Hacker News: leeoniya

New comment by leeoniya in "I set a trap for a book-marketing scammer (2025)"

leeoniya — Sun, 23 Aug 2026 03:12:56 +0000

i also tapped out quickly. how many times can you describe the same thing?

Claude: "Hold my beer."

New comment by leeoniya in "Going Dark, and the era of law enforcement hacking"

leeoniya — Sat, 15 Aug 2026 02:40:16 +0000

> The one shown in Wikipedia is mine.

neat!

https://upload.wikimedia.org/wikipedia/commons/e/e3/Pen_Regi...

assuming you still have it, it wouldn't be the worst idea to update the pic to something without chromatic aberration and motion blur. the wiki pic is probably the worst of the bunch compared to others on google image search :)

New comment by leeoniya in "Stop sending me huge PRs; a rant"

leeoniya — Sat, 15 Aug 2026 02:12:44 +0000

It's possible to do a large amount of AI-assisted work, then do a second/third/fourth pass to break it up into a reviewable stack of self-contained PRs. But it takes time, and there's no such thing as one-shotting it. And it's basically impossible to continuously rebase manually without burning tokens. The way to merge the stack is more-or-less "stop the world".

However, I have yet to see how this will play out with upstream contribs:

https://github.com/moment/luxon/discussions/1796

https://github.com/leeoniya/luxon/tree/leeoniya/perf-patches...

New comment by leeoniya in "Improving Heuristics for A* Pathfinding"

leeoniya — Sun, 09 Aug 2026 05:40:44 +0000

i've always been impressed by NBA*: https://github.com/anvaka/ngraph.path

demo: https://anvaka.github.io/ngraph.path.demo/

Show HN: easy-tz – A fast, 10KB, dependency-free getTimeZonesAt(timestamp)

leeoniya — Thu, 16 Jul 2026 17:55:53 +0000

If you have a time zone picker in your web UI that shows accurate & current time zone abbreviations, there's a good chance it's pulling in 770KB of moment-timezone dependency. And if you were thinking of replacing moment with luxon, this use case would be a potential blocker for shrinking your bundle.

It is surprisingly challenging [and soul-sucking] to work with timezone lore with the goal of reducing this down to something small and fast. There are many performance foot-guns and correctness edge-cases that have to be addressed. With the help of AI (Fable 5 here), this becomes tractable for a single engineer to iterate on in the span of a weekend. I worked through about 12 different prototypes during the process, mixing and matching various strategies and approaches for every aspect of the final result. This should tell you something about how much taste and steering still needs to be applied even to frontier models by an experienced swe to produce something of high production-grade quality, not just, "it works, ship it". If you're familiar with my existing OSS work, I would not ship any level of unaudited slop :) Also, https://github.com/leeoniya/easy-tz/blob/main/comparison.md is not to be missed.

Hopefully one other dev (or agent) finds this useful; I think it fills a missing gap in the frontend timezone-handling ecosystem.

disclaimer: still 0.1.0, so if you use any of the non-default implementations in the lib, expect some turbulence before 1.0.

Comments URL: https://news.ycombinator.com/item?id=48937896

Points: 2

# Comments: 0

New comment by leeoniya in "Dopamine Fracking"

leeoniya — Mon, 08 Jun 2026 13:19:26 +0000

so, Your Attention is All They Need?

New comment by leeoniya in "Gnutella: A Protocol Outliving the World That Created It"

leeoniya — Mon, 25 May 2026 15:37:53 +0000

or was it Windows hiding file extensions by default and you downloaded a .mp3.exe file?

New comment by leeoniya in "PHP's Oddities"

leeoniya — Sat, 23 May 2026 18:50:47 +0000

surprised not to see http://phpsadness.com/ here

New comment by leeoniya in "Meta blocks human rights accounts from reaching audiences in Arabia and the UAE"

leeoniya — Wed, 20 May 2026 16:16:15 +0000

the term "Americans" always bothered me, though it's commonly used to refer to US

New comment by leeoniya in "I believe there are entire companies right now under AI psychosis"

leeoniya — Fri, 15 May 2026 21:13:24 +0000

> "no no, it has full test coverage"

i don't have enough fingers (and toes) to count how many times i've demonstrated that "100% coverage" is almost universally bullshit.

