Ragnar {Groot Koerkamp}
LightNews LightNews
banner
curiouscoding.nl
Ragnar {Groot Koerkamp}
@curiouscoding.nl
970 followers 110 following 1.3K posts
Postdoc on high troughput bioinformatics @ KIT Karlsruhe;
IMO, ICPC, Xoogler, Rust, road-cycling, hiking, wild camping, photography
Posts Replies Media Videos
Reposted by Ragnar {Groot Koerkamp}
jnalanko.bsky.social
The architecture and art at the DSB 2026 workshop is out of this world. The talks are also great! #dsb2026
February 18, 2026 at 10:44 AM
Reposted by Ragnar {Groot Koerkamp}
npmalfoy.bsky.social
The great @curiouscoding.nl present its quadrank work at #DSB2026
February 18, 2026 at 8:28 AM
curiouscoding.nl
QuadRank slides!

curiouscoding.nl/slides/quadr...
QuadRank: Engineering a high-throughput rank
curiouscoding.nl
February 18, 2026 at 1:09 AM
Reposted by Ragnar {Groot Koerkamp}
robp.bsky.social
How would you design a *multithreaded*, *concurrent* & *dynamic* hash table if you are focused specifically on common k-mer workloads, where streaming query & insertion are common? Jamshed, Prashant and I explore this in kache-hash, a cache-friendly k-mer hash table!
www.biorxiv.org/content/10.6...
www.biorxiv.org
February 17, 2026 at 6:49 PM
Reposted by Ragnar {Groot Koerkamp}
biorxiv-bioinfo.bsky.social
kache-hash: A dynamic, concurrent, and cache-efficient hash table for streaming k-mer operations https://www.biorxiv.org/content/10.64898/2026.02.13.705625v1
February 17, 2026 at 5:47 AM
Reposted by Ragnar {Groot Koerkamp}
divbyzero.bsky.social
I tell my topology students that sets are not like doors, which are either open or closed. A set can be open, closed, both, or neither.

I guess I should rethink that assertion. Would you say a "Dutch door" can be both open and closed? (Or neither open nor closed)?
February 16, 2026 at 4:52 PM
Reposted by Ragnar {Groot Koerkamp}
myrrlyn.net
i do not understand the point of living like this and i am glad that i do not suffer from this affliction
timkellogg.me Tim Kellogg @timkellogg.me · 3d
Token Anxiety

i think i mostly echo this for myself. with so much that can be done, i often feel like i *should* be doing something, always
Token Anxiety Nikunj Kothari C @nikunj • Feb 13 A friend left a party at 9:30 on a Saturday. Not tired. Not sick. He wanted to get back to his agents. Nobody questions it anymore. Half the room is thinking the same thing. The other half are probably checking the progress of their agents. At a party. All the parties are sober now. Young people don't drink because they're going back to work after. Not inspired by Bryan Johnson, although that's probably a factor. The buzz they want now runs on tokens per day. I keep noticing it on walks through the Mission. Laptops glowing everywhere. Cafes, sidewalks, heck even park benches. People walking with screens open like a flashlight guiding them somewhere. Less drunk laughter on the streets these days. More keystrokes. Dinner conversations used to start with "what are you building?" That's over. Now it's "how many agents do you have running?" People drop the number the way they used to drop their follower count. Quietly competitive. The flex isn't what you've accomplished anymore. It's what's working while you're sitting here not working. The vocabulary is what really gets me though. People describe models the way sommeliers describe wine. This one has better taste. That one hallucinates with more confidence. Opus is bold, Codex is smooth. They talk about harnesses and reins like they're controlling horses. Invisible whips directing invisible labor. Someone at a dinner said they keep labor. Someone at a dinner said they keep "Claude on a tight leash for code review but give it more slack for creative work." We've started borrowing the language of how we treat animals for something none of us actually understand yet. Waking up and checking what your agents produced overnight is the first thing now. Before coffee. Before texts. You open your laptop and grade homework you assigned in your sleep. Some of it is good. Most needs rework. But you start shipping a plan before you sleep just so you can wake up to more code written overnight. Saturdays became uninterrupted build windows. No meetings, no Slack, twelve hours of you and your agents. Sunday morning X is all terminal screenshots and shipping receipts. "What'd you ship this weekend?" replaced "what'd you do this weekend?" The anxiety is rational, which is why it sticks. Every week some new benchmark drops that makes last month's workflow feel prehistoric. Codex ships overnight processing. Opus gets faster. Context windows double. None of it reduces the pressure. It multiplies it. You can do more now. And someone already is. The window to be first at anything feels like it's shrinking by the day. Literally, by the day. I replaced Netflix with Claude Code. I lie in bed thinking about what I can spin up before I fall asleep, what can run while I'm unconscious. Reading a novel feels indulgent now. Watching a movie without a laptop open feels wasteful. This voice in my head that says "something could be running right now" just doesn't shut off. I'm not even building a company. I'm just addicted to building my random ideas. Everyone here knows they should step away more. That's not the problem. The problem is what your brain does when you try. I still take a x ss walks. The agents come with me now.
February 15, 2026 at 5:04 PM
Reposted by Ragnar {Groot Koerkamp}
tef.bsky.social
which is fine and good but my problem is the notion that code is nothing more than a medium for business interests, it just feels like the same dismissal we normally see from coders about art
February 15, 2026 at 2:36 AM
Reposted by Ragnar {Groot Koerkamp}
robp.bsky.social
I’ve written a post about my recent experiences (successes) with AI coding models; the experiences that caused me to re-evaluate my initial judgements, the surprise I had at what can be accomplished, & some fears I have about these tools. Discussion welcome! combine-lab.github.io/blog/2026/02...
COMBINE-lab - The skeptic’s guide to generative AI assisted coding
An easy-to-use, flexible website template for labs, with automatic citations, GitHub tag imports, pre-built components, and more.
combine-lab.github.io
February 15, 2026 at 4:31 AM
Reposted by Ragnar {Groot Koerkamp}
penryu.bsky.social
Me using git in 2024: "it has a few rough edges, sure, but it's really nice!"

