Julien Roncaglia

@vbfox.net

Software developer 🇫🇷 🇪🇺 Mostly on Mastodon, https://hachyderm.io/@vbfox (Bridged at @vbfox.hachyderm.io.ap.brid.gy)

On one end, the Anthropic team is a massive user of AI to write code (80%+ of all code deployed is written by Claude Code). They ship amazingly fast. On the other hand, seeing these beyond terrible reliability numbers suggests there might be a downside to all this speed:

Bild

Oui, la canicule de 1976, c’était du pipi de chat. Un épisode aujourd’hui ridiculisé par la violence et la répétition des canicules de ces dix dernières années. [1/4]

Bild

Les États-Unis cette nuit, en violation de tout droit international, ont bombardé l’Iran. Quel sera le prix en retour de ces bombes ? Trump et Netanyahou, deux dirigeants d’extrême-droite, sont en train de faire basculer le monde face à une Europe muette et paralysée.

Even accepting the premise that AI produces useful writing (which no one should), using AI in education is like using a forklift at the gym. The weights do not actually need to be moved from place to place. That is not the work. The work is what happens within you.

TypeScript team: rewrites compiler in Go. Go community: what do you mean your new compiler takes more than a minute to compile? Unacceptable. Dishonorable even. We are so sorry for this sub par experience. Not how we do things around here. Two days later: WIP 5x speedup. HN: why pick Go anyway?

cmd/compile: slow escape analysis in large package in the typescript compiler · Issue #72815 · golang/go

Go version go version go1.24.1 linux/amd64 Output of go env in your module/workspace: AR='ar' CC='gcc' CGO_CFLAGS='-O2 -g' CGO_CPPFLAGS='' CGO_CXXFLAGS='-O2 -g' CGO_ENABLED='1' CGO_FFLAGS='-O2 -g' ...

github.com

Took me a while to see it, but it’s a simple vulnerability. The API takes messages as hex strings like “AB12” and passes them to the bignum library. That library supports negative numbers like “-AB12”. However, the sign is dropped when calculating the nonce. Different messages, same nonce 💥

Visual Studio 2022's splash screen shows up approx. 100ms later than Visual Studio 2013 manually testing it. This was confusing to me after I traced it, as relative to their main method entry point, they both take about the same amount of time to render and show the window. What was going on?

Visual Studio 2022 splash screen.