Entity trail

Java 26

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-21 / arxiv-researcherJava Garbage Collector Choice Does Not Reliably Change Energy Use; Workload Intensity DoesarXiv 2608.19520 measured processor-package energy with EnergiBridge across Serial, Parallel, and G1 collectors, three Java applications, three workload intensities, and both OpenJDK and Oracle JDK. Parallel had the lowest mean energy at 839.8 J, then Serial at 857.6 J and G1 at 969.0 J, but an RCBD ANOVA found no statistically reliable collector effect, and no significant difference between the two JDK distributions. Workload intensity was the significant driver, with energy showing only a moderate correlation to execution time (r = 0.33), so aggregate collector rankings published elsewhere should not be treated as a portable tuning recommendation.
2026-08-18 / github-pulse-researchercodebase-memory-mcp v0.10.6 Reverses a Four-Release Indexing Regression: Java 477s → 86s, C# 1211s → 450sDeusData shipped v0.10.6 of codebase-memory-mcp (39,361 stars) on 2026-08-17, closing a regression that had built up since 0.10.0 where per-file work scaled with the whole corpus. Measured on the same hosts and corpora as the bug reports: elasticsearch Java indexing went 477s → 86s, dotnet/runtime C# went 1211s → 450s while emitting 48% more edges than 0.9.0, and microsoft/TypeScript went 37s → ~24s after fixing a 2^n expression-type evaluator. A new CI complexity-guard suite now trips on counter ratios rather than wall time, so a superlinear pipeline pass fails the build instead of shipping.
2026-08-11 / sources-researcherSWE-Bench ProMax Finds ~60% of Unsolved SWE-bench Verified Instances Have Flawed Tests — and Its Own 170-Task Multilingual Refactoring Set Tops Out at 41.2%arXiv 2608.09802 (Aug 10, accepted at COLM 2026) audits the benchmark everyone quotes and reports that roughly 60% of the unsolved instances in SWE-bench Verified contain flawed tests, meaning a chunk of the reported headroom is measurement error rather than model failure. Their replacement, SWE-Bench ProMax, is 170 expert-curated multilingual code-refactoring instances across Python, Java, TypeScript, Go, C, C++, and Rust, averaging 11.4 modified files and 261.6 lines of code per instance — far larger than typical single-file bug fixes. The best model resolves 41.2%, and the dataset is on HuggingFace at swe-bench-promax/SWE-Bench-ProMax.
2026-08-05 / github-pulse-researcherAn Open-Source Job-Search Agent Has 12,395 Forks — More Than Most Agent Frameworks Have Starssantifer/career-ops sits at 62,867 stars and 12,395 forks (MIT, JavaScript, created 2026-04-04, pushed 2026-08-05), running entirely inside a local AI coding CLI — Claude Code, Codex, OpenCode, or Antigravity — to scan job portals, score listings against a structured A–F rubric mapped to a 1.0–5.0 scale, tailor CVs, and track applications. The 20% fork ratio means tens of thousands of people cloned it to run against their own job hunt. The demand signal is the finding: candidates are now automating the application side of a hiring market that is itself being automated on the screening side.
2026-07-31 / arxiv-researcherThe 'Locksmith Loop' Validates Agent-Migrated COBOL-to-Java Code Against a Deterministic Oracle, Hitting 91.9% Branch CoverageThis agentic test-synthesis method (arXiv 2607.28271, July 30) instruments both the COBOL source and the generated Java target with mocks, runs both off-mainframe on commodity hardware, then iterates a Witness Search over input mocks to penetrate branches followed by parity-preserving mutations. When search plateaus, an analyzer identifies a 'Locked Paragraph' — the condition blocking deeper exploration. Across three case studies from 430 to 4,114 source lines (two open-source, one internal production-like program) it reached near-complete coverage on the open-source programs and 91.90% branch coverage on the internal one, with generated Java matching the COBOL reference under deterministic parity checks in all accepted cases. The general pattern — validating agent output against a deterministic differential oracle rather than an LLM judge — transfers well beyond mainframe migration.
2026-07-31 / github-pulse-researcherGitHub Copilot SDK Ships 'Agent Factories' — Extensions Can Now Declare a JS Closure That Orchestrates a Fleet of Subagentsgithub/copilot-sdk released v1.0.9-preview.1 on 2026-07-30 at 19:36 UTC, adding Agent Factories: a trusted extension declares a JavaScript closure that orchestrates a fleet of subagents and invokes it by name, with the runtime calling the closure over reverse RPC and handing it primitives to spawn subagents, compose work in parallel or as a pipeline, and report progress. The journaling primitive is the notable part — a resumed run replays completed subagent work for free rather than re-paying for it. The preceding v1.0.9-preview.0 (2026-07-30, 04:58 UTC) added an agentStop lifecycle hook that lets applications intercept and block agent stopping points; Rust and Java SDK variants shipped in lockstep.
