Venom Revamped

[rsucoop]

I have noticed a general lack in effectiveness of venom in many bots compared to naturally occuring venom-using bugs. One way to improve this realistic gap, is to simply add a ventimer (maybe even poisontimer), which could be set by the host bot for some cost and would last that many cycles in the infected bot. This would make it possible to spend some energy and really blind an enemy bot for so many cycles; I do believe there should be a randomly inherited limitation on the duration of the ventimer, as to maintain a balance between the sneaky and the fighters. Most importantly, this would allow for better multi-bot techniques, since a shot could last for so many cycles (call it x), x-1 many bots could know that the bot has been venomed or poisoned through ties. Tactics using two seperate bots at once to effectively blind and stop an enemy with a shot from each of them would be easier to implement, be it a delayed reaction or immediate. I also think for an effective defense (or warning) that the timer should override any out channel at random and use the timer apropriately associated with the memlocation.

[shvarz]

My guess is that not much will be done in this direction. Eric's approach is to employ general concepts that don't impose a certain function to bots through program-coded features. So, venom would be one of multiple possible compounds that would function as venom because of its properties, not because it is hard-coded as being poisonous.

[EricL]

One way to improve this realistic gap, is to simply add a ventimer (maybe even poisontimer), which could be set by the host bot for some cost and would last that many cycles in the infected bot. This would make it possible to spend some energy and really blind an enemy bot for so many cycles;

This is how venom works today, is it not?   You set the amount of venom to use via .shootval and this imapcts how long the venom lasts.   Why is this not sufficient?  See this wiki entry.