Modern LaTeX Editions of Public-Domain Mathematics Manuscripts

This record is the public front page for an ongoing open project to produce modern, inspectable LaTeX editions and translations of older mathematics and physics manuscripts. Public-domain scans and source/reference materials are collected, transcribed into TeX, checked into readable PDFs, organized by author/work or coherent corpus, and published with source artifacts so readers can inspect, correct, fork, and continue the work.

Landing file surface: The main record is a landing and preservation surface. Older broad Cayley and Weber direct-reader drafts have been demoted from the latest file list because cleaner author/corpus records now provide the front-facing reader surfaces. The material remains preserved through the full repository preservation ZIPs and prior Zenodo versions.

GitHub mirror: Forkable reader PDFs, source TeX where practical, and current manifests are mirrored at https://github.com/KokunoYumeto/modern-latex-manuscripts. Zenodo remains the archival DOI surface and large-file/provenance backstop.

Workflow and replication packet. The direct file Project Workflow and Replication Notes.pdf summarizes the practical pipeline used here: ChatGPT web project sessions help identify and repair works; Codex handles local indexing, scripting, auditing, Zenodo publication, and GitHub mirroring; Kimi K2.6 agent swarms have supplied large first-pass TeX transcriptions; ChatGPT Pro sessions perform review, repair, and translation; and Claude Code has been used as a local LaTeX-repair and translation worker. The current standalone workflow record (10.5281/zenodo.20461174) and the accompanying Workflow Replication Packet - Scripts, Tooling Notes, and Manifests.zip contains sanitized example scripts, tool lists, metadata templates, OCR notes, and operational notes for people who want to reproduce or adapt the workflow.

Raw/source/provenance role. This landing record is also the stable bulk-preservation lane for the project. Large raw source drops, page-image working archives, superseded drafts, full-repository preservation chunks, workflow notes, and broad provenance bundles are kept here under the all-version concept DOI 10.5281/zenodo.20393488. Author and corpus satellite records should stay reader-facing and comparatively lean, linking back here instead of re-uploading multi-gigabyte raw archives on every update.

Current project map.

