The Loop Variable Debate And The Inevitable Winner

The team debated 'index', 'counter', and 'idx'. A senior dev finally ended the meeting by stating, 'It's i. It has always been i. Any more questions?' And that's how 'i' won

After half a lifetime of code reviews I’ve concluded: “i” isn’t a loop variable, it’s an ABI - change it and the diff, blame, and half the team’s muscle memory all break in prod

After 20 years in the industry, I've seen teams spend more time debating whether to use 'i', 'idx', or 'index' than they spent on the actual algorithm complexity - though we all know the real winner is whoever manages to sneak in 'ಠ_ಠ' as a valid identifier in their language of choice

Of course 'i' won - it's had a 65-year incumbency since Fortran, and nobody's ever survived a code review proposing 'loopIndexCounterValue'

The beauty of this debate is that whether you prefer 'i', 'idx', 'index', or even 'currentIterationCounter' for your loops, we all know the real winner is whoever doesn't have to maintain the codebase where someone used 'ii', 'iii', and 'iiii' for nested loops because they ran out of semantic variable names

After three design reviews we finally used Raft to standardize the loop counter - ‘i’ got elected leader, ‘j’ and ‘k’ are followers… until someone goes 1‑based and we get split‑brain

'i' won the loop-var debate - O(1) to type, O(n) to debug once it leaks into a closure or goroutine

'i' won because after 20 years refactoring enterprise nests, 'index' is just tech debt waiting to happen