Me using git in 2025, having used jj for 2 months: "HOW DO YOU PEOPLE LIFE LIKE THIS?!"
February 14, 2026 at 12:45 AM
Reposted by Ragnar {Groot Koerkamp}
redblobgames.com
I'm in between. I'll normally have a few tabs open but I'll sometimes have 20-50 and then I go through them, get what I wanted from them, and close them. It's kind of like the Creative Process diagram from lostgarden.com/2010/08/17/v...
Visualizing the Creative Process
As I coach new developers, I’ve taken to scribbling out the same useful diagram for visualizing the creative process again and again on coffee-ringed napkins.  In order to limit my future abu…
lostgarden.com
February 14, 2026 at 2:37 PM
curiouscoding.nl
Software development is performance art!
adapossum.bsky.social Ada 🏳️‍⚧️ @adapossum.bsky.social · 3d
Not to mention that said dependencies are the only thing that have kept me interested in programming. Maybe it is my autism, but crates like wgpu and bevy just scratch my brain in a deeply satisfying way. Not just the code, but watching them evolve, grow. Watching how they are maintained.
February 14, 2026 at 10:44 PM
curiouscoding.nl
Supplychain attack where I publish 1000 repos with slightly varying BFS implementations, yet all of them include an `rm -rf /` somewhere random inside it.
sachy.dev Sachymetsu @sachy.dev · 3d
Supply chain attacks are a real concern, but handwritten code also has its own liabilities. Ultimately, you cannot escape the cost of code being a liability.

You gotta do your due diligence at some point. But it is kinda weird you don't trust deps but trust LLMs trained on those deps.
February 14, 2026 at 10:41 PM
Reposted by Ragnar {Groot Koerkamp}
sachy.dev
The LLMs are just reproducing the code from those dependencies, now without attribution nor contribution to the source, because that's what trained the LLMs in the first place.

It's still dependencies, just with extra steps and more liability.
brianleroux.bsky.social Brian LeRoux @brianleroux.bsky.social · 3d
Working with LLMs has galvanized my view on dependencies.

Most tasks never needed external packages. But now we can generate vanilla solutions in mere seconds.

Deps were always convenience over necessity. Now they're not even convenient.
February 14, 2026 at 8:52 PM
Reposted by Ragnar {Groot Koerkamp}
stevensalzberg.bsky.social
So now the US is spending the enormous sum of $38 billion to build concentration camps? ICE is calling them "detention centers" but we know what they really are. I cannot believe my country would do this. Everyone should speak out against it wapo.st/3OfarSk
ICE plans to spend $38.3 billion turning warehouses into detention centers
ICE plans to spend $38.3 billion converting warehouses into detention centers, according to planning documents, more than the annual budgets of 22 states.
wapo.st
February 14, 2026 at 4:07 PM
Reposted by Ragnar {Groot Koerkamp}
stephenturner.us
Theseus: Fast and Optimal Affine-Gap Sequence-to-Graph Alignment www.biorxiv.org/content/10.6... 🧬💻🧪 github.com/albertjimene...
Theseus: Fast and Optimal Affine-Gap Sequence-to-Graph Alignment
Motivation: Sequence-to-graph alignment is a central problem in bioinformatics, with applications in multiple sequence alignment (MSA) and pangenome analysis, among others. However, current algorithms...
www.biorxiv.org
February 14, 2026 at 5:00 PM
Reposted by Ragnar {Groot Koerkamp}
tomjanmeeus.bsky.social
Must read (geen betaalmuur) www.nrc.nl/nieuws/2026/...
Opinie | We zijn in acuut gevaar, maar blijven wensdenken over de VS
Geopolitiek: Amerika heeft zich niet alleen van ons af, maar tégen ons gekeerd, ziet Joris Luyendijk. Maar om hem heen doen de meesten alsof er uiteindelijk niets aan de hand is.
www.nrc.nl
February 14, 2026 at 10:52 AM
Reposted by Ragnar {Groot Koerkamp}
lauriewired.bsky.social
CPUs are getting worse.



