1 files changed, 10 insertions, 0 deletions
diff --git a/docs/TODO.md b/docs/TODO.md
index 0db3a23..f58f1df 100644
--- a/docs/TODO.md
+++ b/docs/TODO.md
@@ -13,6 +13,16 @@ that has the full story. Keep this list current as items close or open;
 update the source doc's own entry too, don't let this drift into a
 second stale copy the way `PANEL_SUPPORT_TODO.md` did.
 
+## Fake-monitor incident, root-caused and fixed jointly with the AGS peer session (2026-08-27)
+
+Follow-up to the incident entry directly below. `dotfiles-1a` (AGS) confirmed from AGS's own session log (`~/.local/state/wm-session-*.log` - not `~/.cache/ags/ags.log`, which was stale) that the X-position jump was AGS's own remembered "extend-left" layout restore firing twice, once per fake monitor appearing (`arrange()` places the primary at 0,0 and walks the rest `left -= width`, then normalizes - correct behavior for a *real* second monitor; the only fault was a fake one being in the arrangeable set at all). Fixed on AGS's side: `readArrangeable()` now filters `!m.split && !m.virtual`. Their match was keyed on connector *name*, not index, ruling out the index-shift half of my own original theory.
+
+The Y drift (34 -> 66 -> 97, `full_y` staying `0` throughout every snapshot - confirmed from my own captured `srd monitors` output, so this was never the real monitor's true position moving, only its *usable*/bar-shrunk area shrinking further) was left unexplained by AGS's fix and flagged back to me to root-cause srdwm-side. Found it: `create_virtual_head` (`virtual_heads.rs`) created a real `wl_output` global and pushed the new head into `udev.virtual_heads`, but never registered it in `CompState::outputs` - the list `output_for_wl` (`state/mod.rs`) searches to resolve a client-named `wl_output` back to anything. `protocols/layer_shell.rs::new_layer_surface`'s own fallback for an output it can't resolve is "land on the primary output" (its own comment says so directly) - so AGS's own per-monitor bar, spawned for what it (reasonably, before its own fix above) believed was a new real monitor and aimed at that fake output, silently landed on the real primary output instead, each one stacking its own exclusive-zone reservation on top of the real bar already there. Two fake monitors, two misrouted bars, two zone increments (~32px, ~31px) - exactly the observed climb, and exactly reproducible from reading the code, not guessed.
+
+Fixed by registering (and, on removal, deregistering) a virtual head's `Output` in `CompState::outputs` the same way `bring_up_head` already does for a real one, so `output_for_wl` can actually resolve it and a client's layer-shell surface lands where it was actually aimed. Also added the discriminator `dotfiles-1a` asked for directly: `Monitor::is_virtual` / `MonitorInfo`'s `"virtual"` JSON field (`#[serde(rename = "virtual")]`, `is_virtual` Rust-side since `virtual` is a reserved identifier), `true` only for a fake monitor - AGS's own name-pattern match (`/^FAKE-/i`) was standing in for exactly this and can now be retired in favor of a real field. Full workspace build/test/clippy clean (34 platform tests, up from 33).
+
+Fake monitors are safe to create again. Not yet live-verified against a real repro (needs a restart); `dotfiles-1a` asked for a heads-up before the next live test so AGS's log can be watched at the same time.
+
 ## Live incident: creating a fake/virtual monitor corrupted the real monitor's position, repeatedly, with no further input (2026-08-27)
 
 Asked to demo fake monitors live. `srd dispatch create fake-monitor FAKE-1 1920x1080` was harmless (`eDP-1` stayed at `full_x=0`), but creating a *second* one (`FAKE-2`) immediately moved `eDP-1` to `full_x=1920`, and it kept drifting on its own for at least one more tick afterward (`full_x=3840, y` climbing ~31px per tick) with zero further commands issued. Removing both fake monitors stopped the drift but did not self-restore `eDP-1`'s position; fixed by hand via `srd dispatch set output position eDP-1 0 0`, confirmed restored. The user's actual laptop panel visibly shifted during this.