Code center > Suggestions
Implementing Sound
rsucoop:
Ok, how about the sound waves haev to hit their bodies, and that lets them decifer the information in that wave; which could be a turn command, or a refvar, or a 'hey im in heat' system. And I say dont allow writing to memeory through sound, leave it to the hearing bot.
EricL:
The only thing we have today that has a propagation delay are shots. We could do something here leveraging shots...
Imagine a new "sound" shot type. It propagates out from the shooter in a half circle at regular shot speed, it's range is limited and controllable via .shootval. One diffrence between it and other shot types is that the shooter's speed does not impact the shot's speed. The speed of sound is constant. Another is it can't be seen (other shots types will be visible soon) or maybe it can. Any bot the wave crosses before it dissipates completely is impacted.
What it does when it impacts is a matter of discussion. Perhaps there are a handful of sound in/out pairs. The shooter sets them before firing the sound, the shot impact sets them in the impacted bot.
What happens when a bot is impacted by multple sound shots in a single cycle is TBD. Perhaps each shot has a shooter-settable power that dissipates as a function of range. The "loudest" sound impacted a bot wins...
The .hit sysvars already provide a means for impacted bots to directionalize shots.
If shapes are specified to reflect shots, then they reflect sound shots, allowing for echo-location.
I'd have to impliment some new code to do circle-circle intersection for this shot type for impact detection and there will be a perfromance impact if this becomes heaviliy used, but I would say this option is doable in the mid term.
rsucoop:
--- Quote from: EricL ---The only thing we have today that has a propagation delay are shots. We could do something here leveraging shots...
Imagine a new "sound" shot type. It propagates out from the shooter in a half circle at regular shot speed, it's range is limited and controllable via .shootval. One diffrence between it and other shot types is that the shooter's speed does not impact the shot's speed. The speed of sound is constant. Another is it can't be seen (other shots types will be visible soon) or maybe it can. Any bot the wave crosses before it dissipates completely is impacted.
What it does when it impacts is a matter of discussion. Perhaps there are a handful of sound in/out pairs. The shooter sets them before firing the sound, the shot impact sets them in the impacted bot.
What happens when a bot is impacted by multple sound shots in a single cycle is TBD. Perhaps each shot has a shooter-settable power that dissipates as a function of range. The "loudest" sound impacted a bot wins...
The .hit sysvars already provide a means for impacted bots to directionalize shots.
If shapes are specified to reflect shots, then they reflect sound shots, allowing for echo-location.
I'd have to impliment some new code to do circle-circle intersection for this shot type for impact detection and there will be a perfromance impact if this becomes heaviliy used, but I would say this option is doable in the mid term.
--- End quote ---
What if we did this for multiple shots? We create a sound buffer, which collects all information about a shot right after it impacts the bot, the bots ear goes 1+(1 for each other sound). THen the bot could randomly, or intelligently, sellect a sound its heering and try to interprut it. The only problem would be speed of simulation from that, it would give the bot the ability to hear only commands from a friend, or onyl calls from a mate, or something along that; a sort of ability to decifer its environment. I think we did this with the tin/tout system, only we didnt need 36000 vars
rsucoop:
Could this be used in 2.4x+?
EricL:
--- Quote from: rsucoop ---Could this be used in 2.4x+?
--- End quote ---
If you are asking whether this coudl be implimented on the 2.4x code base, the answer is yes. All it takes is time and effort...
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version