April 9, 2026
Homepage and supporting pages settle into a clearer public structure
This session turned the landing page and the site's supporting-material layer into something more coherent. The homepage navigation is cleaner, the lower half of the page now has a real editorial purpose, and the site finally has a proper set of public support pages instead of scattered placeholders and provisional links.
Now in production
- The shared header now uses a clearer top-level structure: Home, an Atlas dropdown, About, and Journal.
- The homepage lower section is no longer a weak carousel. It now works as a quieter Behind the Atlas band linking to the site's real support pages.
- About, Roadmap, and Methodology & Sources now exist as real public pages in the stable source/build workflow.
What became clearer
- The homepage works better when the lower section is treated as a support layer, not as another discovery carousel.
- About, Project Journal, Roadmap, and Methodology & Sources form a more believable family than mixing those links with half-formed future publishing surfaces.
- The right visual answer for the lower homepage band was not heavy contrast or more decoration, but better spacing, cleaner grouping, and a restrained scroll reveal.
Note:
This was not the planned focus for the session, but it was worthwhile. The landing page now feels much closer to a real public front door than to an exploratory staging ground.
April 8, 2026
Kepler and TESS move onto cleaner runtime data without changing the chart behavior
The exoplanet mission views now sit on a much sturdier data-handling foundation. Kepler and TESS no longer rely on browser-side CSV parsing at runtime. Instead, both views now use generated host-system atlas datasets behind the scenes, while the live chart behavior, scale controls, sidebar grouping, and NASA click-through flow remain the same.
Now in production
- Kepler and TESS now load generated runtime JSON rather than parsing mission CSVs directly in the browser.
- The current Kepler and TESS chart interactions were preserved rather than redesigned during the migration.
- The transit dataset now follows the current generated exoplanet mission runtime files instead of older legacy mission inputs.
What became clearer
- The exoplanet branch can reach data-handling parity before it reaches full scientific cleanup parity.
- Mission views benefit from the same one-system, nested-planets runtime pattern even when their visual logic should remain mission-specific.
- The exoplanet atlas is now in a much better position for a later Kepler and TESS audit pass.
Note:
This was mostly infrastructure work, but it matters. The exoplanet page no longer has to carry its cleanest data handling in only part of the experience.
April 7, 2026
Solar orrery gets fuller modal coverage and clickable outer-region labels
The solar-system orrery is now less likely to feel dead at the edges. Bodies without story stacks can still open useful info modals, the asteroid belt has a stronger regional story identity, and the Kuiper Belt and Scattered Disc labels now act as real clickable atlas regions rather than just decorative text.
Now in production
- Bodies and spacecraft without story-card sets now open info modals instead of doing nothing.
- The Asteroid Belt has a much richer regional story list.
- The Kuiper Belt and Scattered Disc labels now open info modals from the chart itself.
- The asteroid belt now shows a pointing-hand cursor on hover, making its interactivity more legible.
What became clearer
- Outer-region interaction works best when label hit areas are treated as explicit chart-space geometry.
- The solar page benefits from a second modal tier: not every destination needs a story-card stack to be worth clicking.
- Earth still needs a bigger long-term editorial treatment than the current baseline info modal.
Note:
This was less about adding a flashy new feature than about making the solar atlas feel more complete and more trustworthy as a place to explore.
April 3, 2026
Nearby atlas reaches 50 light-years
The exoplanet nearby view has now moved beyond its earlier 10-parsec proving ground into a full curated 50-light-year neighborhood. The important change is not just more systems, but a more trustworthy structure: one host system per point, cumulative discovery history over time, and a nearby dataset assembled through review rather than treated as a raw archive export.
Now in production
- The nearby exoplanet map now covers a curated neighborhood out to 50 light-years.
- The year slider now begins at 1995, matching the first historically relevant systems in the nearby timeline.
- 51 Pegasi now appears as a deliberate rim exception: just beyond the boundary, but too historically important to leave out.
