Issue renumbering, 2026-09-08

When vibe-qc moved to its own repository, the 188 open issues were moved from mpei/vibeqc (project 19) to mpei/vibe-qc (project 34). GitLab’s move creates a new issue with a new number and closes the original with a “moved to” note, so every issue number changed.

The mapping is not derivable from the numbers: moves were applied in ascending old order while new numbers were assigned in sequence, so the ordering inverted – issue 8 became 188. There are 188 distinct offsets across 188 issues.

This table exists because those old numbers are cited elsewhere and cannot be followed otherwise:

• ~125 (#NNN) references in CHANGELOG.md, and more in commit subjects;

• the JCC manuscript’s defect provenance, via the article repository’s handovers;

• bugctl grant and landing records.

Closed issues were deliberately NOT moved. All 776 of them remain on project 19, which is retained as the archive; their numbers are unchanged there.

A machine-readable copy of this table, plus a full pre-move snapshot of every issue including its original labels, is kept outside the repository at _split-archive/ on the maintainer’s machine.

old (project 19)	new (project 34)	title
8	188	BUG 95 [LOW] Marvin immutable release omits advertised D4 dependency
20	187	BIPOLE-PURE-FUNCTIONAL-SYSTEMATIC-OVERBINDING [HIGH]
21	186	perf-bipole-periodic-hf-hours-per-scf-iteration [HIGH]
22	185	BH9-TYPE-I-ISOMERISATION-OUTLIER [MEDIUM]
43	184	BUG 027 [HIGH] GFN2-xTB total-energy residual vs xtb 6.7.1: +245.7 mHa on adenine after the E_re
65	183	Multi-k GAPW fails to converge on expanded cells (missing global-BZ Aufbau and SCF density mixin
66	182	BIPOLE fold-guard known-red block: re-measured re-cut cutoffs await a dedicated edit session
74	181	[HIGH] PlanetX fleet host: memory preflight reads effective allocation 0 MB and refuses every ca
78	180	[HIGH] SCC-DFTB pi-conjugated systems still fail at v0.15.130 (finite-T retry not the default)
81	179	Periodic GDF total-energy offset vs PySCF: DF Coulomb-tail completeness deficit on dense-core ce
82	178	BIPOLE multi-k total energy vs CRYSTAL 23: exchange finite-size gauge (+6.43e-1 Ha on MgO)
84	177	Multi-route band-gap comparison open: LiF pob-TZVP-REV2 (2,2,2) vs GPAW
85	176	Bands/DOS: Si non-converged; 2D h-BN fails closed; negative indirect gap from per-k occupations
86	175	Bulk equation-of-state observables (a0, B, band gap): Si/MgO/LiF scans non-converged or runaway
87	174	Brillouin-zone mesh convergence: unguarded 38.9 GiB Ewald-cache allocation; k=6 non-converged
89	173	DFT+U: NiO target fails with a memory-allocation failure
91	172	Mixed density fitting for heavy cores: overlap matrix loses positive definiteness
92	171	Space-group symmetry demonstration: NaCl reduction exceeds a 1225 GB GDF memory guard
94	170	DF-CCSD(T): explicit DF-vs-conventional pair comparison on the corrected fitting path pending
95	169	CPCM/COSMO cross-code validation: compare by solvent-response components, not raw solvated total
97	168	MgO cluster cross-code RHF consensus contract open at 1.1e-1 Ha
99	167	Native D4 backend: parity coverage H-Ne only; external dftd4 remains default
116	166	BIPOLE-TERMINAL-REBUILD-WITHDRAWS-CONVERGENCE [HIGH]: non-incremental post-loop Fock rebuild dis
117	165	PAPER-ROW-CARRIER-REFUSAL-OUT-DROPS-RUNG-FIELDS [LOW]: an rc-9 .out header reports ‘cutoff=shipp
120	164	CARRIER-PER-CASE-TIMEOUT-KILLS-INSIDE-A-LONG-RESERVATION [HIGH]: 206 members lost to a ~900-1200
121	163	BIPOLE-VS-GDF-CROSSROUTE-440MHA-ON-DIAMOND [HIGH/P1]: same code, same job, same runtime, Coulomb
127	162	[HIGH] GFN2-xTB parameter cache is per-host and unpinned: twin compute nodes offline -> all 78 G
135	161	PAPER-ROW-CARRIER-TYPEERROR-POST-SCF [HIGH]: band/density extraction raises ‘TypeError: only len
139	160	VALIDATION-RUNS-ENERGY-HA-CARRIES-SCF-TOTAL-FOR-CORRELATED-RUNS (reffill wave + v0.15.28/30 legs
142	159	GDF-KPOINT-COLLAPSE-ON-POB-TZVP-REV2 [HIGH]: Si-diamond RKS-PBE (2,2,2) rc 0 at -582.128 Ha, 3.4
145	158	REFERENCE-DECK-DEFECT-CPCM-RUNS-SMD: the CPCM h2o reference family (mb075/ml075*/mf075) runs SMD
147	157	RP217J-REGISTRATION-WRITES-SCF-TOTAL-FOR-CORRELATED-METHODS: the judged-registration tooling sti
149	156	CASE-METHOD-VS-DECK-METHOD-MISMATCH-DLPNO-MONOTONICITY: 8 liakos rows’ case says dlpno-ccsd whil