We’ve pushed the silicon so hard that silent data corruptions (SDCs) are no longer a theoretical problem.



Mercurial Cores are terrifying because they don’t hard-fail; they produce rare, but *incorrect* computations!
February 13, 2026 at 8:11 PM
Reposted by Ragnar {Groot Koerkamp}
zezunja.bsky.social
‘Zo kregen weblog Geenstijl, opinieprogramma Nieuws van de Dag en zanger Dries Roelvink accreditatie, maar Left Laser en BOOS niet (meer).’

Dries Roelvink en extreemrechtse media wel. Wat een wereld.

www.villamedia.nl/artikel/reda...
Redacties Left Laser en BOOS niet welkom in beveiligde deel Tweede Kamer
De journalisten van Left Laser krijgen geen toegang tot het beveiligde gedeelte van de Tweede Kamer, aanvragen voor een dagaccreditatie worden structureel door de persafdeling van de Tweede Kamer afge...
www.villamedia.nl
February 13, 2026 at 8:06 PM
Reposted by Ragnar {Groot Koerkamp}
camillemrcht.bsky.social
"They go to the breeding gulags.” what the heck did I just watch
sturdyalex.bsky.social Alex Andreou @sturdyalex.bsky.social · 4d
And just like that, Gilead was no longer dystopian fiction.

“Just like Hitler imprisoned Gypsies, Jews, communists, you know, all of his political rivals, we have to do the same thing with women… So they go to the gulag first. They go to the breeding gulags.”

www.mediamatters.org/nick-fuentes...
Nick Fuentes: “The number one political enemy in America is women. … They have to be imprisoned.”
www.mediamatters.org
February 13, 2026 at 7:28 PM
Reposted by Ragnar {Groot Koerkamp}
katemckean.bsky.social
Over on threads someone just use ai;dr and we all need to adopt that right quick
February 11, 2026 at 7:56 PM
Reposted by Ragnar {Groot Koerkamp}
ccanonne.github.io
The list of accepted papers at #STOC2026 is out:
acm-stoc.org/stoc2026/acc...

Congratulations to all authors!
STOC 2026 - 58th ACM Symposium on Theory of Computing
acm-stoc.org
February 13, 2026 at 1:27 AM
Reposted by Ragnar {Groot Koerkamp}
bcmerchant.bsky.social
At Google, speech critical of ICE is censored from the company's internal platform, employees who speak out are served with warnings, and leadership refuses to acknowledge a petition that's now gotten 1,200 employee signatures.

Inside the worker-led push to get ICE out of Google:
Google is stifling anti-ICE speech in the workplace as 1,200 employees call on the company to cut ties
The search giant has made it harder and riskier to protest its policies. Workers are still fighting back. Plus: A community takes down a data center, QuitGPT, and the 5 worst Super Bowl AI ads.
www.bloodinthemachine.com
February 12, 2026 at 7:23 PM
Reposted by Ragnar {Groot Koerkamp}
jermp.bsky.social
Just accepted! dl.acm.org/doi/10.1145/...
Modern Minimal Perfect Hashing: A Survey | ACM Computing Surveys
Given a set S of n keys, a perfect hash function for S maps the keys in S to the first m ≥ n integers without collisions. It may return an arbitrary result for any key not in S and is called minimal i...
dl.acm.org
February 12, 2026 at 3:03 PM
Reposted by Ragnar {Groot Koerkamp}
bede.im
Thanks all who listened and asked questions, and @pha4ge.bsky.social for hosting. Now on YouTube.

Pro tip: don't create a 6GB file during a live demo with only 5GB free 😂
pha4ge.bsky.social pha4ge.bsky.social @pha4ge.bsky.social · 6d
Don’t miss Episode 16 of #GenomicHorizons 🧬

Join us on 12 Feb 2026 at 14:00 –15:00 UTC as @bede.im presents: “Practical panhuman host sequence depletion for responsible data sharing.”

Register:
us06web.zoom.us/webinar/regi...

Live stream:
www.youtube.com/live/FRuj2Wm...

See you there!
February 12, 2026 at 4:45 PM