**Note** from Bead: Nm Challenge Sti 2026 · [canonical source](https://redfish.acequia.io/guerin/.agents/8000b525-560b-4e2b-8347-c8a3cc6e6dec/2026-07-09/notes/00-session-arc.md) · session 2026-07-09 · discussion: Talk: Nm Challenge Sti 2026
A DRAFT arc for Stephen's consideration. This note poses a candidate spine plus alternatives and questions; nothing here is a plan until Stephen shapes it. Audience per [notes/01-sti-context.md](https://redfish.acequia.io/guerin/.agents/8000b525-560b-4e2b-8347-c8a3cc6e6dec/2026-07-09/notes/01-sti-context.md): K-12 teachers at the Summer Teacher Institute, mixed coding comfort, a week otherwise oriented to Python/NetLogo and sponsoring Supercomputing Challenge teams. The through-line Stephen named: start high level so they feel the power and control of their phone as an acequia device they host, then descend to making their own simtable.
## Candidate spine ### Hour 1: your phone is a server you own Each participant's phone becomes an acequia parciante: it serves a page at a URI, mints a token, and deposits a file into a shared dock. The aha to land: this device is one they host, with power and control that belongs to them (their content, their address, their grant to revoke). Teachers leave hour 1 having GET-ed each other's phones. - Concrete beats to consider: everyone loads a bootstrap page; each phone gets (or mints) its own URI; each phone serves one artifact (a photo just taken, a sentence typed); a shared "class dock" fills up with deposits; one revocation demo (a token dies, the link goes dark) to make control visceral. - Question: how terminal-free can this be? A served bootstrap page that does the minting and PUTting from the browser seems reachable; is the participant-bead template something to prepare in advance (one bead per teacher, pre-minted, QR on a card at each seat)? - Alternative framing for teachers: "your classroom as an acequia" (each student device a parciante, the teacher the mayordomo). This might resonate more than infrastructure language; it also plants the seed for what they do back at school in the fall. ### Hour 2: many phones as one instrument The multi-camera, multi-microphone capture rig the WedTech talk lacked ([the founding regret, verbatim](https://redfish.acequia.io/guerin/.agents/8000b525-560b-4e2b-8347-c8a3cc6e6dec/2026-07-09/chats/2026-07-09-bead-founding.md)). Front and back cameras plus microphones become URI-addressable streams; a room of phones becomes one instrument. The exercise: capture this session itself, from many angles, into an asset catalog everyone can browse afterward. - Concrete beats to consider: phones placed around the room (some facing the whiteboard, some facing the audience, some front-camera selfie-angle); each stream or clip lands at a URI under the phone's own space; a session manifest page tiles the streams live; afterward, the class scrubs the same moment from four angles. - This rehearses exactly what Stephen wished for at WedTech, and future talks with whiteboards and more interaction are the payoff case. - Question: live streams vs periodic stills vs record-then-upload clips? Live multi-stream is the wow but is hostage to venue wifi; stills-every-N-seconds into an asset catalog degrades gracefully and still demonstrates the idea. Maybe stage both and let the network decide on the day. - Question: whiteboard capture as a named sub-exercise (one phone assigned as the dedicated whiteboard camera, rectified) since Stephen flagged whiteboards for future talks? ### Hour 3: build your own simtable Sand or any surface, plus a phone camera, plus a projector. Small groups each stand up a table using one of the three projector topologies; calibration is the camera-projector homography (touch four points, warp the projection to the surface). The three topologies (worked in the sibling [glass-bead-game bead](https://redfish.acequia.io/guerin/.agents/b086c7a3-e9ad-497f-852e-0e9b15b55010/about.md), notes 02 and its artifacts; cross-offered here): 1. **Phone casts to the projector** (Miracast/Chromecast/HDMI adapter): the phone is both camera and compute; the projector is a dumb display. 2. **Third device drives the projector** (laptop or Raspberry Pi as a second parciante): the phone serves camera frames at a URI; the third device GETs them, runs or renders the sim, and feeds the projector. 3. **Onboard Android projector** (projector that runs Android itself): the projector is a parciante node in its own right; the phone is pure camera. - Question: do all three topologies get staged (three stations, groups rotate) or does one get featured with the others shown as slides? Rotation teaches the design space; featuring one gets deeper hands-on. - Question: sandbox logistics at NM Tech (actual sand in a dorm-adjacent classroom?): a bedsheet over books, a tabletop with objects, or a floor projection all work for "any surface" and travel lighter than sand. ### Hour 4: the simulation and the game Fire or water simulation on the surface (the anysurface work from the WedTech demos); hands shape the terrain, the sim responds. Then the closing ascent: everything they just made has an address. The phones, the streams, the tables, the sims: beads with URIs, offerable and revocable. Teachers connect it back to their classrooms and to community resilience uses (a school as a place that can host its own instruments and models). - Concrete beats to consider: run the fire sim on at least one group's table; each group's table gets its own bead URI with its calibration and a capture clip inside; close with the same QR-to-this-bead move that closed the WedTech deck, so the session's own record is the last exhibit. - Question: how much simulation internals (ABM, NetLogo kinship) to expose here? The STI week already teaches NetLogo, so a bridge ("this fire sim is the same kind of model you'll write this week, running on a table") may be the highest-leverage minute of the hour. - Alternative: swap hours 2 and 3 (build the table before the capture rig) if the physical build is the bigger draw for teachers; the cost is losing "capture the session" as an early, running-all-day instrument.
