fbbe3293
docs: the wasm figure is the final tree's, and the README quotes bytes per keystroke
a73x 2026-09-04 18:04
Commit message
README.md
| Old | New | ||
|---|---|---|---|
| @@ -646,7 +646,7 @@ mux d dump [--vt] # debug: print the authoritative grid (what the screen *shou | |||
| 646 | mux d stats # live sessions by name + clients each; wire stats (deltas vs snapshot bytes) | 646 | mux d stats # live sessions by name + clients each; wire stats (deltas vs snapshot bytes) |
| 647 | mux d upgrade # exec a newly installed binary in place; every session keeps running | 647 | mux d upgrade # exec a newly installed binary in place; every session keeps running |
| 648 | mux d upgrade HOST # push THIS binary onto HOST over ssh, then upgrade its daemon the same way | 648 | mux d upgrade HOST # push THIS binary onto HOST over ssh, then upgrade its daemon the same way |
| 649 | make bench # typing-workload byte-ratio measurement | 649 | make bench # typing-workload bytes-per-keystroke measurement |
| 650 | ``` | 650 | ``` |
| 651 | 651 | ||
| 652 | `mux a` is the agent-facing mode — verbs `status`, `capture`, `send`, | 652 | `mux a` is the agent-facing mode — verbs `status`, `capture`, `send`, |
| @@ -856,7 +856,8 @@ Prototype, eighteen milestones in, current cut tagged `v0.0.1-15`. Linux | |||
| 856 | and Apple-silicon macOS; every gate below was measured on Linux. All | 856 | and Apple-silicon macOS; every gate below was measured on Linux. All |
| 857 | founding kill criteria cleared, measured on real networks rather than | 857 | founding kill criteria cleared, measured on real networks rather than |
| 858 | loopback: reattach into a live full-screen `nvim` in ~5ms from one | 858 | loopback: reattach into a live full-screen `nvim` in ~5ms from one |
| 859 | snapshot; steady-state delta traffic ~1% of snapshot-equivalent cost; the | 859 | snapshot; steady-state delta traffic ~65 bytes per keystroke, and a cold |
| 860 | attach to a fresh session 86 bytes of cells where it was 5644 of VT; the | ||
| 860 | protocol adds ~4ms to a raw ssh round trip (constant, not a multiple of | 861 | protocol adds ~4ms to a raw ssh round trip (constant, not a multiple of |
| 861 | RTT); ten consecutive real network tears resumed hands-off, all | 862 | RTT); ten consecutive real network tears resumed hands-off, all |
| 862 | delta-served, on both ssh and QUIC transports; QUIC cold attach beats | 863 | delta-served, on both ssh and QUIC transports; QUIC cold attach beats |
docs/decisions.md
| Old | New | ||
|---|---|---|---|
| @@ -9080,7 +9080,7 @@ would cost about 340 on a blank 80x24 screen (23 empty rows at 8 bytes each | |||
| 9080 | plus the typed row), so 160 catches that and leaves the honest figure room. | 9080 | plus the typed row), so 160 catches that and leaves the honest figure room. |
| 9081 | The bench still prints the attach snapshot's bytes beside it, for reading. | 9081 | The bench still prints the attach snapshot's bytes beside it, for reading. |
| 9082 | 9082 | ||
| 9083 | `mux_core.wasm` is 19080 bytes, from 356344. That is the whole terminal | 9083 | `mux_core.wasm` is 19075 bytes, from 356344. That is the whole terminal |
| 9084 | emulator leaving the browser client: the wasm build has no ghostty dependency | 9084 | emulator leaving the browser client: the wasm build has no ghostty dependency |
| 9085 | at all now, and what remains is the wire contract, the grid and the replay | 9085 | at all now, and what remains is the wire contract, the grid and the replay |
| 9086 | core. The 356344 figure is main's artefact as built on 2026-09-03 from a tree | 9086 | core. The 356344 figure is main's artefact as built on 2026-09-03 from a tree |