I played Total Annihilation back in the day. One thing that I remember standing out about it was the ability to automate production and tactics to a significant degree. I had factories pumping out drone ships with preprogrammed flight paths, which would engage any hostiles they encounter then resume flights. Then so long as I kept factories supplied with metals, I ensured air superiority and was able to focus on moving land forces, or other things.
Whereas most RTS devolve into micromanagement matches, TA seemed to be more about high-level strategy since automation really let you focus on theatre management.
> pumping out drone ships with preprogrammed flight paths, which would engage any hostiles they encounter then resume flights.
I love the idea of being able to program a collection of routines which you can change the parameters of, or swap in and out, depending on how the battle is progressing.
There's probably a danger, though, that by making the programming language/runtime too capable, the game ends up being "solved" or at least ends up only leaving boring work for the human to do.
The capability was very limited. Basically you could program a looping series of commands for newly manufactured units. "Go to this waypoint #1, then waypoint #2, then this final waypoint, then go back to the beginning and start again." And IIRC you toggle a switch to select whether the units should strictly obey the commands or stop to engage in battle along the way.
Typically this is used to set a destination for new units, so they assemble automatically. But because you could set the commands to loop, and because you could authorize them to auto-engage (using the built-in AI to track and follow targets they run into), you could hack it to have the new units do routine patrols without further player action.
This definitely doesn't "solve" gameplay. It just gives the player access to the game's built-in AI for their own purposes. Of course it's much more useful playing against the computer than other players.
With Spring games, and their open source AI and widget system, if you wanted to you could basically have an assistant AI playing for you... but the hard part (and most likely why after 16 years it still isn't the dominant way to play and/or didn't result in high-profile scandals in tournaments) is probably to do it in a way that you don't keep hindering each other !
> There's probably a danger, though, that by making the programming language/runtime too capable, the game ends up being "solved" or at least ends up only leaving boring work for the human to do.
This would imply that war and generalship in reality are already "solved" problems. It's not a real concern.
For simple enough models of the battlefield, war and generalship probably are solved problems, but having said that, even the game of chess has not been formally solved, and it's certainly possible that an RTS game could have a greater strategic depth than chess.
Whereas most RTS devolve into micromanagement matches, TA seemed to be more about high-level strategy since automation really let you focus on theatre management.