What became clearer
- The nearby view works best when it is treated as an atlas dataset, not a raw table dump.
- Small editorial exceptions are fine when they are explicit and legible.
- The exoplanet branch now has one strong local-neighborhood view and one strong curated landmark view to build from.
Note:
The nearby map is now much closer to the project’s larger ambition: a chart of real neighboring worlds that can be read historically as discoveries accumulate, not just scanned as a static cloud of points.
April 2, 2026
Exoplanet nearby view resets around a trusted local shell
The exoplanet atlas finally got the kind of nearby-data reset it needed. Instead of continuing to rely on the older arbitrary nearby CSV, the production page now uses a curated local-neighborhood dataset built from the confirmed systems within 10 parsecs. The result is a nearby view that feels much more defensible: cleaner names, cleaner distances, cleaner timeline behavior, and a data shape that finally matches the way the chart actually works.
What changed
- The nearby view was rebuilt around a new host-system dataset rather than the inherited row-level CSV.
- Host names and distances were reset from the Wikipedia nearest-exoplanets list, while system coordinates were attached separately through CDS Sesame / SIMBAD.
- The chart now treats nearby destinations as real host systems with nested planet histories, so systems appear once and then accumulate over time as later planets are confirmed.
- The cleaned chart logic was promoted into the live exoplanet route rather than left behind as an iteration branch.
Why it mattered
- The nearby view now feels much closer to the sci-fi stars page in data discipline: one plotted system, one runtime source, one coherent timeline model.
- The local neighborhood finally reads like a real atlas layer instead of a brittle export from an old spreadsheet moment.
- This does not solve the full 50 light-year neighborhood yet, but it creates a trustworthy inner shell to build outward from.
- The exoplanet branch is now in a better position to tackle the harder Kepler and TESS cleanup later without repeating the same structural mistakes.
Note:
This was less about adding a new feature than about making one of the core atlas views finally feel trustworthy on its own terms.
April 1, 2026
Sci-fi atlas modal cleanup and MVP pass
The sci-fi stars page crossed an important threshold in this session. After the earlier story-to-star audit and data-source cleanup, the remaining work was mostly editorial: replacing placeholder blurbs, normalizing tags, and making sure uncertain story years no longer surfaced as misleading metadata. The page now feels much closer to a genuine first release rather than a promising but provisional branch.
What changed
- The remaining placeholder blurbs in the sci-fi dataset were replaced with short one-sentence summaries tuned for the atlas modal cards.
- Story tags were simplified into a much tighter production vocabulary focused on medium and major awards.
- Uncertain in-story years were nulled out conservatively instead of leaving behind inherited guesses or fake precision.
- The modal renderer was updated so missing story years disappear cleanly instead of showing awkward fallback values.
Why it mattered
- The page no longer feels crowded with provisional metadata and placeholder text.
- The chart now sits on top of a dataset that is cleaner both technically and editorially.
- The sci-fi atlas has started to feel like a real MVP branch of the site, not just a rescued prototype.
- Future work can now focus more on expansion and interpretive depth than basic cleanup.
Note:
This was a satisfying kind of work: less glamorous than a new visualization, but the sort of finish pass that makes the whole page feel trustworthy.
March 31, 2026
Sci-fi atlas audit and data-source cleanup
This session was less about adding a flashy new feature than about making the sci-fi stars page more trustworthy. A long content audit removed a large number of weak real-star matches, and the page's runtime data flow was simplified so the chart and the modal now read from the same source instead of drifting between JSON and CSV.
What changed
- The sci-fi atlas went through a star-by-star verification pass focused on removing false or weak star/story links, especially among catalog stars and overfit franchise matches.
- A lightweight browser checklist page was added so the audit can continue in a more structured way as the remaining systems are reviewed.
- The placeholder custom Deneb modal was retired, so only
Tlon now keeps a deliberately custom exception path.
