The fastest operation is the one your algorithm avoids
Remove repeated work with better structures, indexes, sorting strategies, batching, and precomputation.
Independent field notes / Vol. 01
Precise explanations of the machinery, patterns, and tradeoffs that make Python feel deceptively simple.
Read the cover storyThe index
7 notes, edited for signal.
Remove repeated work with better structures, indexes, sorting strategies, batching, and precomputation.
Measure temporary objects, retained memory, and data layout before trading clarity for reuse.
Design repeatable timing experiments that measure the workload you actually care about.
Move useful work across Python, native, I/O, and process boundaries without optimizing the wrong crossing.
Separate Python call semantics from CPython vectorcall, then design APIs that avoid needless argument and boundary overhead.
Use cProfile and pstats to locate expensive call paths before choosing what to optimize.
Inspect how CPython 3.14 adapts hot bytecode to stable runtime types, and where it cannot help.
Editorial policy
No listicles. No folklore. No just trust me abstractions.
Every note should leave you with a better model of the system, not just another snippet to paste.