When should you use unsigned types in C?

Reach for unsigned when you genuinely want wrap-around modular arithmetic or are treating the value as a bag of bits (masks, flags, logical shifts) — not merely because a number "can't be negative".

DO use unsigned for:

1. Modular arithmetic (multiprecision arithmetic)

   • When you want wrap-around behavior
   • Cryptographic operations

2. Bit manipulation (treating as bit vectors)

   • Logical right shift (no sign extension)
   • Flag fields, bitmasks
   • Network protocols, binary formats

DON'T use unsigned for:

1. Loop counters - Can cause infinite loops:

   // INFINITE LOOP!
   for (unsigned i = n-1; i >= 0; i--)
   // When i reaches 0 and decrements, it wraps to UINT_MAX
   

2. Sizes that might be compared with signed - Mixed comparisons are dangerous:

   // Always false if len is unsigned!
   if (len < 0)
   

3. Just because "it can't be negative" - The dangers outweigh the benefits

Best practice: Use size_t (unsigned) for array indices and sizes, but be careful with decrementing loops. Use ssize_t (signed) when -1 might indicate error.