Kairos Status

v0.2.345 ← Kairos
measured …

Still being read An article is read by a lane, then separately written up into findings rows. Only the second reaches the guide, so there are two bars. Only corpora with work left appear here; they are found on disk, so a new one shows up on its own.

DJ editorial sweep

dj-product-updates
READ
7,071 78%
WRITTEN UP
8,229 91%

of 9,082 articles in the queue

9,082QUEUE
7,071FETCHED
7,071READ
8,229IN THE GUIDE DB

Serato The Drop, the Traktor blog, rekordbox editorial and Beatportal. The largest queue on the board, and the only corpus still being fetched.

512 of those 8,741 are sitting only in the dj-product-updates lane and have never reached the main one, which is where the guide is built. The audit tool reads the main lane alone, so it reports 8,229.

1,963 of the sources its rows cite have no reading saved under that name, so nothing can check those anchors against the article. Every lane that fabricated rows this month looked exactly like this.

Finished, and kept on the record7 corpora read end to end and written up. No bar, because there is nothing left to watch; the numbers stay so a corpus that is done does not read as one nobody did.

The lanes What each one is doing, taken from its own commits rather than guessed from a folder name.

keyframe

IDLE

research: review current harmonic demos and analysis evidence

9 hr ago judged by its last commit

  • 00:24 coordination: Claude's Brief 001 report and review, measured in
  • 00:23 sandbox: tonnetz counts all seven triads, the diminished include
  • 00:18 overnight: morning summary; BACKLOG 5 (Perlego anchor spot-check
  • 00:18 overnight G: first load measured on the live site (724 KB brotli

dj-product-updates

RUNNING

research: review remaining early Serato media

6 min ago judged by its last commit

7 uncommitted change(s) in the tree.

  • 08:49 research: review remaining early Serato media
  • 08:42 research: verify pending Serato tutorial media
  • 08:29 research: verify Serato media and timeline sources
  • 08:02 research: disposition Serato sample pack promo

djmag

STOPPED

research: sweep DJ Mag features 1041-1070, 136 findings

3 days ago judged by its last commit

2 uncommitted change(s) in the tree.

  • 02:44 research: sweep DJ Mag features 1041-1070, 136 findings
  • 02:29 research: sweep DJ Mag features 1001-1040, 161 findings
  • 02:14 research: sweep DJ Mag features 0961-1000, 220 findings
  • 02:01 research: sweep DJ Mag features 0921-0960, 168 findings

gemini

IDLE

research: review current harmonic demos and analysis evidence

9 hr ago judged by its last commit

1 uncommitted change(s) in the tree.

  • 00:24 Merge branch 'main' of /Users/joelsamin/Projects/keyframe into g
  • 00:24 coordination: Claude's Brief 001 report and review, measured in
  • 00:23 sandbox: tonnetz counts all seven triads, the diminished include
  • 00:18 overnight: morning summary; BACKLOG 5 (Perlego anchor spot-check

headings

STOPPED

headings: batch 099, PRODUCTION 146, JOIN 39, SET-ROOM 10, LEARNING 6, RECORD 2 (warn: one

2 days ago judged by its last commit

  • 19:02 headings: batch 099, PRODUCTION 146, JOIN 39, SET-ROOM 10, LEARN
  • 18:58 headings: batch 098, PRODUCTION 219, GEAR 36, SET-ROOM 15, JOIN
  • 18:52 headings: batch 097, JOIN 80, PRODUCTION 50, SET-ROOM 49, RECORD
  • 18:46 headings: batch 096, GEAR 139, JOIN 65, PRODUCTION 31, SET-ROOM

inthemix

IDLE

research: sources register and ranked future reading list (Joel, 2026-09-22)

14 hr ago judged by its last commit

1 uncommitted change(s) in the tree.

  • 18:33 Use explicit CDX prefix matching
  • 12:26 Deepen inthemix NLA reach map
  • 12:13 Map surviving inthemix archives
  • 06:36 reading list and backlog: the inthemix archive reconstruction, m

ra

STOPPED

the 205 written RA articles that no findings file covers

3 days ago judged by its last commit

1 uncommitted change(s) in the tree.

  • 21:10 the 205 written RA articles that no findings file covers
  • 20:32 RA sweep batch 58
  • 20:30 RA sweep batch 57
  • 20:25 RA sweep batch 56

rbma

STOPPED

research: write RBMA batch

3 days ago judged by its last commit

1 uncommitted change(s) in the tree.