154	155	PERIODIC-SCF-STAGNATION-SI-K666-GDF [MEDIUM]: energy pinned 14 consecutive iterations at -582.08
160	154	MARU-PM6-NORBORNADIENE-SILENT-E-ZERO [MEDIUM, legacy BUG-024]: maru-only PM6/norbornadiene retur
161	153	MACE-NOT-STAGED-ON-FLEET [MEDIUM, legacy BUG-014]: MACE models unavailable on fleet hosts - stag
162	152	SE-CAMPAIGN-GENERIC-FILENAMES [LOW, legacy BUG-020]: calculation artifacts named input.py/out/qv
207	151	Nonorthogonal SECCM adapters feed indefinite WS overlap matrices to generalized eigensolvers
208	150	Semiempirical periodic tier vocabularies disagree; C++ PeriodicTier is dormant and the .out neve
242	149	GAMMA-CCM-UNION12-VARIATIONAL-COLLAPSE-ON-NON-MINIMAL-BASIS [HIGH]: eq-18 weighted Coulomb super
246	148	TWIN-AMD-CAPACITY-INVISIBLE: three nodes offline; ppn=128 requests are unschedulable and queued
253	147	W1-RV-DECKS-CASE-NOT-JSON-DECODED: 57 archived decks across 7 bundles die at CASE[“charge”] with
254	146	ARTIFACT-STEM-TRUNCATION-AT-DOT [HIGH]: three instances; now manufacturing a FALSE route-mismatc
296	145	Periodic GFN2-xTB energy is not invariant under a lattice translation of an input atom (85 mHa o
303	144	χ-CCM four-center SCF lacks distributed finite-translation output-cell farming
308	143	run_ccm_rhf_gdf at nrep=(1,1,1) is pathological: N=1 costs more than N=5; ~2900 core-hours burne
311	142	Agentic loop: move coordination, fences and rollout identity out of prose into a machine-checked
338	141	Periodic GFN2-xTB analytic gradient/stress are not derivatives of the periodic energy
348	140	GFN2-SECCM AES moment integrals depend on the typed torus representative
408	139	Periodic GFN2 Gamma driver rejects anisotropic supercells: ‘H0 coordination pair distance must b
409	138	SECCM broyden_eyert mixes shell charges only; tblite’s Broyden state vector also carries atomic
421	137	GFN2-SECCM converges into symmetry-forbidden / hopped SCC attractors that pass every screen (Ge
423	136	GFN2-SECCM OpenMP speedup is x0.94-1.44 on 20 threads; serial Python build dominates (87% of wal
425	135	k-route cutoff ladder unconverged at 200 bohr for dim>=2: mHa-scale tails on every 2-D/3-D dense
427	134	ascent reference runner swallows ORCA error termination (exit 0, vq completed, no energy); atom-
429	133	Gaussian-stack image selection is translation-ball, not pair-ball: a Gamma supercell wider than
432	132	GFN2 native D4 dispersion is absent for Z > 10: frozen refdata covers 8 elements, and TM totals
435	131	k-route indirect gap uses an occupancy-partition convention: 1022/1668 gapless rows report a pos
436	130	vibeqc-inhouse-dftb-screening-v1 leaves bulk_si and hbn gapless; benchmark set has no independen
437	129	graphene BZ ladder is non-convergent (K-point parity split): mesh spread is 127% of the reported
445	128	Lattice row/column orientation is silently accepted either way: three independent transposes fou
456	127	.system manifest can escape the job output directory into the checkout root (seen in a multi-fil
473	126	Validation DB: reassemble the 10 BH9 03_12R2-derived comparisons now that the species converges
475	125	SECCM iteration-count target: adenine needs 24.5x xtb (514 iterations) and the ~3x bar has no re
493	124	real-Γ direct-torus SCF has no convergence aids: 160-atom LiH slab+H2 fails to converge in 128 i
503	123	Block Davidson: wrong roots reported as converged for small n_eig, and convergence never signall
505	122	NO PUBLIC PER-K BAND ACCESSOR: every carrier re-derives frontier-band extraction and the class h
507	121	Verification is dispatched by hand: the broker records LANDED but emits no queue a distinct acto
513	120	Article sealed-reference table still fills the BIPOLE MgO row from CRYSTAL’s RETIRED coarse-mesh
516	119	CASSCF analytic gradient: compute_wz=True yields a spurious x component with d-function bases (H
517	118	Article repo commits author home paths and has no pre-commit hook; its DB validator only runs in
523	117	Citation DOIs diverge between code and article repos: emitted .bibtex sends readers to preprints
525	116	Semiempirical manuscript describes a converging periodic OMx driver that has failed closed since
526	115	Dense-core multi-k KRKS residual: -3.8 mHa vs PySCF / -4.4 mHa vs CRYSTAL survives a parity-size
529	114	Ideal-cubic Gamma DFTB0 T=0 ensemble stress is symmetry-correct but differs from dE/deps by up t