New comment by leeoniya in "Golden Testing a CAD Library"

leeoniya — Thu, 14 May 2026 07:07:10 +0000

if you want to do image diffing, use https://github.com/dmtrKovalenko/odiff

however, if you have SVGs already, compare those without rasterizing them first, then if that fails rasterize & odiff the baseline and new on-demand. this way you dont waste a ton of space storing a bunch of binary [possibly evolving] pngs in git.

New comment by leeoniya in "Fast and Easy Levenshtein distance using a Trie (2011)"

leeoniya — Thu, 16 Apr 2026 13:48:49 +0000

Levenshtein distance is often a poor way to fuzzy match or rank. i suspect that in js, even the trie approach would incur significant GC/alloc thrashing or cost of building a huge trie index.

i tried fuzzy matching using a cleverly-assembled regexp approach which works surprisingly well: https://github.com/leeoniya/uFuzzy

New comment by leeoniya in "EPA official in charge of methane regulations was an oil and gas lobbyist"

leeoniya — Sun, 05 Apr 2026 15:24:50 +0000

> Sen. Sheldon Whitehouse

ok, i was confused for a minute.

> And, given that the gas breaks down relatively quickly, this would have been one of the fastest ways to reduce global warming.

s/quickly/slowly?

New comment by leeoniya in "Claude Code's source code has been leaked via a map file in their NPM registry"

leeoniya — Tue, 31 Mar 2026 18:30:56 +0000

> for a 2–3% performance gain

this is highly workload-dependent. there are plenty of APIs that are multiple-factor faster and 10x more memory efficient due to native implementation.

New comment by leeoniya in "Pretext: TypeScript library for multiline text measurement and layout"

leeoniya — Mon, 30 Mar 2026 09:32:02 +0000

ack, see my reply to sibling comment: https://news.ycombinator.com/item?id=47572206

New comment by leeoniya in "Pretext: TypeScript library for multiline text measurement and layout"

leeoniya — Mon, 30 Mar 2026 09:27:00 +0000

no disagreement. i never claimed uWrap did anything more than it does. it was not meant for typography/text-layout but for line count estimation. it never needed to be perfect, and i did not claim equivalence to pretext. however, for the use case of virtualization of data tables -- a use case pretext is also targeted at -- and in the common case of latin alphabets, right now pretext is significantly slower. i hope that it can become faster despite its much more thorough support for "rest of the owl".

New comment by leeoniya in "Pretext: TypeScript library for multiline text measurement and layout"

leeoniya — Mon, 30 Mar 2026 09:17:52 +0000

i don't have a mac to test this with currently, so hopefully it's not the price but a matter of adding a Safari-specific adjustement :)

internally it still uses the Canvas measureText() API, so there's nothing fundamentally that should differ unless Safari has broken measureText, which tbh, would not be out of character for that browser.

New comment by leeoniya in "Pretext: TypeScript library for multiline text measurement and layout"

leeoniya — Mon, 30 Mar 2026 02:59:48 +0000

tried just doing a concat of the 100k sentences with line breaks, it wasnt much faster, ~1880ms.

New comment by leeoniya in "Pretext: TypeScript library for multiline text measurement and layout"

leeoniya — Mon, 30 Mar 2026 00:41:16 +0000

it is not clear from the API/docs how i would use prepare() once on one text and then use layout() for completely different text.

i think the intended purpose is that your text is maybe large but static and your layout just changes quickly. this is not the case for figuring out the height of 100k rows of different texts in a table, for example.

New comment by leeoniya in "Pretext: TypeScript library for multiline text measurement and layout"

leeoniya — Mon, 30 Mar 2026 00:39:10 +0000

correct, it was meant for estimating row height for virtualizing a 100k row table with a latin-ish LTR charset (no emoji handling, etc). its scope is much narrower. still, the difference in perf is significant, which i have found to be true in general of AI-generated geenfield code.