful to separate your simulation from your rendering.
  13. Quote:Original post by Wavinator Case in point: A game with variable time scales runs into the problem of what others are supposed to do while they wait for you to catch up. Someone above said variable timescales are out of the question for multiplayer. I haven't ever seen it work. Haven't heard anything it sounded like it could work. Haven't imagined an approach where it would work. No time freezing, reversals, not even slowdown. At the very least, it would be highly annoying. Edit: What about classic Myst? Can that be multiplayer-ized?
  14. Very Complex Combat

    What you've outlined has pretty much been done or is unfeasible (in real-time fighing games): Quote: More than just HP/MP 1 Life 2 Magic / Mana / Spirit Force 3 Strength / Power / Physical force 4 Stamina if moves drain both magic and strength, players will have to keep looking up (or down) at two meters, which have the same effect of a single meter. Just have one. Every technique spends stamina, and everything is a technique except moving, so stamina will constantly go down until you get stunned. The thing is, when both players have low stamina, the fight is going to get really boring. Unless... one of the players turtles to conserve stamina.... which also makes for a boring fight. Taking out stamina and merging both 2 and 3, you get the standard HP/MP. Quote: Magical elements at any given time you can only have a primary, secondary and tertiary specialization With elements and slots to fit them in, each one granting moves... you'll end up with movesets for each character per each combination of elements, much like you can see in the old Samurai Shodown 3 & 4 (Bust/Slash modes) Quote: Timing Commit times and delays. Look at a game like Street Fighter, you do any technique, it happens instantly and then you keep going This is the core of most fighters. Delay before the attack and delay after the attack. The thing is, if you wanna do a great fighter, analyse the classics and the not so classic, and find out why they worked and why they didn't.
  15. Hello. I've been coding small games on and off for quite a while, but so far always someone else has handled the graphics side for me. Now I have to do it on my own and I'm looking for the proper place to start. I believe the best way to learn something is to jump right in it, so I'm doing my undergrad thesis into something related to geometry subdivision (like ROAM and such) So, I'll probably end up fiddling extensively either with transformations, pixel shaders or vertex shaders, depending on which angle of the subdivision process I wind up attacking (haven't quite decided it yet, decision due friday). The thing is I know very little about OpenGL, just tried some inmediate mode once, that's all. Should I aim for 2.x? 3.0? (been reading the debate) Maybe I should go for DX instead? Perhaps try an engine and just work on what I should be working on (shader/matrix math) and not waste time learning what I'll end up offloading to an engine anyway? I've been looking at Irrlicht, since it's lightweight, free and quick to get started. Thanks! (for the record, I *have* coded simple 2D games before)
  • Advertisement