Issues

Spiral fit dies at iteration 0 on Windows: randint bound exceeds int32

#1594

· Pizzakira opened

on Aug 24, 2026

vc_grow_seg_from_seed: remote normal-grid store still requires a local metadata.json, and dies with an uncaught std::runtime_error without it

#1588

· flummoxjr opened

on Aug 24, 2026

ink_detection: the on-disk volume cache settles at ~2x --cache-max-gb because the budget is enforced twice

#1586

· flummoxjr opened

on Aug 24, 2026

get_ink_metrics.py: the fork guard is dead code, and on Windows every fold dies after the expensive prep

#1583

· EazyHood opened

on Aug 24, 2026

ink_detection: surface representation-family provenance for prepared inputs

#1582

· nerln opened

on Aug 24, 2026

PHercParis4: authoritative frames for historical PointCollections and the published umbilicus

#1569

· Kotoro78 opened

on Aug 22, 2026

vc_render_tifxyz: --prefetch-remote / --remote-url help text describes a staged on-disk cache that is never written

#1555

· Bullo27 opened

on Aug 21, 2026

Conflicting scroll_source metadata for ink/1667/w013_20240304141531_2um

#1551

· joe-needham opened

on Aug 21, 2026

20260701183124-w010-027 (PHercParis4): row 680 rides ~97 vx of dark CT with no predicted surface near z ≈ 16400–16520 (sheet-switch claim withdrawn)

#1549

· tonclap opened

on Aug 21, 2026

[data QA] PHerc0139 w045/w046 duplicate substantially the same traced surface

#1547

· nerln opened

on Aug 20, 2026

new_model fit is structurally unavailable on published catalog lasagna artifacts: shell-dir-crop needs the absent init_shell_dir, cylinder_seed needs umbilicus_json, which ships as an empty string

#1540

· Bullo27 opened

on Aug 19, 2026

umbilicus: remaining consumers and UI of the shared resolver

#1538

· skyward7187 opened

on Aug 19, 2026

Navigation Menu

Uh oh!

Uh oh!

All issues

Spiral fit dies at iteration 0 on Windows: randint bound exceeds int32

vc_grow_seg_from_seed: remote normal-grid store still requires a local metadata.json, and dies with an uncaught std::runtime_error without it

ink_detection: the on-disk volume cache settles at ~2x --cache-max-gb because the budget is enforced twice

get_ink_metrics.py: the fork guard is dead code, and on Windows every fold dies after the expensive prep

ink_detection: surface representation-family provenance for prepared inputs

PHercParis4: authoritative frames for historical PointCollections and the published umbilicus

vc_render_tifxyz: --prefetch-remote / --remote-url help text describes a staged on-disk cache that is never written

Conflicting scroll_source metadata for ink/1667/w013_20240304141531_2um

20260701183124-w010-027 (PHercParis4): row 680 rides ~97 vx of dark CT with no predicted surface near z ≈ 16400–16520 (sheet-switch claim withdrawn)

[data QA] PHerc0139 w045/w046 duplicate substantially the same traced surface

new_model fit is structurally unavailable on published catalog lasagna artifacts: shell-dir-crop needs the absent init_shell_dir, cylinder_seed needs umbilicus_json, which ships as an empty string

umbilicus: remaining consumers and UI of the shared resolver

Uh oh!

Uh oh!

Search results