qselmer.github.io

Personal academic website of Elmer Quispe-Salazar, implemented with Quarto and deployed through GitHub Pages Actions.

Production status: the controlled migration from Academic Pages/Jekyll to Quarto was completed on 8 September 2026. The production branch is currently master; the final pre-migration Jekyll implementation remains preserved at legacy/jekyll-v0.9.

Information architecture

The website is a discovery and presentation layer. Scientific repositories remain the source of truth for code, analyses, manuscripts, software, and teaching materials.

• Research - scientific questions and thematic programmes linking canonical project repositories with selected papers, talks, theses, reports, and software.
• Publications - formal scholarly outputs; conference outputs are separated.
• Talks - conference talks, invited talks and seminars, posters, and workshops when those record types are present.
• Software - curated scientific packages and applications with explicit maturity.
• Teaching - structured courses/training and reusable teaching infrastructure.
• Posts - tutorials, technical notes, methodological explanations, reproducible-science guidance, and an RSS feed.
• CV - professional record and current PDF curriculum vitae.
• More - curated Data Sources, Miscellaneous material, and Contact.

Legacy routes that no longer belong to the public architecture are retained only as deterministic noindex redirects when old inbound links remain meaningful. The retired substantive pages are not rendered as public content.

Metadata sources

Canonical academic-profile metadata is maintained in qselmer/qselmer.

The website synchronizes only the public metadata required for presentation:

• assets/data/publications.json
• assets/data/research-metrics.json

The canonical repository catalogue is fetched transiently during automation because it can contain private-repository metadata. Only curated public subsets are written to:

• assets/data/software.json
• assets/data/teaching.json

Website-specific editorial decisions remain in local registries such as projects/registry.json, software/registry.json, teaching/registry.json, talks/registry.json, data/registry.json, and blog/registry.json (public label: Posts).

Rendered pages receive canonical URLs, Open Graph/Twitter previews, RSS discovery metadata for Posts, and schema.org JSON-LD for the researcher profile, journal articles, published software, the Git/GitHub course, and technical posts.

See AUTOMATION.md for the complete data flow.

Local validation

python scripts/build_publications.py
python scripts/build_conferences.py
python scripts/build_software.py
python scripts/build_teaching.py
python scripts/build_data_resources.py
python scripts/build_blog.py
python scripts/build_research_graph.py
python scripts/build_projects.py
python scripts/build_home.py
python scripts/validate_site.py source
quarto render
python scripts/validate_site.py rendered
python scripts/check_external_links.py _site

The source validator rejects obsolete Academic Pages/Jekyll runtime paths. The rendered validator checks identity consistency, deterministic generated fragments, CV delivery, compatibility redirects, metadata, and internal href/src targets.

Synchronizing academic-profile metadata

python scripts/sync_profile.py

This synchronizes canonical public profile metadata, derives the curated software/teaching subsets, rebuilds generated fragments, and runs source validation.

Production deployment

The production site is deployed through GitHub Pages Actions. .github/workflows/quarto-pages.yml rebuilds deterministic catalogue fragments, validates the source, synchronizes canonical visual assets, renders the Quarto project, validates the rendered site and external links, uploads _site, and deploys it to GitHub Pages on pushes to master.

The first Quarto production cutover was deployed from merge commit 0cf243eb7f68e67fe322bb1036e7ac10e97d4f33 by GitHub Actions run 34276777663.

Privacy and data policy

Restricted fisheries, biological, institutional, or private-repository metadata is not published through this website. The external Data Sources directory stores discovery metadata and provider links only; it does not mirror third-party datasets.

Legacy preservation

The final pre-migration Jekyll production snapshot is preserved at:

branch: legacy/jekyll-v0.9
commit: f84307690cc573b7b2e83a548d2d56102a05218d

Migration design, route parity, validation, and cutover evidence are documented in MIGRATION.md. Historical attribution is documented in NOTICE.md.