← Atlas

Feasibility, evidence and design

The request is feasible as two linked, evidence-aware course atlases. “Best location” is a sourced working identification with alternatives, not a claim of certainty. An archaeological settlement coordinate does not locate a battle, a river crossing, a cave or a narrated voyage.

Refined prompt

Build an evidence-aware atlas for each existing Caesar and Virgil course. Preserve all primary course content and assessment behaviour. Use the Strabo place–text connections and Carthage evidence panels as design references. Connect every Caesar lesson and every Aeneid book to mapped settings or explicit unlocated entries. Use Pleiades records with retained precision, provenance and snapshot checksums. Select preferred identifications from named archaeological institutions and published research; expose alternatives and explain what the evidence establishes. Distinguish settlements, river representative points, regional context, literary foundations, and event locations. Never invent excavation coordinates, ethnic boundaries, routes, ancient coastlines or historical voyages. Put supplementary scholarship in the evidence panel. Provide clean search/filter controls, equivalent keyboard-accessible place navigation, responsive map/reader layouts, and tablet portrait/landscape map focus for Virgil. Provide a browser-local comparison notebook with portable exports, without implying live collaboration or automatic grading. Deliver GeoJSON including null geometries, GeoPackage and a portable QGIS project. Validate source preservation, spatial/source integrity, offline map assets, device-size layouts, touch targets and course deep links. Report live-data and physical-device limitations honestly. Keep publication separate from local implementation.

Architecture and interface

Strabo’s harborscapes links places and readings; Carthage separates site navigation, evidence limits and comparisons. These patterns are adopted without copying their fixed-size layouts. Leaflet is already present in both reference builds and suits a small offline vector catalogue. OpenLayers offers more GIS/projection functions; MapLibre offers GPU vector styles for larger collections. Neither adds necessary value to this scope. All map libraries remain extensible behind a common GeoJSON evidence model.

The atlas has one map, one searchable directory and one reader with place, evidence and workshop tabs. At tablet portrait widths the reader moves below the map; landscape uses two panes where space permits. Map focus removes the reader temporarily while retaining the selection. Browser zoom remains available, touch controls are at least 44 pixels, ResizeObserver handles map resizing, and the directory provides keyboard access independent of map markers.

Data and scope

28 entries; 16 mapped representative points and 12 unlocated entries. All 28 course groups have a crosswalk. The Pleiades snapshot download date is unknown. Its SHA-256 is c24af658cfc8f26edd1f3a2ab40f497f43b21810a039db835eb233d6aad5a665. Its mtime is file metadata, not a publication date. Individual live pages returned an anti-bot challenge; no live refresh is claimed.

Natural Earth land and rivers are generalized modern context, not ancient shorelines or channels. No inferred routes, event tracks, ethnic territories or uncertainty-radius circles are drawn. Published scholarship is linked with short original summaries. The local Perseus Greenough edition audits supplementary Aeneid references; no additional Latin passages are republished. Name matches in the source are a lexical index, distinct from curated book-context associations.

Working together

Students and instructors can compare evidence statuses, follow the exact course reading, record observations and limitations, and exchange exported notebooks. Notes remain local and are not grades. Imported notes merge without deleting prior entries. Shared links retain place and course filters. Future server collaboration should reuse the existing authenticated course system rather than inventing a second account system.

Validation and release

See the atlas audit for structural, GIS and browser results and exact artifact fingerprints. Simulated tablet, phone and desktop viewports do not certify physical iPad Safari, Android browsers or every possible device. The map remains useful without external tiles. Public deployment has not been performed.

Published references consulted

Build references

Strabo harborscapes · Carthage