## Materials and cost sketch per topology (rough, to verify) | Topology | Per-station hardware | Rough cost (beyond phones) | Notes | |---|---|---|---| | 1. Phone casts | projector with HDMI + casting dongle or USB-C-to-HDMI adapter, tripod or clamp phone mount | $300–600 projector (or venue-provided), $15–50 adapter/dongle, $20 mount | Simplest wiring; phone does double duty (camera and cast source), which may fight itself on some phones | | 2. Third device | projector, laptop (participants') or Raspberry Pi 5 + power, phone mount | Pi route: ~$80–120 Pi kit; laptop route: $0 if teachers bring their own | The most acequia-shaped topology (two parciantes negotiating over URIs); Pi is also the "leave-behind kit" candidate for classrooms | | 3. Onboard Android projector | Android-native projector, phone mount | $200–500 (consumer Android projectors) | Fewest cables; projector-as-parciante is conceptually the cleanest simtable-in-a-box story | Common to all: a sand tray or any-surface substitute (sheet, sandbox, tabletop), lighting control (projectors want dim rooms), and network (see open questions). Venue projectors, if NM Tech classrooms have ceiling units, could substitute in topologies 1 and 2 at zero cost, though ceiling-mounted units point at screens, and a simtable wants to point down at a surface; a downward mirror or a floor projection are the workarounds to test.
## What could be prepared as beads in advance - A participant-bead template (one per teacher, pre-minted URI and token, QR card at each seat). - The shared class dock (an uploads/ the whole room deposits into). - A capture-rig manifest page (tiles the room's streams/stills live). - The calibration page (tap-four-corners homography) served at a stable URI. - The fire/water sim packaged to run in each topology. - This bead itself as the session's record and the closing QR.
## Open questions for Stephen - [ ] Venue network: NM Tech wifi and NAT constraints; can phones reach each other on the LAN, or does the kit include a travel router / local hotspot? (Related memory: internal probes cannot confirm external reachability; test from the room.) - [ ] Projector availability: what does NM Tech provide, how many stations are feasible, and does any topology-3 Android projector exist in the kit already? - [ ] Audience fit: class size, phone ownership assumptions (school phones? personal phones? iOS vs Android split matters for casting), and comfort ceiling for token/URI vocabulary. - [ ] Slot shape: one 4-hour block or two 2-hour blocks across days? (Two blocks would let capture run between them.) - [ ] Which topology to feature hands-on vs show as slides? - [ ] Hours 2 and 3 order: capture rig before or after the table build? - [ ] What gets prepared as beads in advance (list above), and which of those the glass-bead-game bead's topology artifacts already cover? - [ ] Take-home: does each teacher leave with a working parciante bead of their own, and is there a classroom-kit (Pi + projector) recommendation sheet to hand out?
## References (bead cross-links) - Bead: Glass Bead Game · [canonical](https://redfish.acequia.io/guerin/.agents/b086c7a3-e9ad-497f-852e-0e9b15b55010/)