Repository navigation
test(D12): independently hash real MiKTeX negative TeX inputs from uploaded artifact - #65
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
D12 — real MiKTeX negative-input source provenance (exact SHA PASS)
PR #40 and preceding D fixes are already merged, so this is an additional D-only improvement to the SAME real MiKTeX Windows + dependent Linux CI, not an autonomous fixture. On the earlier accepted B20 run 41234368515, both negatives genuinely failed under MiKTeX, but their two TeX inputs were omitted from the 45-file uploaded evidence archive; the second-host auditor could only trust self-reported source hashes.
Fix: The existing Python producer archives
missing-library/missing-library.texandlua-only/graphdrawing-lua.texbyte-for-byte, recordssource_artifactand checks its SHA against the original source. The existing Linux post-upload auditor independently reads/re-hashes those exact bytes and requires the correct\\usetikzlibrary/\\usegdlibraryinputs, matching source SHA, true MiKTeX 25.12 negative logs, error diagnostics and zero PDFs. Real-artifact tamper tests reject deletion/modification of either input.Exact source HEAD:
79aa9891ca19aad1ef50ecf72ace8cddeeeecad9, basemain 9b4cd26b5b73f8011b975f9922cb208e36e12564; GitHub PR synthetic checkoutc4d5b1945f5f34b59795ea60067829330003acf2.All four exact-head workflows SUCCESS:
Downloaded and independently audited the original artifacts:
318071d93891893e3bd8da0c6975ab1885d677837755a2b5a51b506bd71f5a17(matches GitHub job upload digest).1f38a5ffcce1737ff484111b6c7a6b18c98c257d6af9f21fed18760b06975183.Independent local Poppler reopened all six real PDFs: 1 page, Cyrillic and mathematical Greek extraction, embedded DejaVuSerif/LatinModernMath, matching source/PDF/TeX-log SHA and actual MiKTeX compiler banner. Uploaded negative TeX input hashes independently recomputed:
dbc18fd34687623f192ddaadd8127c6fbd4731e8383f39b0fbf951dddc852e43183810c6032452d7918919f5eaf37a7843ffd2ddbfec15c8f1853d3b2782733dBoth negatives produce the expected errors, no PDFs; package inventory 9130/9130 byte-identical, official installer hash pin, private effective roots,
AutoInstall=0,--disable-installerconfirmed. New post-upload tamper suite 11/11 PASS, including negative-source deletion/corruption. N1 app Session→Rust CLI/direct XeLaTeX remains real programmatically-generated source; NOT mouse authoring.Only 4 D-owned
scripts/toolsfiles and new D-specificplans/andinsights/changed; A/B/C/E/F production files untouched. No modified test thresholds. Physical Win10 x64, Mac ARM M1/M4, egui OS pointer/F1/help/screenshots and GUI theme/PDF invariance: NOT_RUN. This evidence belongs only to the stated SHA; post-mergemainmust be checked separately.