The routers order reads its model from the build standard (bstd().g, 'keep' is the old next-model-up rule) and the optical upgrade order its line system (bstd().sys). The Ops sheet shows "From the build standard: X" with a button to the sheet. Old saves move a model or line system from an order that was running into the standard when it has none of its own; a paused order's or an unhired intern's setting is dropped. opsParam refuses model/sys and points at bstdSet. The build standard's order is a generator now: every first job that takes a fiber dark goes through the optical order's routing test (opsDark, in the route worker in play), under one shared budget of dark swaps a night (two for Kevin) and the week-long memory for a failed fiber. A fiber one order worked on is the other's no more that shift, and decom keeps waves the standard just upgraded. The plan is checked at every job, since the shift rests between them: a POP gone or changed by hand meanwhile is skipped. The button still runs the whole plan at once on the page.
7 KiB
7 KiB
What is left (as of 2026-11-07)
Work in this fork, pushed to the vanvik Forgejo remote (never to origin). Built so far: the build standard, multiplayer
phase 1, dark mode (#80), IPv4 address space (#150), express fiber (#58) and satellite teleports (#83). Everything below is
planned or known, not built.
Promised and not done yet
- and 2. are built: the router model and the line system are picked in the Build standard sheet only (the
routersandoptuporders readbstd().g/bstd().sys; sanitize moves an old setting intoS.popStd), andstdtakes a fiber dark onoptup's routing test (opsDark()), sharing its budget a night, with one owner per fiber a shift (X.fib).
Next features
- Wave B starts with a shared session clock. Proposed on 2026-11-07, not decided yet: it changes how every
multiplayer game plays. Today each player's game runs in their own browser at their own speed (pause, 0.5× to 8×), so
players in one session are on different days. The proposal: the server keeps the time and every game in the session
follows it. The simulation stays in the browser.
- Server: a session gets a pace when it is created: one game day every 30 s to 5 min (agent mode's list). Never the
normal 2.5 s, which is 1,440 game days an hour. The day only moves while at least one player was seen recently (the
roster already keeps
seen), so nobody comes back to years of unattended play. Every roster answer carries the server's time and the session day. - Browser: the clock is locked to the session.
UI.clockLockedandHOOKS.speed()(50-state.js) already hide the speed buttons and refuse pause, andsave()already leavesuiSpeedalone while locked. The code goes insrc/game/, becausesrc/agent/does not ship (ADR 0004), so it needs its own copy of agent mode's worker ticker (src/agent/85-ticker.js) to keep time in hidden tabs. Agent mode also setsHOOKS.speed: the session wins, and agent mode is off in a multiplayer game. - Catching up: a game behind the session day runs days until it matches, a few per frame, with its progress
shown. It needs its own loop: with the route worker,
runClockholds one day at a time. A day takes about 30 ms on the big test network, so 400 missed days take about 12 s on a desktop and longer on phones. Late joiners start on the session's current day (to check: what in the world depends on the day a game started). - Nobody can pause or speed up. That is what "the same day" means.
- Later, optional: the server plays absent players. It opens an offline player's save in a headless game (jsdom,
as
scripts/lib/mcp-game.jsalready does with its realtime clock) and runs their days with the intern. The browser takes the save back on return through phase 1's save numbers, so the server is just one more device. Up to about 0.5 GB of RAM per absent player. Without it, other players see an absent network stand still. - Not proposed: a server that runs every game with the browser only showing it. Every click would be a round
trip, and the page reads
SandRdirectly everywhere, so it would be a rewrite. Its one gain over the above is verified cards (today the server trusts them). - Phase 2 gets simpler: its wall-clock timing (tenders that close after 10 real minutes,
idle≥ 300 s, notice on each player's own clock, the soak test's "own speed") can move to game days.
- Server: a session gets a pace when it is created: one game day every 30 s to 5 min (agent mode's list). Never the
normal 2.5 s, which is 1,440 game days an hour. The day only moves while at least one player was seen recently (the
roster already keeps
- Wave B: #97 co-op companies (several players run one network with roles, budgets and approvals), host-authoritative
over the multiplayer server, and multiplayer phase 2 (players as carriers in each other's simulations: transit,
peering, PNIs, shared fiber): design in
docs/plans/multiplayer-phase2.md. - Wave C: #144 directional traffic, last and on its own: keep the totals, split them by traffic profile, utilization is the busier direction. It comes after #58 because both touch link capacity (express channels now count too).
Known limitations of what shipped
- Multiplayer: other players' POPs are not drawn on the map (no clean hook yet); cards are what each browser reports,
the server does not verify them; a save over ~60 KB sends only the card when the tab closes; the HUD chip in phone
landscape wider than 760 px was not measured. Behind a reverse proxy, run the server with
--trust-proxy, or every player shares one address for the rate limits. - Build standard: the game menu has 8 tiles in a multiplayer game (7 otherwise), so the last row is not full. The
sheet's quote and "Bring everything up now" still judge a line-system swap without routing (
bstdDarkOk(): a quiet fiber, under 10% of its capacity, or a parallel link with room; any express with traffic holds its fibers,bstdXpBusy()), so the quote is the bill; only the intern routes to test. A quiet fiber with no way around can still go dark from the button. An old save'srouters.modelmoved into the standard (only from a running order with an intern hired; a paused one's is dropped) also becomes what new POPs get. A model or line system in the standard is a ceiling for theroutersandoptuporders (their next-one-up rule only applies with "keep"), as an order's own model was before;optuphas no per-POP exceptions, so it moves fiber at POPs excepted fromstdtoo. - Dark mode (#80): red error text on dark tinted cards is about 3.8–4.3:1 (bold, but below AA 4.5); the share card PNG and the web manifest stay light; the WebGL map was checked by tests only, not by eye in a real GPU browser.
- IPv4 (#150): the intern does not manage addresses (optional in the issue); transfers are instant; lease-out income is not in the investor revenue share; blocks granted to old saves can be sold after two years (a one-off windfall); CGNAT annoyance costs refunds but does not slow broadband growth; the NOC window's address-space button does nothing.
- Express fiber (#58): splicing a whole fiber pair through a city was left out on purpose (every fiber end is a POP; expresses cover the need); loss and MTU incidents on a fiber do not reduce the expresses on it; the fiber panel's "n of max channels" line does not count express slots (the Express section does); regen sites added later do not charge regen cards for express channels.
- Satellites (#83): no WISP zones, no satellite incidents (rain fade, sun outage), no footprints on the map, no satellite tab in the drag-to-link window (ordering is from the city panel); GEO/MEO losses and prices are design numbers; well-cabled islands (Wellington, Hobart) also count as "can be served"; MEO needs 5° elevation (10° would drop 416 pairs); Pitcairn's $120k satellite grant stays Pie-only.
Housekeeping
npm run perf -- --browserneeds Chromium (sudo pacman -S chromium, thenPERF_CHROME=/usr/bin/chromium); only the jsdom benchmark has been run.- Upstream release tags are not on
vanvikyet:git push vanvik --tagsif wanted. - The merged worktrees under
.claude/worktrees/and theirworktree-wf_*branches can be removed.