  • 22:59 Complete RBMA findings corpus
  • 22:54 Add RBMA findings batch 17
  • 22:48 Add RBMA findings batch 16
  • 22:41 Add RBMA findings batch 15

gemini deep research

STOPPED

One report an hour against a 52 topic queue. Reading for a person, not rows for the guide: it writes no findings and lands nothing.

41 hr ago judged by the newest report on disk

52 reports written. 1 of 52 topics still queued.

Could not determine the exact informal financial transfer volumes and cryptocurrency remittance flows supporting clandestine underground electronic pr

Could not quantify longitudinal career termination rates caused directly by chronic tinnitus or sound-induced hyperacusis across working electronic cl

  • 52 what nobody teaches
  • 51 the global scenes the guide barely covers
  • 50 sampling and edits legally
  • 49 scene documentaries and films

Waiting on you

The guide The atlas as it stands: 419 written folders and 135 proposed, built from 38,604 findings. The bar is each heading's share of those findings and the number beside it is how many sources feed it. The ATLAS tab opens them.

THE SET AND THE ROOM
6,567696 src
PRODUCTION
5,822591 src
THE DJS
5,779554 src
CAREER AND THE INDUSTRY
4,570590 src
WHERE IT CAME FROM
4,428542 src
SCENES AND GENRES
2,873459 src
THE JOIN
2,771439 src
THE GEAR
2,320386 src
THE RECORD
1,638342 src
LEARNING AND PRACTICE
1,497441 src
THE ARGUMENTS
1,403397 src
NOT GUIDE MATERIAL
1,099207 src
THE FLOOR
781215 src
NOT YET HEADED
32421 src
Findings in the atlas38,604
Written folders419
Proposed by the research itself135
Folders resting 60% on one source109
Findings in no folder yet3,457

The grouping run that produces the "no research behind it" list was made 9 hr ago over 38,270 findings. There are 50,437 now, so its answer predates roughly 12,167 rows. Rerun it before trusting the gap list.

The app

Versionv0.2.345
Last code change2026-09-26
Versions shipped in the last week25
Scripts / lines of JavaScript24 / 21,407
Sandbox pages90
Page a phone downloads2.2 MB, 0.70 MB zipped

The last five versions:

  • v0.2.345 2026-09-26

    KEY detail crash (Joel: "fix a crash in the V1 KEY tab"): drawDetail ran a bare .forEach(glossApply) over the

  • v0.2.344 2026-09-26

    DEMOS in DEBUG (Joel: "make it accessible through a demo menu on the debug menu of kairos so I can see it on m

  • v0.2.343 2026-09-25

    GLIDE parked (Joel: "park the glide function for now. I will want something like it in eventually but that mod

  • v0.2.342 2026-09-25

    a FRAME tap no longer drops the booth to V1 (Joel: "why is it defaulting to the old v1 booth view whenever i c

  • v0.2.341 2026-09-25

    FRAME in DEBUG (Joel: "this can't just be for demos it has to be for the actual app"): the app at a phone's re

The library 27 reading banks on disk, all swept, none with an undecided section. Of the 6,612 findings those banks produced, 568 are attached to an entry in the old inventory; that inventory is retired, so read it as a rough measure of the banks alone, not of the 38,604 findings in the atlas.

Every outside source, from research/SOURCES.md: 57 read, 19 partly read, 1 on disk and unread, 3 still to verify. ✓ read from the source · ◐ partly read · — held but not read · ? named in a bank, reading not yet confirmed. The corpora count as one row each; their findings run into the thousands.