Data cleanup
- The sci-fi chart no longer depends on a separate runtime CSV for geometry. The plotted star angle is now stored directly with each star entry in the JSON dataset.
- The page loader was refactored so the chart and the story modals now derive from one unified sci-fi data source instead of two parallel runtime files.
- This should make future additions and corrections less brittle, especially when a newly added star needs to appear both in the legend and in the modal layer right away.
Note:
This is the kind of cleanup that matters even when it is not very visible. The atlas works better when its speculative geography feels authored rather than accidental, and that requires both better curation and a cleaner relationship between the data and the page.
March 30, 2026
Transit Times launches as a new atlas companion
A new standalone page joined the site this session: Transit Times, a search-first companion tool that lets visitors look up destinations already mapped across the atlas and compare rough interstellar journey times under a few different propulsion assumptions.
What launched
- The new page is live as
transit-times.html and is designed around quick search, destination filters, and direct links back into the relevant atlas pages.
- It currently draws from the systems already mapped elsewhere on the site: the Sci-Fi Stars atlas, the exoplanet views, and the curated Landmark Systems set.
- The homepage now surfaces it directly by replacing the old placeholder Story Archive slot in the Keep Exploring band.
How it works
- A small build-time generator now merges the mapped source datasets into one shared transit target file, so the page stays tied to the actual atlas instead of a separate hand-maintained list.
- The page compares three rough benchmark journeys: Voyager 1, a 0.1c fast probe, and an optimistic 1g flip-and-burn scenario.
- Targets with missing usable distance values are now omitted from the transit dataset rather than shown with misleading zero-distance placeholders.
Note:
This is a good example of the atlas expanding sideways rather than just getting denser. It is not another map, but a companion tool built directly out of the destinations the project is already charting, and it may become an unusually good front door into the site over time.
March 28, 2026
Landmark Systems promotion and chart orientation pass
The exoplanet atlas gained a more public-facing front door this session. A curated Landmark Systems view is now part of the main production page, and the chart itself now offers a clearer sense of orientation through a smaller set of recognizable sky references.
What changed
- The former
v6 exoplanet branch was promoted into the main exoplanet-systems.html route, so the curated landmark view now lives in the public production page rather than a side experiment.
- The versioned implementation files were moved into the archive after promotion, keeping the active exoplanet route easier to read and maintain.
- The intro copy was rewritten to explain the chart in visitor-facing language rather than internal development terms.
Orientation cues
- The chart now uses a smaller outer marker set built around human-readable references: Galactic Center, Polaris, and Orion.
- These markers are meant to help a casual viewer understand which part of the sky the map is looking toward without turning the page into a full sky chart.
- A brief “How to read this chart” note was added above the controls to explain distance, direction, and the meaning of the numbered markers.
Note:
This is less about perfect astronomical completeness than about making the atlas readable. The aim is to give the exoplanet page a clearer human point of view while keeping the underlying heliocentric structure intact.
March 27, 2026
Static atlas interaction and mobile upgrade
This session brought the two static atlas pages much closer to the solar-system orrery in terms of direct usability. The sci-fi and exoplanet maps now support stronger mobile interaction, sharper high-density rendering, and a more coherent expanded-view language, which makes them feel less like fixed posters and more like instruments you can actually work with.
Sci-fi stars
- The sci-fi map gained a cumulative publication-year timeline, so systems now appear as works enter the historical record and gain rings as additional stories accumulate.
- The expanded-mode branch was promoted into production, bringing the page much closer to the solar orrery's fullscreen reading mode.
- Mobile rendering was sharpened with a proper high-DPI canvas setup, and touch pan, pinch zoom, and tap interaction were added to the production page.
Exoplanet atlas
- The exoplanet page inherited the same expanded-stage and narrow-screen panel system in a new working branch, then that branch was promoted into production.
- The atlas now supports touch pan, pinch zoom, and improved mobile rendering, which makes the near-neighbors, Kepler, and TESS views much more usable on phones.
