Jump to content
  • Advertisement

00Kevin

Member
  • Content count

    204
  • Joined

  • Last visited

Community Reputation

381 Neutral

1 Follower

About 00Kevin

  • Rank
    Member

Personal Information

Social

  • Twitter
    @ProjectXSYS

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Did you have any game in particular in mind that did a good job at that? I'd really like to review it. There is no redoing, but I could add it. My only concern with allowing that players might abuse it to reveal hidden creatures, traps, hazards, and unexplored areas of the map. My original design included an action queue of sorts. In this case the action queue could be executed at any time, and if actions are queued in a particular order, combos and special bonuses would fire. For example, if you decided to jump and attack the two creatures beside you the "Sweeping Leap" combo would be activated. In this way, the player would focus on building combos for extra effects and damage. Players could discover these as they play. Of course, there will be mechanical reasons to make individual weapon attacks that are not queued, for example it might be easier to hit with a non-queued action. The problem is that I'm not sure players would appreciate this. Many might simply want a single button to activate the special maneuver and not build them during their turn. I do know that I plan on creating a magic system that works on a queue. In this case, you can add in special regents other trinkets to your spells as they are being cast. For example, if you want to add a fire effect , the player can add in sulfur or coal to the spell casting queue.
  2. So what I've decided is to embed the confirmation into the very nature of the menu system itself. What I mean by this, is that you can hover over a unit to display all available options, click on the unit to activate the selection of those options and inspect them in more detail, and then finally select the option. In other words, all actions always have two clicks.
  3. oh right, I totally forgot about that weapon wheel. The RPG game I'm creating allows the player to create robot / cyborg characters only. I made this choice because I didn't want to create a ton of clothing options. With robots, equipment slots are reduced to hand held items and a few augmentation items. It's also a multi-character party based game, allowing the player to create up to 6 characters. So the total number of items a player must deal with is distributed across many characters, allowing for more realism. With that said, I do like the idea of looting corpses and chests. I just have to think about how that would work with such a system.
  4. Yes, I'm actually considering the U7 approach. It would certainly reduce the level of coding required. I also really like a feature of Utlima Underworld's inventory, which allowed you to have containers inside containers. Inventory Tetris is something I'd like to avoid, but I can't deny the fact that some players love organizing their inventory.
  5. If you had to make a game that didn't use icons for inventory (like WoW , Diablo, etc) what solutions would you propose? My current game has an inventory system that is icon and paper-doll based, very similar to WoW, but I'm considering innovative alternatives. For one, I like the idea of allowing the player to manipulate actual 3d objects. My thought is that the player could drag and drop the object onto the character directly. The game might spin the character/camera around, zoom into an equipment location, and/ or highlight valid equipment locations for the selected object Of course, this means that characters would have actual backpack objects on them that would open up to reveal their contents (3d objects). It certainly nods to realism, but that might be fine if it's done correctly. Any thoughts?
  6. As for my previous solution, it fails when the correct action is to maintain maximum range. Maybe players won't like that, but I'll leave that to them to decide.
  7. That is very true, but ranking actions against the creatures intelligence score is something I've already written. My hope is that play testing will end up curtailing my stupid . Of course, I'm not trying to make the AI overwhelming, I just don't want it to perform stupid actions. For example, if the player is at the back of a building with a single window, It makes far more sense if the AI doesn't walk down the stairs, open the front door, and run around the building to attack. I'd rather see the AI simply move towards the window and start shooting.
  8. So the solution I've settled on for now is to let my path-finding system do all the work. You simply find a path to the target. Then you scan that path for points within range and loop until you find the one with a line of effect via a ray cast. The problem is that this solution ignores things like windows which might be closer. One solution to this is to create a second grid with a much smaller granularity (0.25 size squares) and use it for the initial path-finding. This would allow the path-finding system to pass through windows and small openings.
  9. That's interesting I'll have to give this some thought. To be honest, I'm still trying to visualize this solution. As an alternative to ray casting I wrote a Bresenham 3D line drawing algorithm that returns a list of tiles. In fact, my tilemap already has obstructions and occupied squares identified.
  10. I'm not sure that helps all that much considering the positions of units often change. Known positions are rarely useful. In addition the environment can often change as well. With that said, an initial raycast towards each opponent is useful. I'm considering compressing the bit array because quite frankly it's the ultimate solution most of the time. I'll need a compression method that still allows for the data to be read without un-compressing the entire file.
  11. One other complication is that the correct way to test for line of effect and cover is to use the corners of each square and not the center point. This makes this problem even more cpu intensive when raycasting.
  12. Thanks for your help. I will definitely give these constraints a try. One complication I didn't mention is that in my case the tile grid is 3d. Which means if I have a small level that is (50x25x50) for example, I'm looking at 3,906,250,000 combinations (50x25x50x50x25x50) for a static scan. Yikes! that's a 3 GB file per level, assuming I store byte (bool) values and not bits to the file. The bi-directional property will reduce this, but at the moment I'm not sure about the proper algorithm to implement it. Iterating over a 6 dimensional array is easy, but it will certainly take forever and it's not optimal. I know there is a way to iterate over all the possible combinations taking bi-directional factor into consideration, but I haven't worked out the structure for such a loop yet. Ray casting will definitely need to be avoided if the the line was previously scanned from the opposite direction. Regardless, it is still a monster to manage in either case. I might have to re-evaluate the need for such a system and consider less optimal solutions for the AI. The initial system I wrote picked up all the tiles at a radius out from the target and performed pathfinding on each tile. The problem is that the longer the range the slower the system (more tiles need to be scanned) .And if you limit the tiles to those between the attacker and the target, some wall shapes (like U shaped walls) cause this technique to fail. Thanks
  13. I'm looking for solutions to a Line of Effect system I'm creating. Specifically, I'm looking for a real-time algorithm that helps an AI unit find the closest tile in range that has a line of effect to its target At the moment, I'm using raycasting (in Unity) to determine what tiles have lines of effect (no obstructions) to each other. This information is then used by the AI to locate an unobstructed targeting position within range. The end result is that every tile has a list of other tile locations it has a clear line of effect to. The problem is that it takes forever to scan and makes the map far too static for my needs. Anyone implement something similar? Thanks
  14. I guess it depends on how complex the combat system is. If it allows for burst attacks or multi-target attacks it's much more difficult. Highlighting all possible targets isn't easy, especially when many of them are off screen. In addition, a single action might work very differently against each target. Multi-target attacks like picking 1 to 3 enemies within an AoE is more difficult. Of course, it all depends on how much control you want your player to have. There is also the problem of hitting invisible units . For example, if you suspect there is an invisible / hidden creature in a particular square you might need to use an area attack or select the square to attack and not the creature. I personally don't like confirmation dialogs, but for some reason XCOM uses them extensively. I guess play testing revealed that it was necessary.
  15. I agree, confirmations can really get in the way. XCOM uses the right mouse button for some confirmations and it has' become the expected format for many modern TB games. I'm trying to design my game without confirmations or at the very least have it as an optional feature. IMO, three clicks is way too much. Anything that slows down game play in the TB game is to be avoided. Selecting the target, the action, and then a confirmation button is just too many clicks. At the moment, I'm looking at Field of Glory 2 for some ideas. I really like how all the action icons are displayed in context near the cursor. When the mouse is button is clicked the action icons pivot vertically and enlarge. At that point, you can hover over all the possible actions and evaluate the action in context with the selected target via the tool tips. Clicking on the icon again is like a confirmation. This is a good solution since it reduces the number clicks down to 2.
  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

Participate in the game development conversation and more when you create an account on GameDev.net!

Sign me up!