Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[LarkinVB]

What are you working on now LarkinVB? Is Vicious Byte still active?

Vicious Byte dispersed. We all did the game in our spare time when Matrixgames asked to release it. That was a lot of work. Nowadays I just do some modding (AI for Dawn of War or some Disciples 3 mods) and help my friend, who coded the factory, in his private game project - Zombie N4cers.

I don't think I will release the TOS source code as it is ugly, uncommented - just a mess.

[DarkAnt]

Haha, all completed project code is a complete mess. If you want to finish projects then there will be hacks It actually took me a while to learn this as I always wanted everything to be perfect and consequently would never progress very far.
Please don't be embarrassed by your code! It would be helpful to me at the very least.

I searched for Zombie N4cers, but google didn't come up with anything.

Happy coding and thanks again for the fun game.

[Muz]

Even without a source code, something like the design document would be really cool. Or at least the formulas. It's very rare to see a game of as much detail as ToS, a lot of people would be happy to learn something from it.

Anyway, posting a LP here, updating it in a few days once I get back from Canberra:
http://www.bay12forums.com/smf/index.php?topic=70645.0


Oh, and I realized that for some simple, destructive fun and getting the hand of game mechanics, duels are awesome. I just make a duel team with no long term planning, a single specialist pilot of each type. There's usually enough salvage and payment from battles that I can just sell off titans and buy new ones to test out which titans are the most fun.

[mrchinchin25]

Sounds like this game scratches two of my boxes - managing a team that can die and gain experience, and equipping them to fight things.

Oh and mechs... lovely lovely mechs...

[DarkAnt]

The manual is fairly comprehensive. It certainly explains the modifiers that I've had questions about. I can also see how this would generate incredibly messy code if you tried to apply the correct modifiers all at once for a specific action. If you're still with us LarkinVB, how did you go about handling all of the checks in the code?

[LarkinVB]

I can also see how this would generate incredibly messy code if you tried to apply the correct modifiers all at once for a specific action. If you're still with us LarkinVB, how did you go about handling all of the checks in the code?

I apply the correct mofiiers for a specific action. This is not that messy.

[DarkAnt]

The corner cases looked to me like they could be troublesome, but when I look at the docs there's typically only one or two special cases that need to be handled with everything else using the same rule. So as long as the special cases (like tesla bolts having a 12% chance to cause internal damage) are kept to a minimum then it keeps things fairly simple (simple in ToS terms at least   )

What was the most challenging part (code wise) in writing ToS?

[LarkinVB]

What was the most challenging part (code wise) in writing ToS?

Very challenging was the network code to keep the game in sync and only display what the active jock should see/read/hear.
Also very challeging was to code the AI as it has to take into account so many variables and it was written on the fly without a design concept.
Challeging was to add missions/campaigns afterwards and to interact with the editor which was written by someone else located in south arabia.

Team management was difficult as we had members in germany, south arabia, USA and portugal with different amounts of spare time. Remember we were not doing this as a job ! Most stuff was coded by me and a friend located in the same city, otherwise it would have been very hard to complete the project.

[mainiac]

I'd just like to drop in and say thanks.  I've only played around with this game a little so far, but I see that you guys have done some impressive work.

[Zangi]

Wow, sounds interesting.  (Downloading...)

[DarkAnt]

Last question I swear. How did you implement the line of sight? I was thinking about it today and my first thought was that you could use Dijkstra's Shortest Path algorithm using the hex LoS modifiers (relative tile height, titan height, trees, smoke, distance from titan) as edge costs. I think the only thing you'd have to modify is how the graph representing the hexagonal grid connections is constructed. The edges would have to be directional with each directed edge  directed away from the titan so you couldn't do silly things like look around a mountain (actually a bit more restrictive than this).

I suppose the LoS graph would have to be as large as the maximum number number of tiles a titan could possibly see. When I was playing on a random 22x22 map a titan in the center of the map could see end to end I think. This would make the graph pretty big. To compute it for every titan every time any of the LoS graph hexes are modified seems somewhat expensive: O(n), 484 nodes. I guess you can maintain the LoS graph by just propagating out changes to the graph (moving your titan would be a bit challenging). I don't know, maybe I'm way off here. What did you do?

[LarkinVB]

Dijkstra's Shortest Path ? Never heard of. 

I tackled the LOS problem early on (1998) but remember that I used a very easy formula from a hex based boardgame.

Calculate source height, destination height and scan the path from source to destination for blocking hexes (maxheight).

return FALSE in code below is no LOS for that hex. i is iterating from source to destination, dist is the distance.

Code: [Select]

  if ( maxheight >= sheight && maxheight >= dheight )
{
  free(lhex);
  return FALSE;
}
  else if ( maxheight < sheight && maxheight > dheight &&
i >= dist - i )
{
  a = sheight - maxheight;
  b = maxheight - dheight;
  if ( a <= b )
{
  free(lhex);
  return FALSE;
}
}
  else if ( maxheight > sheight && maxheight < dheight &&
i <= dist - i )
{
  a = dheight - maxheight;
  b = maxheight - sheight;
  if ( a <= b )
{
  free(lhex);
  return FALSE;
}
}


[Zangi]

Apparently my designs suck... or my jocks suck.  One or the other... or both  Might have to do with fitting too many things on my recons.  Not enough weapons.
Getting whomped by rookie bots in 1v1 and 2v2.

[DarkAnt]

A sure fire way to win in the early recon matches is to go duelist spending all of your points on light/recon, close combat, electronic warfare and scanner. You might also want to put some points into leadership, business, and damage control (a seperate jock for each one) to cover team expenses.

[Akura]

And don't forget Close Combat. Seriously, Power Axes cause 10 damage(and unless you really suck, generally have really high to-hit), Chainsaws do 15(haven't checked the Vibro Claws yet). Unless you really scimp on everything else, like armor or something, the biggest thing a recon Titan can mount is a Small Meson Gun, which do only 4, at much less accuracy and more space(for ammo). Now, add CC weapons to something that's faster than any other mech out there, and smaller(does Titan size matter on hit-roll calculations?), and you've got yourself a tiny little threat.