Patterns
An approachable and distributable lang for systems / algorithms
As you remove code's language and context specific formalities it becomes a ⧉pattern. A creatively condensed, and accessible vessel for process logic:
// attendance incentive system
sessions = #
streak = #
Miss()
streak = Min(0, streak - 1)
if (streak < 0)
then sessions--
Show()
streak = Max(1, streak + 1)
if (streak is not divisible by 4)
then sessions--
AI language models like OpenAI Codex can both convert to and from this format effortlessly. Inadvertently making it the most powerful way to open source pieces of code across languages and toolsets.
Best practice conventions seem very flexible in my experience though there's plenty of room for experimentation and iteration.
Division
By giving our third party issues a non zero chance in the sun, we temper their engagement and account for them more readily. That way issues won't fester until they boil over and swing out the other direction.
working model
// a system for determining the outcome of an election
// resulting in an normal distribution of outcomes
// that better maps onto the voting populace
cutoff = 0.5 / # of options
option.chance = option.votes / total votes
option.chance = max(option.chance - cutoff, 0)
option.chance /= total chances
roll = random(0.0-1.0)
foreach option in options
if roll < option.chance
then return option as winner
else
roll -= option.chance
- Try your luck instead of dwelling on unresolved issues
- Blame chance rather than your neighbor
- Soften our many one sided political spheres
real world application
Running open "pre primaries" using less divisive voting systems like this while leveraging tech to minimize the overhead. Can safely prove out these policies and better support aligned candidates