Books

✓Last Night a DJ Saved My Life (French translation)READBill Brewster & Frank Broughton · READ in part (ch 5, 6, 9, 12, 13, 16), then retired: every back-translated quote replaced from the English text
✓Last Night a DJ Saved My LifeREADBill Brewster & Frank Broughton · READ, second pass
✓Last Night a DJ Saved My Life: Deep CutsREADBill Brewster & Frank Broughton · READ in full: all 21 chapters with printed page numbers, 2026-09-06, plus a Levan pass across all 21; then read again independently by ChatGPT/Codex overnight 2
✓The Record Players: DJ RevolutionariesREADBill Brewster & Frank Broughton · READ in full: all 46 interviews and the introduction, 2026-09-24, in 18 batches gated one by one (every anchor verbatim); earlier use was Levan-only
✓Keep On Dancin'READMel Cheren · READ
✓How to DJ (Properly): The Art and Science of Playing RecordsREADFrank Broughton & Bill Brewster · READ
✓Dance Music ManualREADRick Snoman · READ, ch 35 and appendices A-D, J
✓Sweet Anticipation: Music and the Psychology of ExpectationREADDavid Huron · READ, chapters 1, 9, 10, 13, 14, 15
✓The Geometry of Musical RhythmREADGodfried Toussaint · READ
✓Tuning, Timbre, Spectrum, ScaleREADWilliam A. Sethares · READ
✓Unlocking the GrooveREADMark J. Butler · READ
✓Sonic BodiesREADJulian Henriques · READ
✓Acoustics and PsychoacousticsREADDavid M. Howard & Jamie Angus · READ
◐Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers, Rooms and HeadphonesPARTIALFloyd E. Toole, Sean Olive & Todd Welti · PARTIAL: sections 5.3.1, 9.6.1 and 9.6.4 only; used as a secondary account of Toole & Olive 1988 (JAES 36(3):122-142), which is still unread
✓Geometry and Topology in MusicREADMoreno Andreatta, Emmanuel Amiot & Jason Yust (eds) · READ, all three chapters (incl. Yust, Gómez; Noll skimmed)
✓Introduction to GeometryREADH. S. M. Coxeter · READ
✓A Geometry of MusicREADDmitri Tymoczko · READ
✓"Behind the Groove" (Larry Levan interview), Collusion No. 5READSteven Harvey · READ
✓journal articles, incl. the JPMS 2011 real-estate paperREADTim Lawrence · READ
✓How to Make Great Music MashupsREADPaul Zala · READ
✓The Remix ManualREADSimon Langford · READ
✓DJing For DummiesREADJohn Steventon · READ
✓DJ SkillsREADStephen Webber · READ, incl. ch 14 and 19 addendum
✓Beyond BeatmatchingREADYakov Vorobyev & Eric Coomes · READ
✓DJ Culture in the MixREADAttias, Gavanas & Rietveld (eds) · READ
✓Everyday TonalityREADPhilip Tagg · READ

The twelve design sources (DESIGN-NOTES.md)

◐Bret Victor, "Magic Ink" (2006)PARTIALPARTIAL
◐Don Norman, Things That Make Us Smart (1993)PARTIALPARTIAL
◐Matthew Butterick, Practical TypographyPARTIALPARTIAL
◐Flight-deck alerting, "the dark cockpit"PARTIALPARTIAL
◐Karl Gerstner, Designing Programmes (1964)PARTIALPARTIAL
◐Josef Müller-Brockmann, "Grid and Design Philosophy" (1981)PARTIALPARTIAL
◐Herb LubalinPARTIALPARTIAL (the bank says READ)
✓Edward Tufte, The Visual Display of Quantitative InformationREADREAD (cited pages)
◐Jeff Johnson, Designing with the Mind in Mind (2010)PARTIALPARTIAL
◐Alan Cooper, About Face, 4th ed. (2014)PARTIALPARTIAL
◐Colin Ware, Information Visualization, 2nd ed. (2004)PARTIALPARTIAL
✓Jef Raskin, The Humane Interface (2000)READREAD (cited pages)

Magazines (the printed publications)

◐MuzikPARTIALcomplete issues 1995-2003 as PDFs · PARTIAL, by theme
◐Mixmag, the magazinePARTIALprinted issues, distinct from mixmag.net: 3 issues in the 2026-09-10 Gemini notebook; Issue 664 (Lilly Palmer) digital edition in Downloads; UQ's ProQuest holds Dec 1994 and Jan 1997 to Dec 2000 · PARTIAL: 3 issues; Issue 664 NOT READ; ProQuest issues requested from Mark
◐DJ Mag, the magazinePARTIAL1 issue in the 2026-09-10 Gemini notebook · PARTIAL: 1 issue

