Hacker News: dmarwicke

New comment by dmarwicke in "NPM to implement staged publishing after turbulent shift off classic tokens"

dmarwicke — Thu, 08 Jan 2026 15:12:49 +0000

2FA publishing still doesn't work for me. just use legacy tokens at this point, gave up trying to figure out what's wrong

New comment by dmarwicke in "Go.sum is not a lockfile"

dmarwicke — Thu, 08 Jan 2026 15:09:28 +0000

npm's package.json and package-lock.json get out of sync constantly on my team. at least go only has one file to mess up

New comment by dmarwicke in "Opus 4.5 is not the normal AI agent experience that I have had thus far"

dmarwicke — Tue, 06 Jan 2026 18:24:07 +0000

this is just optimizing for token windows. flat code = less context. we did the same thing with java when memory was expensive, called it "lightweight frameworks"

New comment by dmarwicke in "Show HN: DDL to Data – Generate realistic test data from SQL schemas"

dmarwicke — Tue, 06 Jan 2026 15:38:22 +0000

does it handle skewed distributions? faker's always been useless for this - like, your test data ends up with everyone having 5 orders when real data is all long tail

New comment by dmarwicke in "Show HN: Yeet – CLI tool to find and reclaim disk space"

dmarwicke — Tue, 06 Jan 2026 15:34:52 +0000

how does this decide what's safe to delete? i've nuked docker caches before and broken builds in annoying ways

New comment by dmarwicke in "Building a Rust-style static analyzer for C++ with AI"

dmarwicke — Mon, 05 Jan 2026 15:22:39 +0000

curious what the token costs look like on a real codebase. opus ain't cheap and C++ headers get big fast

New comment by dmarwicke in "Databases in 2025: A Year in Review"

dmarwicke — Mon, 05 Jan 2026 15:16:05 +0000

we had to restrict ours to views only because it kept trying to run updates. still breaks sometimes when it hallucinates column names but at least it can't do anything destructive

New comment by dmarwicke in "Show HN: CacheKit – an async Rust cache abstraction from a production system"

dmarwicke — Sun, 04 Jan 2026 15:47:21 +0000

couldn't find anything about invalidation in the docs. how does that work? usually where these abstractions fall apart for me

New comment by dmarwicke in "Show HN: Phantom Guard – Detect AI-hallucinated package attacks"

dmarwicke — Sun, 04 Jan 2026 15:43:57 +0000

does this end up flagging legit packages that just have 'ai' or 'gpt' in the name? feels like half of pypi would trigger at this point

New comment by dmarwicke in "Show HN: Boxed – Sovereign exec engine for AI agents (Vercel Sandbox inspired)"

dmarwicke — Sat, 03 Jan 2026 16:27:09 +0000

curious about the startup latency in practice. docker containers even with warm pools still feel sluggish for agent loops. e2b does firecracker and it's noticeably snappier

New comment by dmarwicke in "Show HN: Snowflake Emulator – Local Snowflake Development with Go and DuckDB"

dmarwicke — Sat, 03 Jan 2026 16:23:37 +0000

how well does the flatten() translation work in practice? every time i've used localstack or similar the queries work locally then break in subtle ways once deployed

New comment by dmarwicke in "Show HN: Jetbase – A Python database migration tool (Alembic alternative)"

dmarwicke — Sat, 03 Jan 2026 16:20:05 +0000

the sql-first thing is interesting. main thing that bugs me about alembic is i never know what order migrations will apply in when there's been a merge. how does jetbase handle branching?

New comment by dmarwicke in "Building an internal agent: Code-driven vs. LLM-driven workflows"

dmarwicke — Thu, 01 Jan 2026 20:49:35 +0000

hit this with support ticket filtering. llm kept missing weird edge cases. wrote some janky regex instead, works fine

New comment by dmarwicke in "Show HN: Arrayops – 100x faster Python array operations with Rust"

dmarwicke — Thu, 01 Jan 2026 20:46:24 +0000

does this work with multidimensional or just 1D? polars had the same limitation, ended up just sticking with numpy

New comment by dmarwicke in "Show HN: Basehook – Webhook management system built on Postgres"

dmarwicke — Wed, 31 Dec 2025 20:08:56 +0000

why postgres for this? feels heavy for a queue. tried something similar with redis and it was way simpler

New comment by dmarwicke in "Show HN: Repair-JSON-stream – Fix broken JSON from LLM streaming (1.7x faster)"

dmarwicke — Wed, 31 Dec 2025 20:05:41 +0000

does this handle the thinking blocks that claude puts in? those break json.parse constantly when it streams reasoning before the actual response

New comment by dmarwicke in "Show HN: 22 GB of Hacker News in SQLite"

dmarwicke — Tue, 30 Dec 2025 21:19:38 +0000

22gb for mostly text? tried loading the site, it's pretty slow. curious how the query performance is with this much data in sqlite

New comment by dmarwicke in "Postgres extension complements pgvector for performance and scale"

dmarwicke — Tue, 30 Dec 2025 21:16:19 +0000

does this actually fix metadata filtering during vector search? that's the thing that kills performance in pgvector. weaviate had the same problem, ended up using qdrant instead

New comment by dmarwicke in "Show HN: Mini-vLLM in ~500 lines of Python"

dmarwicke — Mon, 29 Dec 2025 21:19:56 +0000

does this do continuous batching or just static? couldn't tell from the code

New comment by dmarwicke in "Postgres and ClickHouse forming the default data stack for AI"

dmarwicke — Mon, 29 Dec 2025 21:16:33 +0000

how are people syncing these? i just have a cron job running overnight and it feels dumb. tried debezium but it was overkill