@Anil Madhavapeddy
Professor of Planetary Computing, University of Cambridge
Relays (1)
- wss://relay.ditto.pub – read & write
Recent Notes
The directed graph of voting results for LLM usage is very Debian (and perhaps something applicable to mainstream voting on complex issues!) https://vote.debian.org/~secretary/gr_llm/
We'll need a few more zeros on that compensation line!
The 3rd Programming for the Planet talk videos are now live, with notes on https://anil.recoil.org/notes/pldi26-propl and a fediverse mirror at https://watch.eeg.cl.cam.ac.uk/c/propl24/videos. enjoy!
I feel two out of three ain't bad for me on the Dolly Parton advice scale: "Never ignore your roots, your home, or your hair”
Hi nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq0ujnx96rzj2m7kl9yqhyplqhxuf55el43gdqw8xxpm7pdk3ph6gqwfaz86,
IMO #responsibility is with the #operators. They can mitigate, monitor, patch and ...
@nprofile1q... I think the software supply chain is so complex and intertwined now that they’d struggle to figure out who are what to fund! A success disaster of the commons
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq0ujnx96rzj2m7kl9yqhyplqhxuf55el43gdqw8xxpm7pdk3ph6gqwfaz86 Here is my pov as a maintainer.
I agree with a lot of your options, but I think they ...
@nprofile1q... I agree with you that nothing has structurally changed, but the ability to chain multiple exploits within seconds seems like a new capability...
I've had to respond to multiple OSS security issues recently and the wild thing is that agents can now generate exploits just on the *rumour* of a bug. This throws security embargoes out the window, as the fix is less important than the knowledge of its existence. What on earth are open source maintainers supposed to do next? https://anil.recoil.org/notes/rumour-is-the-exploit
4
nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqhxatvn8sm87kkhrnqku097p5ddqhmhwtn385v3are3gk208p3dqqpvlv9g nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq0ujnx96rzj2m7kl9yqhyplqhxuf55el...
@nprofile1q... @nprofile1q... @nprofile1q... I think this 'ePDF' is a rendering of the PDF and not HTML; it doesn't reflow on resize for example. I wonder where the TAPS HTML5 output goes, I cant find any papers in the DL that expose it.
1
I'm delighted to announce that our formerly plain PDF ICFP 2026 paper (https://anil.recoil.org/papers/2026-package-calculus.pdf) has been UPGRADED with advanced AI summaries (if you pay up for the premium ACM account) and is under impenetrable cybersecurity protection (if you click through painful captchas and wait 10 seconds) and is way slower due to being hidden behind a web-based epdf reader rather than using your native PDF client. https://dl.acm.org/doi/10.1145/3828699.
1
Connecting the dots for global biodiversity action; two big perspectives out in @nprofile1q... this week lead by Bill Sutherland and Andy Gonzalez from the RS/NAS biodiversity forum last summer. My thoughts: https://anil.recoil.org/notes/nas-rs-biodiversity-papers
Seeing so many LLM-based spam attacks hitting my elderly relatives now. And meanwhile, I have to explain why the 'Apple Store' is not the same as the 'App Store'. We've collectively made a hash of partitioning trust in mobile devices, while making them a centrepoint of identity online
[blog] Packaging Dart binaries as .deb and .rpm etc.
nFPM makes it very easy to put your binaries into a Debian .deb or RedHat Package Manager .rpm file. I also took the opportunity to slide in the n...
@nprofile1q... sometime this century I gotta get this working for OCaml