1.Simple flat‑file or SQLite solutions are often enough for early‑stage products
“diskcache is a good choice… For small amounts of data, a JSON file does the job… a JSONL file and a single Go binary will literally outlive most startup runways.” – vovanidze
2. The industry is over‑engineered – a tendency to reach for Kubernetes, multiple managed DBs, and “cloud‑only” stacks where none are needed
“people wildly underestimate the OS page cache… seeing startups spin up managed PostgreSQL + Redis clusters + Prisma on day 1 just to collect wait‑list emails is peak feature vomit.” – vovanidze
“we’ve basically been brainwashed to think we need Kubernetes and 3 different databases just to serve a few thousand users.” – vovanidze 3. Critics point out that the “no‑DB” narrative glosses over essential operational concerns such as backups, schema migrations, HA, and data‑loss risk
“Your article completely ignores operational considerations: backups, schema changes, replication/HA. As well as security, i.e. your application has full permissions to completely destroy your data file.” – evanelis
These three themes capture the dominant viewpoints: the practicality of lightweight storage, the critique of hype‑driven over‑complexity, and the warning that simplistic file‑based approaches can hide serious operational pitfalls.