Skip to content

lance-graph-dt64: DR addressing as the dr module; thin-provisioned extent (restored) - #1882

Merged
AdaWorldAPI merged 5 commits into
mainfrom
ccr-45a41fb4-0j3i0l
Oct 31, 2026
Merged

AdaWorldAPI merged 5 commits into
mainfrom
ccr-45a41fb4-0j3i0l

Conversation

@AdaWorldAPI

Copy link
Copy Markdown
Owner

What

DR's addressing moves into the bind store as lance_graph_dt64::dr, lifted from dr-math-binding-probe. "DR routes, DT interprets" becomes a module boundary: dr names nothing from the binding side, and a test pins that.

In dr From
Width, Axis, Extent, Claim, MAX_RESERVED, RouteVersion dt64
RouteLog: route, retract, routed_at, claimed dt64 BindStore, which keeps its API and delegates
FrameStore, Frame, FrameRevisions, FrameRefusal; Focus, Scope, Guards::check; AxisKind, ClassAxis probe, D-DRMC-1
LocalMap, address, coordinates, rotate_yz probe, D-DRCA-1

Two changes against the probe:

  • The probe's Axis { class, kind } is ClassAxis here, because Axis is the reservation axis.
  • LocalMap and address take a tile Width instead of a fixed 64, so DR256 tiles are addressed the same way.

Not lifted:

  • relcomp, where_dr3, population, recipe and the hhtl16 folds are computations, proofs and oracles, so they belong to Quack. relcomp also needs egglog and sha2, which a contract-only crate must not.
  • dr4::route produces values, so it is execution.
  • bind_dr64 waits on the MethodSignature vs MethodRef identity question (EXO-50).

The probe is unchanged; its owner switches to dt64::dr.

Restored: the thin-provisioned extent. #1871 merged before its last two commits were pushed, so 9014fa07b (reserve tiles, claim only routed ones, DR256_GLOBAL_X64, OutsideReservation, ReservationTooWide) and plan addendum 11 never reached main. They are cherry-picked here.

Board:

Checks

  • 27 tests pass (21 binding, 6 dr); clippy (-D warnings) and fmt are clean.
  • Disable runs on 044c9cb63, each turning a test red:
guard removed failing tests
the foreign-store check 2
the source-revision refusal 1
the focus guard 1
LocalMap width read as 64 1
address width read as 64 1
a use crate:: in dr.rs (the boundary test) 1

The thin-provisioning disable runs from #1871's follow-up (4 guards) still hold.


Generated by Claude Code

… routed ones

A fixed W64/W256 width per axis capped a class at 256 identities, so
Jina's 7,620 nodes could not be bound. Each axis is now a reservation
of DR64 or DR256 tiles (Axis { tile, tiles }), up to the whole DR256
identity space (DR256_GLOBAL_X64, 65,536 positions). Reserving claims
nothing: BindStore::claimed(class, version) counts the identity tiles
that hold a routed position, and its cost (tile_bytes per claimed tile)
is stated, never allocated. route refuses a position past the class's
identity reservation (OutsideReservation); an axis wider than it can
address is refused (ReservationTooWide), never truncated.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAqKMPh2Seh2WLccCksAdY
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAqKMPh2Seh2WLccCksAdY
…r-math-binding-probe

"DR routes, DT interprets" as a module boundary inside the bind store:
dt64::dr holds positions, tiles, reservations (Width, Axis, Extent,
Claim), the route log (RouteLog: route, retract, routed_at, claimed),
the frame witness (FrameStore, Frame, FrameRevisions, FrameRefusal),
the guards (Focus, Scope, Guards::check), the typed axis (AxisKind,
ClassAxis) and the global->local pointer table (LocalMap, address,
coordinates, rotate_yz). It names nothing from the binding side; a test
pins that. BindStore keeps its API and delegates to RouteLog.

Lifted from dr-math-binding-probe (D-DRMC-1, D-DRCA-1) with two
changes: the probe's Axis { class, kind } is ClassAxis here, and
LocalMap/address take a tile Width instead of a fixed 64. DR's
computations (relcomp, WHERE, folds, dr4 values) stay out: they are
executions, Quack's. The probe is unchanged; its owner switches.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAqKMPh2Seh2WLccCksAdY
…1869 and #1871

PR_ARC_INVENTORY gains the #1869 and #1871 entries that were missing,
LATEST_STATE records main after them, STATUS_BOARD moves the
D-DT64-6-BS row, and a board entry records the dr module. Addendum 12
also records that #1871 merged without the thin-provisioned extent,
which this branch restores.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAqKMPh2Seh2WLccCksAdY
@coderabbitai

coderabbitai Bot commented Oct 31, 2026 •

Copy link
Copy Markdown

Important

  • Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

Run configuration
  • Source mode: Auto
  • Repository: Unavailable
  • Branch: Unavailable
  • File: No configuration file from this repository applied
  • Commit: Unavailable
  • Layers: Organization UI
  • Review profile: CHILL
  • Plan: Advanced
  • Run ID: f4ae9c9a-7b4d-47fe-9532-89c0a8f21475
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

regenerate-and-diff failed on DUPLICATE ID D-DT64-6-BS: the new entry's
heading reused the bind-store entry's id. entries_index.py keeps an
existing row's id, so the stale row was removed before regenerating.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XAqKMPh2Seh2WLccCksAdY
@AdaWorldAPI
AdaWorldAPI marked this pull request as ready for review October 31, 2026 02:53
@AdaWorldAPI
AdaWorldAPI merged commit ce70ad0 into main Oct 31, 2026
13 checks passed
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 31, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-31T02:53:46.423014Z e99b360 Draft marked ready
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

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