Repository navigation
✨ Betty speaks ready answers shortest first and moves a long voice queue to chat - #18627
clubanderson wants to merge 2 commits into
Conversation
…eue to chat Quorum's say queues clips and plays them in order, so hive keeps an estimate of the queued speech. Spoken answers ready in the same loop turn are said back to back, shortest first; one that would wait behind more than 20 s goes to chat with one spoken "I've put the rest in chat"; an ack behind more than 8 s of speech is skipped. Answers to addressed requests may take meetings.chat_answer_burst (3) extra chat posts a minute past the chat rate. Part of #18590. Signed-off-by: Andy Anderson <andy@clubanderson.com>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
Open PR overlap checkBase branch: Conflicts are checked with every PR merged onto the current tip of
|
Mutation testing5 mutant(s) in changed files are not reached by any test (NOT COVERED). A maintainer may apply the
Verdict: |
Part of #18590. Signed-off-by: Andy Anderson <andy@clubanderson.com>
|
Combined into #18635 to get through CI and review together. |
Part 3 of the concurrent-requests work on #18590: ordering of parallel answers and the chat burst.
Voice. Quorum's
sayreturns{queued:true}once a clip is accepted and plays clips in call order (quorumdocs/contract.md), so she cannot talk over herself, but hive cannot see when she stops talking. The session now keeps an estimate of the queued speech (15 characters a second, at least 1 s a clip) and:drainPersona, up to 16) and speaks the spoken answers among them back to back, shortest first;voiceBacklogMax), and says "I've put the rest in chat." once (at most every 30 s, localized for en/es/fr/de/pt/it);ack_no_voicedrop, as before for an unspoken ack).Chat burst. New
meetings.chat_answer_burst(default 3,HIVE_MEETINGS_CHAT_ANSWER_BURST): oncemeetings.chat_posts_per_minute(6) is spent, answers to addressed requests (and voice answers falling back to chat) may take that many extra posts a minute. Spek's posts, room-question answers and queue notes keep the plain rate.Combined with #18623 (parallel model calls) and #18625 (attribution), answers post as soon as each is ready and their prefix makes the order unambiguous.
The run-loop change is one added line after
s.onPersonaReply(ctx, rep); #18592 adds a case just above it, so whichever merges second rebases trivially.Tests (
live/voicequeue_test.go): three ready answers spoken shortest first with the long one moved to chat and the note spoken, an ack skipped behind 12 s and spoken behind 2 s, burst past the rate for answers but not for Spek and none when unset, speech-time estimate and back-to-back backlog.🤖 Generated with Claude Code