Skip to content

docs: the roadmap's section 7 records the maintainer's decisions - #83

Merged
tap merged 1 commit into
mainfrom
claude/sample-rate-solutions-comparison-mqc190
Oct 12, 2026
Merged

tap merged 1 commit into
mainfrom
claude/sample-rate-solutions-comparison-mqc190

Conversation

@tap

@tap tap commented Oct 12, 2026

Copy link
Copy Markdown
Owner

What this changes

docs/ROADMAP.md section 7 now records each question as it was asked, with the maintainer's decision under it (from an interview on 2026-10-12). The passages that depended on those answers now state the decided version.

Q Decision Where it lands
Q1 USB asynchronous mode, with the Pico 2 (RP2350) as the asynchronous-mode device on its own crystal. The bench runs twice, so timestamps ship in 1.0. A11 gains the Pico 2's UAC2 asynchronous-mode firmware, which is not in the tree today. B3 becomes "on paper, then on the same bench". C1 lands in 1.0; if the second pass slips, it becomes the first 1.1 item. The verdict rule (section 6) now includes the timestamp servo in async's 1.0.
Q2 No outside consumer links a C ABI. MAJOR is the ABI. FetchContent and add_subdirectory only. C3: no separate ABI version (the bit-packed tap_sr_<engine>_version() is the probe); install()/find_package deferred.
Q3 Confirmed as landed in A1 (#80). Section 7 only.
Q4 Tag DspTap independently and assert its minimum version. A9
Q5 Probably our own Pico 2 firmware (the SDK's default). This is the maintainer's best recollection, unconfirmed, and is recorded that way. Section 7 only. A3 landed regardless (#81).
Q6 Hard gate, ±3 %. A8

The status line now says execution started on 2026-10-10 on the roadmap's recommendations and that the answers confirmed them.

Why

The original request for #79 asked for the section 7 answers to be recorded as decided. The answers given there were placeholders, so A1, A3 and A4 went ahead on the roadmap's recommendations. This PR records the real answers.

Verification

  • scripts/check_retired_words.py: clean.
  • Every passage that branched on a Q1/Q2/Q4/Q6 answer (A8, A9, A11, B3, C1, C3, section 6) was read and updated to state the decided version.
  • Docs only. CI's book and style jobs are the remaining gate.

Notes for the reviewer

  • Q1 adds a firmware item. It assumes a Pico 2 USB audio firmware that enumerates as a UAC2 asynchronous-mode device. Nothing in the tree does that today (pico2_cyccnt and pico2_dualcore are measurement firmwares), so A11 now names it.
  • Q5 is recorded as tentative.

🤖 Generated with Claude Code

https://claude.ai/code/session_016GKzq1VSdcMYCRMhddiMaf


Generated by Claude Code

Q1: the bench is USB asynchronous mode, with the Pico 2 as the
asynchronous-mode device on its own crystal (its USB audio firmware joins
A11), and it runs twice, so the timestamp servo is a 1.0 item (B3, C1, the
verdict rule). Q2: no outside consumer links a C ABI; MAJOR is the ABI;
FetchContent and add_subdirectory are the only supported consumption at
1.0 (C3). Q3: as landed in A1 (#80). Q4: DspTap tagged independently, the
minimum asserted (A9). Q5: probably our own Pico 2 firmware, by the
maintainer's recollection, unconfirmed; A3 landed regardless (#81). Q6: the
host ratchet is a hard gate at ±3 % (A8). Each question stays as asked,
with its decision beneath it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016GKzq1VSdcMYCRMhddiMaf
@tap
tap merged commit 17298d3 into main Oct 12, 2026
31 checks passed
@tap
tap deleted the claude/sample-rate-solutions-comparison-mqc190 branch October 12, 2026 15:45
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants