Voice admin: ui_socket_path() returns admin.sock instead of ui.sock #34
Labels
No labels
prio_critical
prio_low
type_bug
type_contact
type_issue
type_lead
type_question
type_story
type_task
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
lhumina_code/hero_voice#34
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
The
ui_socket_path()function incrates/hero_voice_admin/src/main.rsreturnshero_voice/admin.sockdespite its name and intent. Hero_router proxies/hero_voice/ui/to<svc>/ui.sock, so the iframe gets a 404 / connection-refused.Root cause
Should be
ui.sock. Sibling crates (hero_office_admin,hero_aibroker_admin,hero_compute_admin) all bind their UI listener at<svc>/ui.sock. The top-of-file doc comment also still referencesadmin.sock.Repro
ls ~/hero/var/sockets/hero_voice/— admin.sock present, ui.sock missing.Expected
Voice iframe loads HeroVoice UI; the socket dir contains
ui.sock.Fixed in
50e033f6c3. Awaiting CI before closing.Reopened to correct the record.
The original premise was wrong. Hero_router's
resolve_ui_socket(hero_router/crates/hero_router/src/server/routes.rs:916) tries<svc>/admin.sockfirst (canonical) and falls back to<svc>/ui.sockas legacy. Voice was already bindingadmin.sock(per the Phase-1 migration inc9a3299,web.sock → admin.sock), so the iframe was working. Only the function NAMEui_socket_path()was misleading.Original "fix" commit
50e033f(changing the binding toui.sock) was a regression away from the canonical convention. It's been force-removed fromdevelopment. Replacement commitdc9af3fjust renames the function toadmin_socket_path()so the source matches reality.Closing as not-a-bug. The misleading-name cleanup ships in
dc9af3f.