Creepso/fluxa-core
1
0
Fork 0
mirror of https://github.com/FluxaMedia/fluxa-core.git synced 2026-08-04 07:56:00 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

docs: mark §11 (desktop JSON tax) resolved after verifying fluxa-desktop

Browse source 
Per CLAUDE.md's rule to verify against fluxa-desktop/src-tauri call
sites before touching FluxaCore: the plan's premise for this section
no longer holds. dispatchAction/completeEffect/getSnapshot in
fluxa-desktop's src/core/engine.ts already go through dedicated Tauri
commands (engine_dispatch/engine_complete_effect/engine_snapshot) that
call FluxaCore::* directly in Rust, not through core_invoke. FluxaCore
itself has also grown from the "8 methods" this doc assumed to 21
(cast/AirPlay/Chromecast/Roku helpers), all with real call sites
verified via grep. The remaining core_invoke Tauri command is a
generic fallback for infrequent methods where a dedicated command
per method would be boilerplate with no measurable win, since Tauri
IPC serializes as JSON at the JS<->Rust boundary either way. No code
change needed; also corrected the stale "8 methods" claim in
CLAUDE.md's core_api description to reflect the current count.
This commit is contained in:
KhooLy 2026-07-07 15:10:23 +03:00
parent 4ce837b125
commit 603a94bf6c
1 changed files with 13 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

22
docs/robustness-plan.md
View file

@ -225,16 +225,20 @@ Statics move to `OnceLock`; the preference-derived ones get a one-entry cache ke
the preference string (a `Mutex<Option<(String, Regex)>>` is enough — the preference
changes once per settings edit, not per stream).
### 11. Desktop pays JSON tax it doesn't owe
### 11. Desktop pays JSON tax it doesn't owe — RESOLVED, already fixed on the desktop side
`fluxa-desktop` links this crate natively yet still calls `core_invoke(method,
args_json)` — serialize on the Tauri side, parse here, serialize the result, parse it
back. For the 8 `FluxaCore` methods plus the hottest `core_invoke` calls (dispatch,
completeEffect, stream ranking), expose typed Rust signatures
(`fn dispatch(&self, action: AppAction) -> DispatchResult`) and let desktop call them
directly. The JSON path stays for Swift/WASM, where it's the genuine boundary. Verify
against `fluxa-desktop/src-tauri` call sites before adding anything to `FluxaCore`
(per CLAUDE.md rule).
Verified against `fluxa-desktop/src-tauri` call sites (2026-07) per the CLAUDE.md rule, and
this section's premise no longer holds. The hottest paths — `dispatchAction`,
`completeEffect`, `getSnapshot` in `src/core/engine.ts` — already go through dedicated Tauri
commands (`engine_dispatch`, `engine_complete_effect`, `engine_snapshot` in
`src-tauri/src/lib.rs`) that call `FluxaCore::headless_engine_dispatch_json` etc. directly in
Rust, not through `core_invoke`. `FluxaCore` has also grown from the "8 methods" this doc
assumed to 21 (cast/AirPlay/Chromecast/Roku helpers included), all with real desktop call
sites. The one remaining `core_invoke` Tauri command is a generic fallback for the long tail
of infrequent methods (`playbackPreparePlan`, `preferencesSchema`, etc.) — writing a
dedicated Tauri command per method there would be pure boilerplate, since Tauri IPC already
serializes args as JSON crossing the JS↔Rust boundary regardless of whether the command is
generic or dedicated; there's no double-parse to eliminate. Nothing left to do here.
### 12. Measure before and after