Websites and online archives

✓DJhistory.comREADinterview and article archive · READ
◐Resident AdvisorPARTIALthe written archive: 3,420 features and city guides (1,071 Exchange podcasts and 252 films are audio and video, on the backlog to transcribe) · PARTIAL: 729 written articles still queued in research/ra-corpus/TODO-SWEEP-2.txt
✓mixmag.net, the websiteREAD4,778 editorial features (25,473 news items untouched) · READ: all 4,778 features, landed 2026-09-23
?djmag.com, the features archiveVERIFY1,070 features and cover features, 2.0 million words · CATALOGUED in full, NOT YET WRITTEN UP
◐Red Bull Music Academy lecturesPARTIAL438 lecture transcripts, practitioner voices · PARTIAL: 25 of 438 read
◐DJ TechTools and Digital DJ TipsPARTIAL2,342 instructional articles, 2006-2026 · PARTIAL, 4 of 2,342 in main
✓Tutorial sites surveyREADCrossfader, DJTT, DDJT, Point Blank, Mixed In Key, Ableton Learning Music · READ as a survey
✓Editorial sitesREADRA, Mixmag, DJ Mag and others for culture · READ as a pass
✓DJ academiesREADsix syllabuses (Crossfader, DDJT, Vibes, Point Blank, two universities) · READ; Point Blank BA, DDJT modules, NextGen returned 403
✓Ishkur's Guide to Electronic MusicREADv3, music.ishkur.com · READ
✓Albert Bregman's McGill siteREADauditory scene analysis exposition · READ (the book itself is not)
✓Dmitri Tymoczko's Princeton papersREADthe free papers behind A Geometry of Music · READ (the book too, see A)
✓Richard CohnREAD"A Platonic Model of Funky Rhythms", MTO 22.2, 2016 · READ
✓In Sheep's Clothing HiFiREAD"A look inside the hi-fi equipment behind Paradise Garage" (2023), insheepsclothinghifi.com · READ once; registered 2026-09-23 from the conversation audit
✓WikipediaREAD"Wikipedia:Signs of AI writing" · READ; registered 2026-09-23 from the conversation audit
✓Design reference setREADWCAG 2.2; Pickering, Inclusive Components; Butterick, Practical Typography; Apple HIG; Material Design 3; FAA Human Factors Criteria for Displays (2007); Few, Information Dashboard Design (2006); Nielsen Norman Group · Mixed, by file (edition pass 2026-09-25): READ, WCAG 2.2, Apple HIG, Material 3, FAA criteria; PARTIAL, Few (chapter 4 from source, chapter 3 from the contents

Papers and documents held as files

✓Vliegen & Oxenham 1999, JASA 105:339READREAD
✓Vliegen, Moore & Oxenham 1999, JASA 106:938READREAD
✓Deeks, Gockel & Carlyon 2013, JASA 133:377READREAD
✓van de Par & Kohlrausch 1997, JASA 101:1671READREAD: came unrequested with Vliegen & Oxenham, and sets the limit on what they can be used for
✓Lustig & Tan 2019, "All about that bass: audio filters on basslines determine groove and liking in EDM"READREAD 2026-09-05
✓Laroche & Dolson 1999, "Improved phase vocoder time-scale modification of audio", IEEE Trans. Speech and Audio Processing 7(3):323-332READREAD in full, all ten pages, 2026-09-26
✓Driedger, Muller & Ewert 2014, "Improving time-scale modification of music signals using harmonic-percussive separation", IEEE Signal Processing Letters 21(1):105-109READREAD in full, 2026-09-26
✓Grofit & Lavner 2008, "Time-scale modification of audio signals using enhanced WSOLA with management of transients", IEEE Trans. Audio, Speech, Lang. Proc. 16(1):106-115READREAD in full, 2026-09-26 (chart values approximate)
✓Bello et al. 2005, "A tutorial on onset detection in music signals", IEEE Trans. Speech Audio Proc. 13(5):1035-1047READREAD in full, 2026-09-26
◐Donaldson 2011, "Extending the phase vocoder with damped sinusoid atomic decomposition of transients", MA thesis, McGill (supervisor Depalle)PARTIALPARTIAL: front matter and chapters 1, 2, 4, 5, 6 read; chapter 3, appendices and references not
◐Pike, Surrey PhD, "Timbral constancy and compensation for spectral distortion caused by loudspeaker and room acoustics" (2015 or 2016)PARTIALPARTIAL: 581 pages, selected chapters read (some as text, some as images); cites Toole & Olive 1988 only once
✓Macnamara et al. 2014, deliberate practiceREADREAD
✓Sweller et al. 2019, cognitive load theory twenty years onREADREAD
✓Bjork & Bjork 2020, desirable difficultiesREADREAD
✓Kudaravalli 2024, reminiscence bumpREADREAD
✓Equipment manuals: Allen & Heath Xone:92 (guide and block diagram), Xone:96; Pioneer DJM-900NXS2, DJM-A9, DJM-V10, DDJ-FLX4, CDJ-3000; Rane Seventy, Seventy-Two MkII; Denon X1850 Prime; PlayDifferently Model 1READREAD for DJ-NOTES; CDJ-3000 read verbatim on key 2026-09-13
—Mixmag Issue 664 (Lilly Palmer), digital editionNOT READNOT READ; on the reading list
✓Found free for the ASA pass (UQ-REQUEST, 2026-09-08), eight papers, resolved 2026-09-25 against the banks. No local copy of any.READREAD: van Noorden 1975, Oh 2022, Siedenburg 2026, Williams 2025, Honingh 2015, McAdams 2019 ch 2 (each shows page numbers, quotes or method detail). PARTIAL: Jh
✓Acoustilog Inc., "Disco Acoustics: The Paradise Garage", plus ten page scans (gif, png) of Fierstein & Long, "State-of-the-Art Discotheque Sound Systems", AES convention paper, c. 1980-81READREAD, per GARAGE-AES-NOTES.md; one scan checked by the agent 2026-09-25 (disco5b.png, its T60 table matches the notes). The saved html holds only site menus. Th
?Schreiber & Müller 2018, TempoCNN, the deepsquare-k16 modelVERIFYUSED IN THE APP, not read as a paper: k16 chosen over k4 on a 39/40 benchmark against Joel's tagged library (Joel, 2026-08-08, session 8adf7f83 #6862: "go with
✓Yamamura's Dojo, Super Mario Maker 2 (Nintendo 2019), via the Super Mario WikiREADREAD: all 50 lessons, mined into research/GAME-DESIGN-NOTES.md (play panel, BACKLOG 42). Nintendo's dialogue; the first game-design source
✓Game design as pedagogy: about 100 source copies (Juul's papers, Gee, Malone and Lepper, Andersen 2012, Lomas 2013, Ryan 2006, Swink, Koster, Iwata Asks x10, Hayashida, Portal commentary, Cook, Ableton, Rhythm Heaven, Rhythm Doctor, Harmonix, DJ Hero, Rhythm Workers studies, Android latency)READREAD, digested into research/GAME-DESIGN-NOTES.md §5 to §11; each file's header gives its origin; one preprint (Jamey 2024) is a private research copy only

