Hacker News: qyph

New comment by qyph in "EFF is leaving X"

qyph — Thu, 09 Apr 2026 18:54:53 +0000

https://x.com/elonmusk/status/2030202550259962338

New comment by qyph in "Researchers complete first human trial on viability of enteral ventilation"

qyph — Wed, 22 Oct 2025 01:04:04 +0000

I wonder if this could be used for doping in aerobic sports? Could this elevate overall oxygen intake in a healthy person?

My vague understanding is that oxygen intake is a big limiting factor in aerobic activities hence measurement of things like vo2max in sports science. ‘Blood doping’ has similar benefits though it’s also about having more blood period.

It seems unlikely that one could take a big enough suppository to help in a meaningful way in a marathon, but in a middle distance race lasting only a few minutes…

New comment by qyph in "Leaving Google"

qyph — Sun, 11 May 2025 04:10:44 +0000

Gmail wasn't created with 20% time.

https://time.com/43263/gmail-10th-anniversary/#:~:text=Gmail...

New comment by qyph in "Sublinear Time Algorithms"

qyph — Mon, 24 Feb 2025 03:46:54 +0000

https://en.wikipedia.org/wiki/AKS_primality_test though it's number theory, and concerned with numbers of size n, rather than lists of length n.

Also relevant: https://www.cs.yale.edu/homes/aspnes/pinewiki/Derandomizatio...

New comment by qyph in "Sublinear Time Algorithms"

qyph — Mon, 24 Feb 2025 03:44:44 +0000

Hyperloglog analyses generally assume access to the full data stream, and so are O(n) at a minimum. Perhaps by running hyperloglog on a sublinear sample of the dataset you'd get an algorithm in this class.

New comment by qyph in "Toni Morrison's Rejection Letters"

qyph — Sun, 31 Mar 2024 06:58:11 +0000

You have it precisely backwards. "Developing scenes in which the reader discovers what kind of people they are" is synonymous with "show" and "being told" is literally a form of "tell."

New comment by qyph in "If we lose the Internet Archive, we’re screwed"

qyph — Thu, 06 Apr 2023 21:10:25 +0000

From your article: > The United States became a party in 1989.

The person you are responding to is writing about American Copyright. https://en.wikipedia.org/wiki/Copyright_law_of_the_United_St.... The first american copyright law is from 1790, which does in fact predate the bern convention, and our life of the author plus 50 years rule is from the copyright act of 1976, 13 years before the US ratified the bern convention.

So yes, it is an 'American thing' as are all issues of law in the US. And if you think that this kind of convention is meaningfully binding, just checkout the history of the US and other major powers with regards to various other international treaties, like the ICC https://en.wikipedia.org/wiki/International_Criminal_Court.

New comment by qyph in "Twitter's Recommendation Algorithm"

qyph — Fri, 31 Mar 2023 20:11:33 +0000

Those aren't redundant or collinear though? Maybe you are surprised they didn't encode this as an integer "num_images"? It is fairly common to one hot encode ordinal variables with only a few common/possible values this way.

New comment by qyph in "Is this poison ivy?"

qyph — Sat, 04 Feb 2023 00:01:42 +0000

Blackberry has thorns and serrated leaves so you can distinguish it fairly easily regardless of season. When there are leaves on the plants there aren't really many plants that are easy to confuse with poison oak around here. Just look for leaves that come in sets of 3 leaflets with rounded/smooth lobes.

New comment by qyph in "Estimating square roots in your head"

qyph — Fri, 03 Feb 2023 03:38:58 +0000

Your method apparently isn't very good seeing as a square with area 2300 has sides of length ~48

New comment by qyph in "Show HN: Can you lose at Wordle if you tried?"

qyph — Wed, 01 Jun 2022 01:50:07 +0000

The goal is to avoid entering the word. In this case 'focal' was the word. You entered 4 guesses, but only had 1 valid guess, + the 'wordle' left so it was impossible to finish without undoing.

New comment by qyph in "Thanks to interbreeding, just 7% of our DNA is unique to modern humans: study"

qyph — Wed, 28 Jul 2021 22:26:27 +0000

Thanks, your comment makes the basic differences clear to me.

New comment by qyph in "Thanks to interbreeding, just 7% of our DNA is unique to modern humans: study"

qyph — Wed, 28 Jul 2021 20:52:48 +0000

Is this for some subset of our DNA? Quick skim of the article didn't tell me. But there is a popular factoid that human dna is 98.8% similar to chimpanzee dna on average. Is that factoid false?

New comment by qyph in "Stack Overflow sold to Prosus for $1.8B"

qyph — Wed, 02 Jun 2021 20:32:10 +0000

Perhaps they want liquidity, a more diverse portfolio, etc.

New comment by qyph in "Apple has gone to extraordinary lengths to make scroll bars invisible"

qyph — Wed, 10 Mar 2021 20:42:10 +0000

The problem I always had with it is that the little symbols are inscrutable, and mouse-over for a description is inconsistent. Because of this I found the ribbon far less discoverable than text based menus.

New comment by qyph in "Bitcoin network average energy consumption per transaction compared to VISA"

qyph — Mon, 08 Mar 2021 19:17:31 +0000

Do you think the entire us economy would disappear if bitcoin became the dominant currency of global trade? People will still want to trade the same goods and services.

New comment by qyph in "Google to stop selling ads based on your specific web browsing"

qyph — Wed, 03 Mar 2021 18:51:20 +0000

In floc no personal data or web history is uploaded to the cloud. This is the essence of "federated learning" - your browser does the big ml computation, then just uploads the result.

New comment by qyph in "Ask HN: Does LinkedIn rate limit when browsing your own contacts?"

qyph — Tue, 23 Feb 2021 20:56:00 +0000

LinkedIn was down earlier today.

New comment by qyph in "Chat Bubble Blindness"

qyph — Wed, 05 Aug 2020 20:06:29 +0000

A z test relies on the normal approximation, no? I don't think that is appropriate with proportions so close to 0.