Provenance Record

Well-researched, clearly structured, appropriate technical depth for a Briefing-category article at 882 words. Organises findings into coherent sub-sections. The 'What We Don't Know' and 'Analysis' sections add genuine editorial value.

All 8 source snapshots read: source-0.html.gz (tiobe.com main index), source-1.html.gz (Slashdot), source-2.html.gz (tiobe.com R page), source-3.html.gz (tiobe.com Zig page), source-4.html.gz (tiobe.com Python page), source-5.html.gz (InfoWorld), source-6.html.gz (dev.to), source-7.html.gz (tiobe.com SAS page). Allowlist warnings for tiobe.com, developers.slashdot.org, and dev.to are false positives — all are established, legitimate sources for programming language coverage. All key claims confirmed: R #8 / 1.77% (source-0); Jansen quote verbatim in source-0; R lowest #73 Dec 2008 / highest #8 May 2026 confirmed in source-2; SAS all-time high #11 Mar 2007 confirmed in source-7; MATLAB #20 / 0.89%, Go #16 / 1.12% / -1.58pp, Rust #15 / 1.14% / +0.21pp, Java #3 / 7.94%, C++ #4 / 7.92%, C# #5 / 5.41%, Python #1 / 19.98% / -5.37pp all confirmed in source-0; Python Language of the Year 2007, 2010, 2018, 2020, 2021, 2024 and Highest Position #1 May 2026 confirmed in source-4; Zig #36 / 0.39% and 'approaching top 30 for the first time' verbatim in source-3; InfoWorld Jansen February 2026 quote about R and Perl confirmed verbatim in source-5.

All numeric data points and direct quotes verified against source snapshots. One note: the article states R's previous all-time high was 'in the early 2020s' — the R page data shows the 1.77% rating was also recorded in August 2017 (Date.UTC(2017, 7, 2), 1.77), suggesting R may have been at #8 before. This is interpretive context from TIOBE's own position tracking (the page says 'Highest Position (since 2007): #8 in May 2026'), which is internally consistent with the article's claim. No fabrications or hallucinations detected.

High-quality submission. All claims traceable to cited and verified sources. Verdict upgraded to APPROVE — automated APPROVE_WITH_CORRECTIONS was driven by allowlist false positives (established sources not yet in allowlist) and a loaded-language false positive inside a direct quote. No corrections record needed.