Tags: gethopp/hopp
Tags
feature: show call info about other users (#324) * add: make call-center the same in room and calls * add(backend): call state in redis 1:1 calls: user:call:{userID} → JSON {peer, room} — set on accept, deleted on end or WS disconnect (peer notified via call_end). Rooms: user:room:{userID} → roomID — set on join, deleted on leave or WS disconnect. * add: export endpoint for call presence * add: show in frontend with which user a user is in a call * refactor: have many users in peer list in the redis entries * add(backend): handle join call from the backend * improvement: use http endpoint for join call * add: semi working 3 users in one call * improvement: broadcast call state changed for faster ui updates * fix: ui fixes * improvement: clean callState redis on server start * add: tests for callstate * fix: fixes * fix: frontend fixes * chore: review feedback * chore: github bot comment * bump version --------- Co-authored-by: konsalex <konstantinos.alexoglou@grafana.com>
fix: hang on rapid start/stop threads (#323) * fix: hang on rapid start/stop threads The screensharing window could hang the main event loop when it was closed and then opened again. This could happen because the close could happen while the redraw_thread was in request_redraw, which needs the main event to be free in order to continue and at the same time a new open would wait from the main event loop for the previous redraw thread to join. This change removes all the joins from the codebase as they don't really needed for protecting us from zombie threads. * chore: bump release
fix: races in incoming calls (#318) * fix: drop websocket with pings and new ping interval * revert: dedupe ttl * fix: disable call button when receiving a call * fix: do not render call-center after 60s of an initiated call * fix: add did not answer timeout when we are calling * chore: bumb version * remove: dead code * review: feedback
PreviousNext