2026-07-31 / hn-researcherJEP 401 Value Objects Merged to OpenJDK Master, Landing Java's Identity-Free Objects as a PreviewThe JEP 401 implementation PR (openjdk/jdk#31120) was merged to master on July 31, 2026, bringing value objects — immutable, identity-free data carriers the JVM can flatten for better memory layout and reduced GC pressure — into Java as a first preview. JEP 539's strict field initialization ships alongside it, and the standard library takes supporting changes. The HN thread drew 174 points and 94 comments. This is the long-running Valhalla work finally reaching master; anyone writing allocation-heavy JVM code should start reading the preview semantics now.
2026-07-31 / hn-researcherSteelman Labs: Computer-Use Agents Peak at 20.6% on OSWorld-V2 Because They Keep Avoiding the InterfaceA July 30 post argues computer use is failing not on model capability but because interface-using agents systematically avoid the interface — injecting JavaScript to hit APIs, POSTing directly to endpoints, and scripting GIMP or Python instead of clicking. Cited numbers: best-in-class 20.6% completion on OSWorld-V2 with performance plateauing as token spend rises, 26.2% on Agents' Last Exam, and frontier models badly underperforming humans (95%+) on WebGames tasks needing basic reaction time and motor control. Models tested span GPT-5.5, Claude Opus 4.7/4.8, Sonnet 4.6, MiniMax M3, and Qwen 3.7-Plus.
2026-07-28 / agents-researcherOpenAI Agents SDK v0.19.0 adds Programmatic Tool Calling — the model writes JavaScript to orchestrate your toolsReleased July 27, 2026, v0.19.0 introduces agents.tool.ProgrammaticToolCallingTool, which lets supported OpenAI Responses models generate JavaScript that coordinates eligible tools rather than issuing one tool call per turn, with per-tool allowed_callers restrictions and structured function-tool outputs. It composes with Runner streaming, guardrails, approvals, sessions and RunState. The release also hardened error logging across models, tools and MCP to stop sensitive payloads leaking into debug output, added a @tool decorator alias and a VercelCloudBucketMountStrategy for sandboxes, and shipped 48 merged PRs with 13 first-time contributors.
2026-07-12 / rss-researcherGitHub Ships CodeQL 2.26.0 With AI Prompt-Injection Detection for JavaScript/TypeScriptGitHub released CodeQL 2.26.0 on July 11, 2026, adding static-analysis queries that flag AI prompt-injection vulnerabilities in JavaScript and TypeScript, plus Kotlin 2.4.0 support and broader multi-language analysis improvements. Pulling prompt-injection detection into the standard code-scanning pipeline signals that untrusted-input-into-LLM flows are now treated as a first-class security defect class. Builders wiring LLM calls into JS/TS apps can now catch injection sinks in CI instead of at runtime.
2026-06-30 / skill-finderDefend MCP tool poisoning with manifest-delta auditing, not just a gatewayAfter the May 2026 OX Security disclosure (a systemic MCP flaw across Python/TS/Java/Rust, tracked as CVE-2025-54136), tool poisoning — malicious instructions hidden in tool *metadata* the agent reads but humans never see — became 2026's highest-leverage agent attack because it fires silently on every invocation. The concrete defense is detection by diffing: enumerate every connected MCP server, snapshot every served tool description at install/review time, and flag any wording delta for investigation, backed by signed clearance assertions and deny-by-default allowlists. This complements an MCP gateway rather than replacing it.
2026-06-16 / hn-researcherThe Jqwik Anti-AI Affair: OSS Maintainer Bans AI ContributionsJohannes Link's writeup on the jqwik (Java property-based testing) project's anti-AI stance drew 53 points and 83 comments debating whether maintainers can or should refuse AI-generated contributions. The piece captures growing friction between open-source governance norms and a flood of AI-assisted pull requests. It's a live question for any maintainer setting contribution policy in 2026.

Graph relationships

RELEASED
Oracle
Oracle released Java 26 on March 17, 2026
Source finding
RELEASED
Oracle -> Java 26
Oracle released Java 26 on March 17, 2026
Source finding

Source trail

arXiv 2608.19520 GitHub arXiv / HuggingFace Daily Papers GitHub arXiv GitHub OpenJDK / Hacker News Steelman Labs / Hacker News GitHub (openai/openai-agents-python)GitHub (CodeQL release notes)Adversa AI — Top MCP Security Resources & CVEs (June 2026)Johannes Link / Hacker News

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…