1999 command and conquer outperforms 2020 AOE2:DE?

How are those old school command and conquer games from the 90s still have better unit pathing than a game in 2020. You can select way higher amount of troops and compute mass numbers of units but DE, no lets limit it to sixty and if there are more than 1000 troops on the map it will crash. These command and conquer games were ages ago and are still out performing in certain areas. Age of empire need to step there game up. If you haven’t yet, try out tiberian sun. It’s a great game. (not as good as AOE2 of course).

1 Like

While I agree that pathing could be better, there are certain limitations when it comes to refurbishing old engines/games. If you want to flex on AoE2 with CC:TS pathfinding: there’s a few things to consider:

  • Pathfinding isn’t a standalone feature. The devs develop a pathfinding model that factors in their multiplayer/LAN model, the map design with its exceptions, unit hit boxes, etc. In turn, pathfinding models affects how the animations and combat are designed. A different game has different pathfinding needs, which is obvious from the way bridges work in CC. In AoE2, I believe dropping a bridge on a map is somewhat like painting terrain as ‘walkable’. In CC: TS, they had to write pathfinding exceptions for bridges. So there’s an obvious difference in the way they modeled the interactions between units and maps aka pathfinding.
  • The original model was catered to the hardware of the time (think CD-rom speeds, 56k dialup modems and millenial pentium processors with 1gb ram & 64-256mb GPUs or some such). Designing your game with that in mind makes it hard to adapt the core of the game later on. Hardware has changed a lot and despite developers having access to f.e. DirectX to circumvent all that, there’s still quite a bit that depends on it. Not sure how definitive (pun intended) or hardware-restricted those core models are for AoE2 CD/HD/DE (and how much they’ve changed between versions of the game) but to some extent, some models like the AoE2 pathfinding may not be properly adaptable to fit modern paradigms because of that.
  • Aside from the hardware and inter-connectedness of features, it may just be that they haven’t found a solution to the problem yet, as path-finding algorithms have to be written on a case by case basis and I can imagine they’re trying to optimize the current algo instead of throwing the old one out in favor of a better more modern one. Both approaches have constraints. Writing a new one could cause a lot more time spent bughunting and having to rewrite/redraw animations to fit the movement, among other thing. Writing exceptions or optimizing in other ways might be more straightforward, but then your pretty much limited by the foundational algo’s speed.

I hope that helps clear things a bit. Maybe the devs can give us more insight? Pwetty pwease?

Edit: suggested reading, https://www.academia.edu/24714838/Artificial_Intelligence_for_Games_2nd_edition_ (yeah, it’s big and 10 years old but still a very comprehensive and seminal work)

4 Likes

I remember this series of PC games.

Well, @GunTh0r somehow stole what I wanted to say. Adding another point to his post…you would actually appreciate game making more whenever you read technical aspect of the game building. You start to admire the end product. You should definitely read the PDF book on path-finding if you get a chance.

Anyways, you did made a fair/valid point. I’ve made several flash games, I used to work for Nitrome, and we generally implement path algorithms in engine. It all depends on how well the algorithm is implemented, and what kind of “obstacles”, “Terrain” and “Angle” algo sees. The latter part is a huge advantage for Command and conquer. In fact every game implemented with such design, had an flat isometric view. Whereas you see Age of Empires, and even the later command and conquer iterations, you see pathing issues. You can blame it on all on the technology and design.

1 Like