- A cleaner release workflow also emerged: once a versioned branch is promoted, its working files now move into the archive so the main routes stay easier to read and maintain.
Note:
This was not just polish. It was the point where the static atlas pages started to inherit some of the solar orrery's confidence: expanded reading space, tactile navigation, and mobile behavior that feels intentional rather than merely tolerated.
March 25, 2026
Archive cleanup and search-readiness baseline
This session was about making the public site feel more intentional. A large group of old test pages, prototype routes, and support files were moved out of the published surface and into archive space, while the live pages gained a cleaner baseline for search appearance through canonical URLs, page descriptions, and a first pass at sitemap and robots files.
What changed
- Older route experiments, reference files, and historical support material were moved out of the public
Code/ surface into archive space so the live site no longer exposes as many obsolete branches.
- The root repo documentation was refreshed so the source/build workflow, archive split, and current public structure are easier to understand at a glance.
- The public pages gained canonical URLs and clearer page descriptions, giving search engines and social previews a more deliberate signal about what each route represents.
- A conservative
sitemap.xml and robots.txt were added for the current stable public pages rather than the broader repository history.
Why it matters
- The project now presents a much cleaner public face: fewer stray routes, fewer accidental archive pages, and a more obvious separation between current site and historical working material.
- Search results should have better page-level context over time because the site is now supplying explicit descriptions, canonical targets, and a published sitemap.
- This was less about chasing generic "SEO tricks" than about making the current site easier to identify, crawl, and represent accurately.
- The cleanup also makes future architecture work easier, because it is now clearer which files belong to the living site and which belong to its development history.
Note:
The project is still evolving, so this should be read as a baseline cleanup and indexing pass rather than a declaration that the information architecture is finished. But the public surface is now much more coherent than it was a day earlier.
March 24, 2026
Shared header rollout and exoplanet atlas promotion
This session finished a piece of site infrastructure that had been left half-solved after the footer pass, then used that stronger shell to finally bring the exoplanet atlas up to the same standard as the newer pages. The headline change is not just that the exoplanet page is now responsive, but that it was rebuilt through a safer test-track process and then promoted into the source-built production workflow.
What changed
- The stable source-built pages now share one production header partial and one production footer partial, instead of carrying more duplicated top-of-page markup.
- The exoplanet atlas was rebuilt through
v2 and v3 test tracks before replacing the live route, which made it possible to test responsive behavior without destabilizing the public page.
- The near-neighbors view now has a proper solar-space gutter and a more explicit distance projector, so the local map reads more like the Kepler and TESS atlas views while staying true to its own dataset.
- The exoplanet page now lives in the same source/build workflow as the other stable pages, with
site-src/pages/exoplanet-systems.html compiling to the public Code/exoplanet-systems.html output.
What improved
- The exoplanet canvas now renders crisply at responsive sizes instead of inheriting the slight blur caused by fractional width measurement.
- The Sun interactions are now more coherent across the atlas pages: the center can act as a doorway back to the solar-system orrery, and the view menu now makes that relationship clearer.
- The side legend is more useful than before, with a subtle hover state and chart-linked interactions that make the list feel like part of the map rather than an afterthought.
- Several edge cases were cleaned up in the process, including tooltip behavior, back-button state after leaving for the solar page, and broken legend ordering caused by invalid Kepler distances.
Note:
This does not mean the exoplanet atlas is finished. The Kepler dataset still needs cleanup, and there is still room to unify responsive symbol scaling more deliberately across the scientific and speculative maps. But this is the first time the exoplanet page has felt like part of the current site rather than a surviving older branch.
March 24, 2026
Shared footer build system and subscribe testing
This session was less about a single visual page and more about removing a source of recurring friction. The production footer is no longer just copied around by hand. A lightweight build step now lets the stable pages share one footer source, which made it practical to clean up route drift and test multiple newsletter backends without editing five pages every time.