Secondary research by another AI

?Gemini Deep Research (the kf-lanes/deep-research lane)VERIFY19 reports so far, one per topic from a 52-topic queue, written from live web searches, each with its sources listed · UNVERIFIED SECONDARY: nothing in them has been checked against the pages they cite, and report 07 carries at least one likely misattribution (Vince Aletti). Eve

The repository Where the folder has grown heavy, and what would make it lighter without losing anything.

Whole folder751 MB
Research454 MB
Git history, of which loose objects188 MB / 63 MB
Research documents no decision covers48
Old chat exports, also uncovered130
Typefaces re-sent on every deploynone, served as files

Done on 21 September. The git history was packed, which returned 92 MB and changed nothing else. The hosted page now serves its typefaces as files instead of carrying them inlined, so a version bump no longer re-sends 422 KB; the single-file build still inlines everything, as it has to. The ledger now lists the documents it has never ingested, which previously vanished rather than reading as undecided. Those still need a sweep or a reason each, and they include the one the project instructions tell you to read before redesigning any surface:

  • research/prototypes/KAIROS-HANDOVER.md
  • research/prototypes/NeedleDrop_Handoff.md
  • research/prototypes/README.md
  • research/prototypes/RHYTHM-HANDOVER.md
  • research/prototypes/SESSION-HANDOVER-2026-08-25.md
  • research/prototypes/designing-kairos.md

Five faults found in the audit tool, and fixed Each was verified against the lanes' own git history and files, then corrected in tools/audit.py on 21 September. On 24 September audit.py was retired and this board became the one measurement. Kept here because the shape of the error is worth more than the correction.

  • RA written up read 501, not 601

    The tool counted findings only in the main repo, so a hundred RA articles written up in the gemini lane were invisible. It now counts every worktree, and says plainly how much has not yet reached the lane that builds the guide.

  • The catalogue-index warning pointed the wrong way

    It always said the index was behind. It holds 3,828 rows against 2,268 readable files, so it is 1,560 ahead of what has actually been read and 425 behind the file count. It now says which, and against what.

  • DJhistory was accused of claiming 57 mined

    That sentence is a running total from batch 26. The pass stopped writing it and has since passed batch 54. A progress document appends, so a figure it stopped restating is history. The tool no longer treats one as a live claim.

  • DJhistory was accused of a wrong queue size

    The document carries a dated correction saying 96 was wrong and the manifest holds 99. It was being flagged for the error it had already fixed. A document that corrects itself is no longer counted as disagreeing.

  • Zero sections undecided does not mean the research is done

    That ledger lists 141 sources and not one of them is a corpus. It covers the documents and the banks only, 6,612 of the 50,437 findings in the database.

Every number on this page is measured from files and from git at the moment shown at the top. Nothing is read from a progress log or a ledger, except where the page says a document claims something, and then the claim sits beside the measurement. Regenerate with tools/status.sh, which remeasures, rebuilds and redeploys.