Every investigation begins with questions rather than answers. This roadmap describes the directions in which Wide Open Windows is currently expected to grow. It is not a fixed development plan, but a record of intentions that remains open to revision as the project evolves.
At the same time, this page serves as a living development history. Each new version records not only what is planned next, but also what has been achieved. In that sense, the roadmap functions as both a roadmap and a changelog, documenting the gradual emergence of the project over time.
The long-term ambition is not merely to publish philosophical texts, but to develop a navigable landscape of experience in which books, essays, observations, perspectives, resources and reader reflections gradually become part of a single interconnected exploration.
These intentions are not promises. New questions may prove more interesting than existing plans, technical decisions may need to be reconsidered, and parts of the project may develop in unexpected directions. This page will therefore continue to evolve together with the investigation itself.
Current version: 0.5
Completed in August 2026.
Version 0.5 is the first stage at which the principal architecture of Wide Open Windows can be regarded as coherent and stable. The project is no longer merely a collection of interconnected pages. It now functions as a navigable digital environment with several complementary ways of moving through it.
The Rhizome remains the structural backbone, while Your Path, Explore, Recent Updates and Discover provide different forms of orientation. A visitor can search deliberately, retrace a recent route, follow conceptual relations or encounter less predictable connections without being forced into a single hierarchical menu.
The publication infrastructure is now largely automated. TheBrain remains the structural source, while the update pipeline rebuilds page metadata, the Rhizome, the search index, sitemaps, RSS resources, llms-full.txt and the dedicated WOW semantic JSON representation.
Structured JSON-LD is generated automatically for public pages. Books, essays, perspectives, observations, collections and project pages can therefore be represented in machine-readable form without maintaining a second manual metadata system.
The public read-only API is also operational. It exposes published pages, groups, search, complete semantic data and API health information through JSON endpoints, while remaining deliberately restricted to read-only access. It introduces no accounts, writing permissions or database-driven application layer.
Reader Reflections has become a site-wide private response mechanism. Readers can respond directly from content pages without entering a conventional public comment system. Submissions are moderated, arrive privately by email and are never published automatically.
A mobile audit has been completed as part of this version. The principal navigation, Rhizome, Explore, Discover, Paper Mode and Reader Reflections have been tested and adjusted for small screens. The project therefore remains usable as the same environment across desktop and mobile, rather than becoming a separate simplified mobile site.
Completed in 0.5
- Rhizome navigation integrated with the page structure generated from TheBrain.
- Your Path navigation for retracing recently visited pages.
- Integrated Explore search with generated search index, filters and related results.
- Automatically generated Recent Updates.
- RSS publication infrastructure for selected new and updated pages.
- Automatic JSON-LD generation for public pages.
- WOW semantic JSON as a machine-readable representation of pages, groups, concepts and rhizomatic relations.
- Public read-only API for pages, groups, search, semantic data and API health.
- Discover as an additional mode for encountering meaningful but less predictable pathways through the site.
- Site-wide private Reader Reflections with moderated SMTP delivery.
- Mobile audit and responsive refinements for the principal interfaces.
Version 0.6
Version 0.6 builds on the semantic and technical infrastructure completed in 0.5. The emphasis shifts from constructing the underlying machinery to using that machinery more directly within the investigation itself.
Concept pages
Important concepts that recur across books, essays, perspectives and observations can acquire lightweight pages of their own. These are not intended to become dictionary entries or a rigid conceptual taxonomy. They function instead as semantic junctions: places from which a recurring idea can be followed through the different contexts in which it appears.
Concept pages should emerge from the existing material rather than impose a new structure on it. Their role is to make recurring ideas more visible while preserving the rhizomatic character of the project.
AI integration
The semantic JSON, structured metadata, language-model resources and public API now provide a basis for experimenting with AI-assisted exploration.
The aim is not to attach a generic chatbot to the website. Any AI integration should support the investigation by helping readers find relations, formulate questions or discover relevant windows, without presenting an AI-generated synthesis as the final meaning of the project.
AI should therefore function as another navigational and interpretive layer within the landscape, not as an authority standing above it.
Structured-data refinement
The automatically generated JSON-LD will be refined further. Page images will be represented as Schema.org ImageObject structures with explicit width and height rather than only as image URLs.
Version 0.6 remains deliberately limited in scope. Its purpose is not to turn Wide Open Windows into a software platform, but to explore what becomes possible now that its human-readable and machine-readable structures are connected.
Version 1.0
Version 1.0 will mark the point at which all pages planned for the first public release contain meaningful content and the project as a whole has been reviewed for coherence, accessibility, metadata quality and long-term maintainability.
Every public page will present both a human-readable and a machine-readable representation. Readers, search engines, external applications and AI systems will therefore be able to navigate the conceptual landscape while preserving the distinction between content, structure and presentation.
Discovery itself will remain part of the experience. Readers can move through searching, suggested paths, recent updates, the Rhizome, Discover, concept pages and selected semantic relations rather than relying on a single conventional navigation structure.
The public read-only API and semantic resources introduced during development will by then have matured into stable interfaces for published pages, metadata and selected rhizome relations, still without accounts, writing permissions or a database-driven community platform.
Version 1.0 will not mean that the investigation is complete. It will mean that Wide Open Windows has become sufficiently coherent, stable and substantial to move from development into its permanent public form.
Beyond version 1.0
No fixed development plan exists beyond the first public release. By then, the project may already have evolved in directions that cannot yet be anticipated.
Some possibilities include:
- A complete Dutch edition becomes available alongside the English version, with both languages remaining synchronised as the project evolves.
- The books become available in additional printed and digital editions.
- Selected reader reflections gradually become part of the landscape itself, forming a curated dialogue rather than a conventional discussion forum.
- Essays, books, observations, perspectives and concepts become increasingly cross-linked through automatically generated semantic relationships.
- Readers can create personal collections, or “Windows”, by bookmarking pages into their own thematic journeys.
- A searchable atlas of lived experience gradually emerges from essays, observations, concepts, worlds and selected reader reflections.
- Machine-readable resources and the public API continue to develop, making the conceptual organisation of the project increasingly accessible to researchers, external applications and future generations of AI systems.
Wide Open Windows is not intended to become a static archive of finished texts. Its long-term ambition is to remain a living investigation: a place where new questions continue to generate new windows through which reality may be seen differently.
It is not merely a website about experience. It is an attempt to create a digital environment in which experience can be investigated, revisited and gradually reorganised through exploration.
Leave a reflection
If this page prompted a thought, question or observation, you can send it to Wide Open Windows. Nothing is published automatically.