What changed
- A small source/build workflow was introduced so the stable pages now compile from
site-src/pages into the public Code/ output directory.
- The footer markup and footer CSS were moved into shared partials, so changing the production footer now means editing one file instead of manually syncing every current page.
- The main stable routes were normalized around
exoplanet-systems.html, scifi-systems-extrasolar.html, and solar-system-orrery.html, with shared-worlds.html added as a simple placeholder page.
- The homepage and other production-facing links were updated so the main navigation hooks now point to the current stable filenames instead of older working branches.
Subscribe tooling experiments
- The footer subscribe UI was tested first against Kit using a direct form post rather than the full provider widget, which fit the custom site design better but exposed deliverability and reporting frustrations.
- Beehiiv was then tested as an alternative. Its smaller embed fit the footer visually, but browser behavior and blocker sensitivity reinforced that third-party embed scripts are still a weak point for this kind of shared footer.
- Even so, the new footer architecture made these experiments much easier: the provider could be swapped once in the partial and rebuilt across all stable pages.
- For now the goal is not to perfect newsletter infrastructure, only to prove that the footer can be wired to a real backend while the site remains in a low-stakes stage.
Note:
The important gain here is architectural. The site now has a path away from drifting shared chrome, and that should make future header/footer cleanup much less painful than the last few rounds of manual syncing.
March 23, 2026
Sci-fi stars release-candidate pass
Tonight shifted the extrasolar science-fiction page from a rough but promising shell integration into a page that now feels like a credible second atlas experience beside the solar-system orrery. The work was less about adding dramatic new features and more about giving the page a better structure, clearer framing, and a calmer visual hierarchy so the chart could breathe.
What changed
- The old sci-fi page was rebuilt into a new iterative branch and then promoted to
scifi-systems-extrasolar.html as the current release-candidate filename.
- The page now uses the shared site header and footer, a lighter interior shell, and a more deliberate top section with title, explanatory copy, and grouped controls.
- The chart and legend now live inside one unified experience panel, with a custom grid layout closer to the solar page instead of a brittle Bootstrap column split.
- The Sun at the center now acts as a doorway back to
solar-system-orrery.html, making the one-light-year gutter feel intentional rather than empty.
What improved
- The legend was restyled with a cleaner sans-serif catalog feel, a more refined inset scrollbar treatment, and dimensions that sit more comfortably beside the chart.
- The chart wrapper, controls card, and legend now share tighter geometry and border logic, which makes the whole page feel less improvised.
- A page-specific draw script now measures the canvas more carefully, which fixed the slight softness caused by fractional layout width.
- The sci-fi stars route is now wired into the homepage, the solar page, the journal, and the docs as a real current experience instead of a lingering prototype filename.
Note:
This page still feels less alive than the solar-system orrery, but it now has a clearer identity. The next gains probably come not from more shell polish alone, but from interpretive overlays, richer motion, and deeper system-focused interaction.
March 22, 2026
Journal page and editorial shell direction
This session was about finding a more believable reading-space language for the project. The dev journal moved away from a dark, dashboard-like treatment and toward a lighter editorial page with the shared site shell around it. Just as important, the project’s design references became clearer: more serious editorial and print-derived than game-like or purely atmospheric.
What changed
- The journal became a real page with the shared site header and footer and a lighter body better suited to reading.
- Footer/header inconsistencies surfaced again, reinforcing the need for a more reusable shared shell strategy later.
- The site now has a stronger case for two related visual languages: darker atlas instruments and lighter editorial reading spaces.
Design references that clarified things
- Works in Progress, Cabinet, Tablet Magazine, Aeon, and The Baffler all helped sharpen the editorial direction.
- The Pudding and New York Times graphics remain important references for richer longform layouts that can break out of the body column when needed.
- The project’s desired tone was framed more clearly as “more Le Guin, less Halo.”
