Update output stereo state on write - #31
Open
relative23 wants to merge 1 commit into
Open
Conversation
relative23
pushed a commit
to relative23/oscmix-desk
that referenced
this pull request
Aug 17, 2026
michaelforney/oscmix#31 Rebased onto upstream main, which turns out to be the same commit as our pin -- upstream has not moved since 2025-06-01, so the patch applies directly. The binary built from the PR branch is byte-identical to the one the measurement was taken against. Merges clean, 1 commit, +28/-1.
relative23
pushed a commit
to relative23/oscmix-desk
that referenced
this pull request
Aug 17, 2026
Six places opened their own socket and knew 127.0.0.1. The tidying is the smaller half; a test asserts no seventh appears outside backend.py, because each one makes the own-state-path option the roadmap wants to keep open more expensive. The half that matters is Traits. The dependency on oscmix's *behaviour* was invisible: it lived in the control flow as timing constants with nothing naming what they were for. Three declared properties now, each checked rather than believed: - reports_link_state_on_write = False. The sole reason LINK_ECHO_TIMEOUT, LINK_SETTLE and LINK_SYNC_BLIND_DELAY exist. Measured by instrumenting upstream (patches/README.md): setlevel() reads stereo=0 for a pair that was just linked. - dumps_playback_matrix = False. Asserted against the recorded dump, and the reason the playback matrix is re-established not verified. - reports_unchanged_registers = False. Why a timeout on the echo is normal rather than an error. So when michaelforney/oscmix#31 lands and the pin moves, flipping one flag is the change instead of a search through the control flow -- in the order ADR 0008 fixes: bump, measure, then delete. A pure extraction otherwise: every timing decision stays with the caller, because a backend that decided timing would be the old control flow with an extra indirection. Verified beyond green tests -- 50 soak cycles, and both paths exercised on the device: the normal one (5 registers confirmed) and the blind one with UDP 8222 held.
setlevel() reads out->stereo to decide whether a /mix write addresses one channel or the pair. With .set=setbool that flag only changes once the device echoes the register back, so a /mix sent right after /output/N/stereo takes the unlinked branch and never writes mix[1]. setinputstereo() already updates inputs[] on write; do the same for outputs. Nothing changes on the wire.
relative23
force-pushed
the
cache-sync-output-stereo
branch
from
August 26, 2026 18:44
d105a87 to
56641bc
Compare
relative23
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
setlevel()readsout->stereoto decide whether a/mixwrite addresses one channel or the pair. With.set=setboolthe flag only changes once the device echoes the register back, so a/mixsent right after/output/N/stereotakes the unlinked branch and never writesmix[1].setinputstereo()already updatesinputs[]on write; this does the same for outputs. Nothing changes on the wire.Observed on a UCX II by logging
out->stereoinsetlevel()after/output/5/stereo 1followed by/mix/5/playback/1: 0 before the patch, 1 after.