• Main corpus and bulk preservation record: [##########] 100/100 this landing record remains the broad cumulative surface, with reader PDFs, artifact ZIPs, manifests, and full-repository preservation material. Latest version: this Zenodo record page; concept DOI: 10.5281/zenodo.20393488; earlier broad cumulative record: 10.5281/zenodo.20410262.
• EGA working English translation and source package: [###-------] substantial working translation and source package; [##########] French originals present EGA I-II are largely inherited from the community translation. Current project additions include EGA 0_IV sections 15 through 23 and EGA IV sections 1 through 21 as substantive working translations. The current record includes a broad EGA working reader, a separate 514-page EGA IV sections 1-21 working reader, all eight NUMDAM French originals as direct top-level PDFs, and TeX/source packets in ZIP artifacts. All-version record: 10.5281/zenodo.20414353.
• SGA working English translation: [#####-----] corpus-wide working translation; active strict SGA 5/6 stream Current SGA record with coherent cumulative readers, French references, source artifacts, and strict source-checked working translations. Latest published record: 10.5281/zenodo.20488197; all-version record: 10.5281/zenodo.20410947.
• Non-European and multilingual mathematical manuscripts: [#########-] reader-facing multilingual release; bulk provenance on this landing record Consolidated release for Chinese, Indian/Sanskrit, Islamic/Arabic, and historical-reference mathematical works. The satellite is now intentionally lean: readable work-level PDFs and compact current TeX/work packets live there, while large raw archives and broad provenance remain on this landing/preservation DOI. Latest published record: 10.5281/zenodo.20488731; all-version record: 10.5281/zenodo.20410957.
• Chinese mathematical classics: [########--] 80/100 focused corpus slice for Chinese mathematical works including Jiuzhang Suanshu / Nine Chapters, Li Ye, Qin Jiushao, Sunzi, Yang Hui, and Zhu Shijie material, with language-separated readers and source artifacts. Current record: 10.5281/zenodo.20435670.
• Indian and Sanskrit mathematical classics: [########--] 80/100 focused corpus slice for Aryabhata, Bhaskara II, Brahmagupta, and related Sanskrit/Indian mathematical works, with original-language drafts, translations, and source artifacts. Current record: 10.5281/zenodo.20435677.
• Islamic and Arabic mathematical texts: [########--] 80/100 focused corpus slice for al-Kashi, al-Khwarizmi, al-Tusi, Omar Khayyam, Robert of Chester/Karpinski, Rosen, Ruska, and related Arabic/Islamic mathematical works, with original-language drafts, translations, and source artifacts. Current record: 10.5281/zenodo.20435687.
• Historical reference texts for non-European mathematics: [#######---] 70/100 focused reference slice for Said al-Andalusi, al-Muqaddasi, Ibn al-Nadim, Ibn al-Qifti, Smith-Karpinski, and related contextual reference texts used for terminology and historical comparison. Current record: 10.5281/zenodo.20435690.
• Heinrich Weber: [######----] broad draft available; literal repair stream active through Volume I section 107 Author page with source-checked German/English repair stream for Lehrbuch der Algebra, broad older drafts, TeX sources, component PDFs, and provenance artifacts. Latest published record: 10.5281/zenodo.20488622; all-version record: 10.5281/zenodo.20412153.
• Emmy Noether: [#####-----] paper-level high-fidelity edition in progress through Papers 41-42 Author page with cumulative German modern-LaTeX and English translation readers, paper-level German/English pairs, TeX sources, scans, and build reports. Latest published record: 10.5281/zenodo.20488499; all-version record: 10.5281/zenodo.20412587.
• Deligne papers: [##########] 90/90 original-language reference PDFs preserved; English/French working translations continue Dedicated Deligne record with original-language references, English/French working drafts where available, paper-level sidecars, render checks, and provenance material. Latest published record: 10.5281/zenodo.20487802; all-version record: 10.5281/zenodo.20410853.
• Carl Friedrich Gauss: [#######---] 3,000+ pages of modern-LaTeX working drafts available; detailed proofreading continues dedicated Gauss author page with cumulative Werke band readers and source artifacts. The current local pipeline has addressed several known mathematical transcription defects and preserves source material for continued checking. All-version record: 10.5281/zenodo.20410934.
• Bernhard Riemann: [#######---] 70/100 author page for selected mathematical papers and a broader Gesammelte Werke complete-draft reader, with TeX/source/provenance artifacts. Current record: 10.5281/zenodo.20434317.
• Cayley, Dedekind, and Dirichlet: classical algebra and arithmetic shelf: [###-------] Cayley front-facing source-checked slice readers; [#########-] broader source material preserved focused shelf for modern LaTeX working drafts by Cayley, Dedekind, and Dirichlet. The current record promotes 13 Cayley volume-level source-checked slice readers assembled from the current source-checked pickup ranges. The artifact ZIP preserves the broader Cayley source inventory, including rendered slice PDFs, TeX chunks, source trees, and audit material. Dedekind Bands I-III and one Dirichlet selected-works grouping remain available as working drafts. All-version record: 10.5281/zenodo.20414787.
• Minkowski, Hecke, Landau, Sylvester, Steinitz, Hensel, Oka, Hausdorff, Grassmann, and Killing: [######----] 60/100 author-cluster record for cleaner reader PDFs and source/provenance artifacts. The Landau reader is the English Goodman translation of Elementary Number Theory; Sylvester collected-papers source scans for Volumes I and IV are preserved in the artifact ZIP for future modern LaTeX typesetting. All-version record: 10.5281/zenodo.20411006.

Workflow provenance. Review sessions identify and prioritize older public-domain or otherwise suitable works for transcription and translation; readers are welcome to suggest works or point to better public PDFs/scans. Local archive tooling downloads and indexes scans and source packets, prepares working folders, audits files, organizes public releases, and publishes through the Zenodo API. Large-scale machine-assisted transcription systems convert scans to TeX drafts, which are preserved as source/provenance artifacts. Review passes check, repair, and translate the TeX. Public organization is by author, work, or coherent corpus; reader-facing files use author/work/corpus titles, while provenance archives preserve source material for reconstruction.

Full repository preservation update. The current working inputs and source packets are included directly on this already-linked original/main record as ten chunked ZIP files, plus a manifest and README. There is no separate public raw-source DOI. These are provenance/backstop files for reconstruction and continuation; the reader-facing records remain organized by author or corpus.

File organization. Top-level PDFs are reader-facing files intended for browsing and direct download. ZIP files contain TeX/source material, source scans or reference material where appropriate, OCR text, provenance, manifests, audit outputs, and repair queues. Records are being split by author or coherent corpus while this main record remains the single landing page and bulk preservation surface.

Coherence note. Redundant records are being handled conservatively: cleaner author/corpus pages become the recommended public entry points, while older broad shelves are retained as preservation surfaces until their contents are fully superseded, checked, and linked. This avoids losing work while making the front-facing archive progressively easier to browse.

Quality note. These are working scholarly drafts, not final critical editions. Technical PDF audits check whether files open and extract text; they do not prove mathematical correctness. Some drafts still need layout repair, source comparison, theorem-number checking, and human proofreading. Availability is preserved while cleaner organized versions are built.