tanuki-cd.com

The static website for the TANUKI-CD research program.

The site intentionally uses plain HTML and CSS. It has no build step and no client-side JavaScript. Serve this directory with any static web server.

Pages

• index.html — program overview and recent work
• research/index.html — papers, preprints, research interests, and related literature
• code/index.html — software projects and repository policy
• about/index.html — program name, methodology, people, and AI responsibility
• ai/index.html — FAQ on generative AI use in mathematics research

Subpages use directory URLs such as /research/, /code/, /about/, and /ai/. Requests without a trailing slash resolve to the same pages. The custom error page is stored in 404/index.html.

Updating a publication

Preprints and published papers use one record in research/index.html. Copy the existing publication-card structure, then give the article a stable id that will not change when the paper is published.

When a preprint is published:

1. Change the status class and label from publication-status--preprint to publication-status--published.
2. Add the journal citation, publication date, and DOI.
3. Keep the preprint URL and any code or data links.
4. Update the corresponding entry in the Recent Work section of index.html when it is among the three newest records.

Updating a project

Project cards live in code/index.html. The project name is the primary stretched link and must point to the canonical repository on git.tanuki-cd.com. Secondary paper, documentation, and GitHub-mirror links must use the project-secondary-link class so they remain independently clickable.

The TANUKI-CD forge is authoritative. GitHub is a mirror.

Updating members

Member cards live in the People section of about/index.html. Store portraits locally under art/, provide descriptive alternative text, and avoid time-sensitive biography wording when a stable description is possible.

Local preview

python3 -m http.server 3000

Then visit http://localhost:3000/.

Production deployment

The public site uses https://tanuki-cd.com/ as its canonical address. The nginx example in deploy/nginx.conf.example serves the static pages, redirects HTTP and www requests to that address, and displays 404/index.html for missing pages while retaining HTTP status 404.

The example also redirects the old public .html URLs to their directory URLs, preserving query strings. These redirects require the updated nginx configuration.

Pushing this repository does not install the nginx configuration. Apply it through the server's deployment process, setting its document root and existing TLS certificate paths first. The certificate must cover both hostnames. Preserve the server's existing certificate-renewal configuration and other virtual hosts. Check the resulting configuration with nginx -t before reloading nginx.

Create an archive containing only the public site files from the repository root:

tar -czf /tmp/tanuki-cd-site.tar.gz \
  index.html research/index.html code/index.html about/index.html ai/index.html 404/index.html \
  styles.css robots.txt sitemap.xml LICENSE \
  art/fano.svg art/tanuki-banner.png art/tanuki-banner.webp \
  art/quaternionic-icon.png art/fastcd-icon.svg art/algebraity.webp \
  assets/fonts/lmroman10-bold.otf assets/fonts/GUST-FONT-LICENSE.txt

Extract that archive into the configured document root. The nginx worker needs read access to these files and traversal access to their parent directories.

After deployment, verify:

• The homepage, four inner pages, stylesheet, images, font, robots.txt, and sitemap.xml return 200 over HTTPS.
• HTTP and HTTPS www requests redirect to https://tanuki-cd.com/, preserving the requested path and query string.
• A missing path returns 404 and renders the custom Page not found page.
• The Open Graph image is publicly reachable and a shared link displays the intended title, description, and image.