Archive
Hands-on Python dataclass vs slots vs dict lab: instantiation and attribute ops/s plus RSS for classic, slots, dataclass, namedtuple, dict on localhost.
Hands-on getaddrinfo cost lab: localhost vs numeric IP vs AI_NUMERICHOST, plus hosts/stub names with real p50/p95 µs. DNS sinkhole labeled honestly here.
Hands-on atomic rename vs overwrite lab: tempfile+os.replace vs truncate-write with/without fsync. Real ops/s on localhost; crash-safety limits labeled.
Hands-on uuid.uuid4 vs secrets.token_hex/urlsafe vs os.urandom ops/s on localhost. Throughput only; uniqueness and collisions not measured in this lab.
Hands-on base64 vs hex encode/decode lab on 32MiB buffers with real MB/s measurements in both directions on localhost.
Hands-on Python re vs str methods lab on a 32MiB localhost corpus: count, find, line prefix, case-insensitive, and email scans with real measured MB/s.
Hands-on posix_fadvise SEQUENTIAL vs RANDOM vs NORMAL lab on localhost: cold and warm 256MiB reads after drop_caches with real measured p50 MB/s and RPS.
Hands-on fork COW vs multiprocessing spawn lab: PSS/Private_Dirty after fork and page dirtying, plus spawn start latency. Real numbers on the lab box.
Hands-on Python json vs orjson vs msgpack lab: encode/decode MB/s on records, nested, and unicode payloads. Real p50 numbers measured on localhost box.
Hands-on SHA-256 vs BLAKE2b vs xxHash hash throughput on 64MiB buffers (SHA-1/SHA3 labeled). Real MB/s with OpenSSL and xxHash versions on the lab box.
Hands-on SQLite WAL vs DELETE vs OFF journal_mode lab: insert/update RPS (FULL/NORMAL sync), plus WAL checkpoint cost. Real p50 numbers on localhost box.
Hands-on zstd vs gzip vs lz4 on 64MiB payloads (real /usr text, mixed, urandom). Measured compress/decompress MB/s and ratio at default levels. No Docker.