Hacker News: DennisL123

New comment by DennisL123 in "Claude Code is unusable for complex engineering tasks with the Feb updates"

DennisL123 — Mon, 06 Apr 2026 20:55:55 +0000

Gotcha. It seemed though from the replies on the github ticket that at least some of the problem was unrelated to effort settings.

New comment by DennisL123 in "Issue: Claude Code is unusable for complex engineering tasks with Feb updates"

DennisL123 — Mon, 06 Apr 2026 18:14:22 +0000

Happy to have my mind changed, yet I am not 100% convinced closing the issue as completed captures the feedback.

New comment by DennisL123 in "We should revisit literate programming in the agent era"

DennisL123 — Mon, 09 Mar 2026 07:47:59 +0000

If agents can already read and rewrite code, literate programming might actually be unnecessary. Instead of maintaining prose alongside code, you could let agents generate explanations on demand. The real requirement becomes writing code in a form that is easily interpretable and transformable by the next agent in the chain. In that model, code itself becomes the stable interface, while prose is just an ephemeral view generated whenever a human (or another agent) needs it.

New comment by DennisL123 in "OsmAnd’s Faster Offline Navigation (2025)"

DennisL123 — Fri, 27 Feb 2026 17:30:09 +0000

Yes, the idea was to transition to a variant of MLD that lends itself to the offline use case and would use an updated modeling to be more space efficient while preserving performance. There exists an incomplete (and private) PoC written in Rust.

New comment by DennisL123 in "OsmAnd’s Faster Offline Navigation (2025)"

DennisL123 — Fri, 27 Feb 2026 07:34:42 +0000

Who knows .. ;)

New comment by DennisL123 in "Ask HN: Why does Google Maps still use mercator projection?"

DennisL123 — Thu, 22 Jan 2026 08:17:53 +0000

Agree, mercator projection isn’t great. But it’s pretty simple.

All of the mapping apps are rooted in paper maps. That’s what most people find accessible in a natural way.

So, in any 2D world view some projection must be chosen, and you can fundamentally chose between true angles or true size. Because of that choice any projection is a distortion. Choosing true angles has advantages when it comes to turning projected data into something like turn instructions in your nav app. And then again, mercator projection is easy to use. So, bottom line it’s a mix of people are used to it and simplicity of using it.

New comment by DennisL123 in "Manus AI 100M USD ARR"

DennisL123 — Sun, 28 Dec 2025 09:55:30 +0000

Last minute I sold 200 €/$£/¥ worth of product. At this rate and because online sales is 24/7, I am hitting 200*60*24*365 = 105.120.000 ARR. tada!

New comment by DennisL123 in "Germany outfitted half a million balconies with solar panels"

DennisL123 — Mon, 06 Oct 2025 07:37:43 +0000

In the US the price varies a lot by state [1].

[1] https://www.electricchoice.com/electricity-prices-by-state/

New comment by DennisL123 in "Move over Dijkstra: New algorithm just rewrote 70 years of computer science"

DennisL123 — Sun, 05 Oct 2025 18:24:08 +0000

Yep, that’s the one. There are a number of follow-up papers that engineer individual aspects of the implementation.

New comment by DennisL123 in "Move over Dijkstra: New algorithm just rewrote 70 years of computer science"

DennisL123 — Sat, 04 Oct 2025 11:21:04 +0000

I was referring to variants of Customizable Route Planning. Easiest to implement is likely the partitioner of Sommer et al, and a unidirectional multi-level dijkstra.

New comment by DennisL123 in "Move over Dijkstra: New algorithm just rewrote 70 years of computer science"

DennisL123 — Fri, 03 Oct 2025 14:08:05 +0000

OSRM founder, here. Yes, you are right, many of the speedup techniques are related. My personal opinion is, tho, that looking at the identification of important nodes is best captured by the ideas of applying partitioning to multi-level dijkstra and by what’s called hub-labels. The latter has a close relationship to Contraction Hierarchies.

New comment by DennisL123 in "The unreasonable effectiveness of modern sort algorithms"

DennisL123 — Sun, 14 Sep 2025 07:59:09 +0000

Efficiency, not effectiveness. They are all effective in the sense that they produce sorted results. Even the non-modern sort algorithms are effective in the sense that the results are correct. This should be about the efficiency with which they do it, right?

New comment by DennisL123 in "Hashed sorting is typically faster than hash tables"

DennisL123 — Thu, 11 Sep 2025 14:52:50 +0000

A way to give radix sort a performance boost is to allocate uninitialized blocks of memory for each bucket that are of size n. It exploits the MMU and you don’t have to worry about managing anything related to buckets potentially overwriting. The MMU is doing all the bookkeeping for you and the OS actually allocates memory only on page access.

New comment by DennisL123 in "SIMD Binary Heap Operations"

DennisL123 — Thu, 14 Aug 2025 18:58:19 +0000

A winner tree uses extra space, doesn't it? That might exclude it from certain applications to be an alternative. Four-ary heaps are roughly (ymmv) twice as fast as binary heaps by exploiting cache locality in a better way for small key/value types. And it seems to be a sweet spot since eight-ary heaps don’t deliver additional improvements.

New comment by DennisL123 in "Mozilla Firefox – Official GitHub repo"

DennisL123 — Tue, 13 May 2025 06:48:11 +0000

A BUILD.md could be useful.

New comment by DennisL123 in "Faster sorting with SIMD CUDA intrinsics (2024)"

DennisL123 — Tue, 06 May 2025 06:18:24 +0000

Interesting stuff. Not sure if I read this right that it‘s 16 und 32 bit values of integers that get sorted. If yes, I‘d love to see if the GPU implementation can beat a competitive Radix sort implementation on a CPU.

New comment by DennisL123 in "Shortest-possible walking tour to 81,998 bars in South Korea"

DennisL123 — Thu, 24 Apr 2025 05:12:22 +0000

OSRM lead dev here. Love to see this large of an instance being solved.

New comment by DennisL123 in "Show HN: OpenTimes – Free travel times between U.S. Census geographies"

DennisL123 — Tue, 18 Mar 2025 19:48:35 +0000

It precomputes partial routes that are combined at run time. :)

New comment by DennisL123 in "GPT-4.5"

DennisL123 — Fri, 28 Feb 2025 07:11:20 +0000

tl;dr: doesn‘t work as expected and we sank a ton of money on it too.

New comment by DennisL123 in "My washing machine refreshed my thinking on software estimation"

DennisL123 — Tue, 18 Feb 2025 09:59:12 +0000

Shitty tools lead to wasted time. True for programming, and apparently also for appliance installations.

Good job on pulling through, though.