Note:
The journal page is still a baseline rather than a finished template, but it now points toward a more serious and more literary article system for future essays and atlas notes.
March 20, 2026
Touch controls, mobile testing, and a realism detour
This session shifted from page polish to interaction behavior. The solar-system orrery now responds to touch on real devices: one-finger drag pans the chart, two-finger pinch zooms it, and taps on bodies still open the story modals. A short experiment with giving Sedna a selectively eccentric orbit was also useful, even if it mostly confirmed why the current circular abstraction remains the cleaner default.
Now working
- Expanded mode now supports touch pan and pinch zoom on mobile browsers.
- Contained mode now supports the same touch behavior, which makes the chart much more usable on phones.
- Tap-to-open modal behavior was restored by handling touch taps explicitly instead of depending on synthetic mobile click behavior.
- Testing the local dev server over Wi-Fi from a phone is now part of the workflow, which should speed up future mobile iteration.
What became clearer
- A single true eccentric orbit looks intriguing in isolation, but it becomes visually jarring when everything else remains perfectly circular.
- Cubic scaling is not a good fit for a truthful eccentric-path exception, because the compression breaks the geometry too aggressively.
- Linear distance makes that kind of orbit read more plausibly, but plausibility alone does not mean it belongs in the current visual system.
- The mobile state is now functional enough to live with for the moment, even if it is not yet elegant.
Note:
Tonight was a good reminder that not every interesting scientific detail belongs in the default view. Some of the more realistic orbital ideas may want to return later as a dedicated mode rather than as isolated exceptions.
March 19, 2026
Solar-system orrery release-candidate pass
This was the long integration pass that turned the solar-system orrery from a promising prototype into a page that can plausibly stand as the current release candidate. The biggest change was not visual polish by itself, but the decision to treat contained view and expanded view as two deliberate modes of the same page.
Major gains
- The page now has a contained default view and an in-page expanded view instead of relying on browser fullscreen.
- The immersive mode was rebuilt around its own canvas host, which finally restored the prototype feeling without discarding the newer shell language.
- The expanded matte treatment solved the canvas blur problem by separating the visual border from the sizing element.
- The below-chart atlas index in contained mode became a responsive grid, which makes much better use of the width when the legend drops beneath the chart.
Smaller but important
- The expanded controls are draggable again, and the keyboard shortcuts for controls, legend, reset, and scale mode are more coherent.
- The homepage now points to the right solar page instead of the old prototype.
- The shared docs were updated so the project now recognizes the solar-system orrery as a first-class current page.
- Temporary debug logging in the drawing code was removed once the page stabilized.
Note:
This was also the point where several longer-term questions became clearer: Earth probably needs special handling, moon-heavy systems may need richer drilldowns or modals, and 3D is not really a side quest so much as an eventual expression of the project’s deeper identity.
March 18, 2026
Landing-page cleanup and project memory work
This session was less about one dramatic page change and more about making the project remember itself. Alongside some landing-page cleanup and routing fixes, the main work was documenting what the solar-system branch had become, where its weak spots still were, and how future ideas like Earth, moons, and richer subviews should be carried forward.
What was put in place
- The homepage links were updated so the journal and solar-system page point at their intended destinations.
- The solar-system page got its own dedicated documentation page to record current behavior, follow-ups, and performance notes.
- An initial dev-journal habit started taking shape so major implementation decisions would not disappear into chat history.
- Project docs were refreshed so the current site state is easier to reconstruct and hand off later.
Why it mattered
- The project now has a better internal memory, not just a set of pages.
- Questions about Earth, moons, realism, and future 3D directions now have explicit places to live before they turn into implementation work.
- The distinction between interactive atlas pages and editorial reading pages started to sharpen in a useful way.
Note:
This was also the session where it became easier to say that the site probably needs two visual languages: darker, more instrument-like shells for the maps, and lighter, more editorial reading spaces for journals, essays, and reference pages.