TooMuchCoding Newsletter
If you’d like new issues by email, subscribe below (unsubscribe anytime):
Issue #34
Third-party cyber evaluations in OpenAI models, TeamCity RCE vulnerability, JDK 27 early-access features, and more.
•
4 min read
If you’d like new issues by email, subscribe below (unsubscribe anytime):
Third-party cyber evaluations in OpenAI models, TeamCity RCE vulnerability, JDK 27 early-access features, and more.
An autonomous AI agent escaped its sandbox. Three major model releases landed. Cisco firewalls are getting hacked. Also, OpenTelemetry profiles are live.
xAI sues users over Grok CSAM, Spring Boot 3.x reaches end of life, and OpenAI's GPT-Red tries to patch itself. A wild week in the JVM ecosystem.
Spring Cloud Contract moves to Stubborn.sh, Apple sues OpenAI, CISA adds critical CVEs, and JDK 27 early-access arrives.
Claude Sonnet 5 lands as Anthropic's most agentic Sonnet yet, Fable 5 returns under cybersecurity limits, plus actively-exploited SharePoint and SimpleHelp CVEs - a quiet holiday …
JDK 27 defaults to G1GC, Spring 3.5.x hits EOL, Google loses more AI talent to Anthropic, and Cisco zero-days are in the wild. Plus the usual security chaos.
Weekly newsletter for senior JVM/Java developers. This week: government blocks AI models, FortiBleed hits 86k devices, and Project Valhalla finally ships.
Issue #26 of TooMuchCoding Newsletter for senior JVM/Java developers. Anthropic cancels Fable 5 and Mythos 5 for everyone, CISA goes full panic mode with 3-day patch deadlines, and …
Recursive self-improvement, HTTP/2 bombs, JDK 27 rampdown, and why your observability stack needs to get smarter.