Hacker News: ankitml

New comment by ankitml in "The only scalable delete in Postgres is DROP TABLE"

ankitml — Mon, 15 Jun 2026 00:21:08 +0000

This is directionally correct approach. Deleting a large chunk of rows, in a large table does lead to unpredictable-bad behaviour for a while until those dead tuples are handled.

I have used a very similar strategy by forking repack client https://github.com/reorg/pg_repack/pull/326 This works out of the box with rds/cloudsql etc.

New comment by ankitml in "New quantum state of matter found at interface of exotic materials"

ankitml — Mon, 04 Aug 2025 12:38:18 +0000

So your claim is spin ice as a concept doesnt exist in particle physics and would not be discussed outside solid state?

New comment by ankitml in "Gemini 2.5 Deep Think"

ankitml — Fri, 01 Aug 2025 23:43:48 +0000

I have ultra. Will not be renewing it. Useless, at least have global limits and let people decide how they want to use it. If I have tokens left, why can't I use it for code assist?

New comment by ankitml in "Photon transport through the entire adult human head"

ankitml — Wed, 18 Jun 2025 12:47:39 +0000

Hypothetically my ancestors were cheetah and not hominoid apes. One needs to prove foundational hypothesis before larger claims that can be built on top of those.

New comment by ankitml in "Photon transport through the entire adult human head"

ankitml — Tue, 17 Jun 2025 23:25:16 +0000

Teleport is not a thing a photon can do. Leave alone across brain

New comment by ankitml in "US Streetlights Are Turning Purple"

ankitml — Tue, 17 Jun 2025 23:02:12 +0000

Design choices must include failure modes. IMHO, utterly stupid to only consider happy path.

New comment by ankitml in "Watching AI drive Microsoft employees insane"

ankitml — Wed, 21 May 2025 11:29:18 +0000

GitHub is not the place to write code. IDE is the place. Along with pre CI checks, some tests, coverage etc. they should get some PM before making decisions..

New comment by ankitml in "Haskelling My Python"

ankitml — Mon, 21 Apr 2025 22:15:50 +0000

It is possible to test the chaining though, if you know your data well. If not, those edge cases in the data quality can throw things off balance very easily.

New comment by ankitml in "Dead trees keep surprisingly large amounts of carbon out of atmosphere"

ankitml — Mon, 14 Apr 2025 15:24:58 +0000

Dead trees can exist as furniture, flooring and buildings too. Stored carbon with ancillary usage.

New comment by ankitml in "Google is publishing the home addresses of developers without their consent"

ankitml — Fri, 28 Mar 2025 14:08:18 +0000

Amazon did that to a physical product my wife is selling. I was very annoyed. But it ended up being a configuration but default being `show address` state is definitely annoying.

New comment by ankitml in "I tried making artificial sunlight at home"

ankitml — Fri, 28 Mar 2025 14:05:41 +0000

you do you. NIR is the part of sunlight that makes you "feel" good or alive or whatever nice thing you are looking from being outside.. Ill stick with NIR is essential component for calling stream of photons as sun-like.. or not sun-like.

New comment by ankitml in "I tried making artificial sunlight at home"

ankitml — Thu, 27 Mar 2025 22:07:19 +0000

Yep, that was my first assessment. Cant call it sun without NIR.

New comment by ankitml in "Sharding Pgvector"

ankitml — Thu, 27 Mar 2025 02:51:16 +0000

Well, I have a brute force strategy for pgvector working reasonably well. Individual, partial indexes. It works for all those queries with category_id=<> clauses. You only need an index for larger categories, for categories with rows below a threshold you dont need index a KNN/dot product would work.

New comment by ankitml in "Sharding Pgvector"

ankitml — Thu, 27 Mar 2025 02:44:55 +0000

Depends on the hardware, sharding layout, bulk/individual insert and table bloat. Looks like you are missing pg_repack..

New comment by ankitml in "670nm red light exposure improved aged mitochondrial function, colour vision"

ankitml — Thu, 20 Feb 2025 23:53:17 +0000

get a broad spectrum red light bulb, that is most well understood and is least risky. Cons is the targetted spectrum would be only a small portion of it but good thing is it is a 4$ cheap test. It worked for me for smaller issues, like minor burns, cuts etc.

New comment by ankitml in "670nm red light exposure improved aged mitochondrial function, colour vision"

ankitml — Thu, 20 Feb 2025 23:49:15 +0000

Vitamin D is not an outcome of standing by fire. It is from UV spectrum which is entirely absent from fires. It needs high temperature fire like fusion to be emitted. Totally doable by sun but not by your campfire.

New comment by ankitml in "670nm red light exposure improved aged mitochondrial function, colour vision"

ankitml — Thu, 20 Feb 2025 23:48:06 +0000

Morning and evening is entirely infrared (80-90 percent with rest of it visible). No UV.

New comment by ankitml in "670nm red light exposure improved aged mitochondrial function, colour vision"

ankitml — Thu, 20 Feb 2025 23:46:52 +0000

i dont know what you mean by science. There are literally 1000s of research papers showing mitochondrial "horsepower" with red light on every type of tissue. Cells heal themselves as first thing when they get extra energy. Do you want your neighborhood clinic to validate before trying some light samples out?

New comment by ankitml in "670nm red light exposure improved aged mitochondrial function, colour vision"

ankitml — Thu, 20 Feb 2025 23:45:00 +0000

skepticism without homework is stupidity. The above snark demonstrates effectively.

New comment by ankitml in "670nm red light exposure improved aged mitochondrial function, colour vision"

ankitml — Thu, 20 Feb 2025 23:43:01 +0000

infrared, yes. Although depends on skin colour wrapping the skull. Melanin absorbs everything under the sun. It has very high absorption of UV but somewhat absorbs infrared too in this range of frequencies. Darker the color of skin, higher the melanin and higher the absorption by skin.