536	113	test_geomopt.py::test_run_periodic_geomopt_imports_and_runs is a deterministic BipoleFoldUnrelia
539	112	Six tests/test_out_format_snapshot.py golden tests are red on main (molecular rhf/rks/hessian/td
547	111	vq queue directory is never reaped: 16,374 specs, 99.9% terminal, delay new scheduler dispatch b
550	110	Staged ORCA UKS reference decks emit the non-keyword UPBE; the stager fix reached one of 25 copi
551	109	ORCA CCSD(T) reference for H2/STO-3G aborts in MDCI (K(C)-AO-direct not implemented) while exiti
552	108	RP248 ladder payload forces write_molden_file=True and blocks 37 of 86 cases before any chemistr
554	107	The reaction-field step is implemented twice: the Gaussian macro-iteration and the MSINDO fock_e
555	106	qcil submit_wave.py cannot submit to twin at any version: it reads release[binary], which every
556	105	S22 validation decks pass max_iter/conv_tol_energy to run_job, which accepts neither: 22 runs al
558	104	Implement the COSMO-RS/COSMOSPACE thermodynamic layer: conductor-surface record, sigma profiles,
559	103	import vibeqc raises when the native core cannot satisfy SKALA registration, breaking docs-build
567	102	vibe-qc DFT single points sit 17 to 44 uEh/atom above ORCA 6.1.1 on all six qcil wave001 DFT pai
568	101	qcil wave tooling is pinned to 0.15.45 and mis-parses 0.15.157 artifacts: validator strips .inpu
573	100	TREXIO I/O: writer and reader in vibe-qc (HDF5 and text back ends) and a TREXIO reader in vibe-v
577	1	fleet: unify the vibe-qc deployment layout across nodes (three different layouts on five hosts)
638	99	LiH/pob-TZVP-rev2 PBE multi-k GDF holds 29.4 GB peak RSS at nbasis=13 (19x the MgO/TZVP leg) on
646	98	GAPW and GDF disagree by 0.3007 Ha on LiH in the molecular limit: the no-blow-up guard overshoot
648	97	Third CWD-artifact leak class: 37 call sites pass output= a bare relative string, which the guar
649	96	Periodic GDF does not scale to a full node: T_eff saturates near 17 threads, 0.27 efficiency at
650	95	A convergence aid silently routes Gamma GDF onto the legacy fallback, which never auto-sizes the
652	94	test_uno_reference::test_orthonormal_and_occupations_vs_pyscf is red on main and absent from the
655	93	AICCM front door M2 (gamma half): pin the per-unit-cell artefact contract of the two neutral pro
656	92	χ-CCM D126: compact point-symmetry quotient for translation-pair work
658	91	bugctl: the L157 verification-brief sequence cannot work as documented (land closes the grant pr
659	90	AICCM front door M3a: wire the four-centre runner arm (method=”aiccm”, variant=”four-center”)
661	89	χ-CCM D127: compact real-torus AO point-action kernel
662	88	The RSGDF dense-core held class covers every all-electron pob-TZVP-rev2 cell, not one: 28 to 200
664	87	Multi-k GDF is over-bound by a mesh-independent Madelung-scale amount: 0.46 Ha from the Gamma fa
665	86	Fleet PySCF 2.6.2 cannot converge a periodic meta-GGA grid and silently returns a 28 mHa-wrong r
670	85	AICCM four-center external-XC front door drops grid/domain controls and certifies unsupported dr
671	84	Periodic alternative backends misexecute and misreport non-SAP initial guesses
672	83	SKALA next-release Twin runtime and scientific validation campaign
674	82	BIPOLE corrected exchange split cuts its erfc arm at cutoff_bohr with CRYSTAL’s unbounded alpha
676	81	test_multik_roks_fused_kernels_match_serial_fold is nondeterministic: fused vs serial iteration
678	80	Bare KS options still select the legacy grid on run_wb97x_d, the double-hybrid core, optimize_mo
680	79	Periodic SAD initial guess is un-normalized: tr(D S) = 5.31 for a 4-electron LiH cell (+0.73 e)
681	78	ROHF/GDF public route refuses its own default initial guess (SAD) since bbe7b9001: six tests of
683	77	χ-CCM D128: bind localized occupied action to its physical overlap
684	76	Molecular density-mode initial guesses do not preserve target charge and spin populations
686	75	Direct native SCF-with-JK entry points ignore selected initial guesses
687	74	READ initial guesses lose basis provenance and accept invalid periodic occupations
688	73	Nested SCF and ASE workflows drop the selected initial guess
689	72	DLPNO-CCSD extended residual domain is 300-400x slower than the legacy recipe and barely truncat
690	71	BIPOLE multi-k RKS/UKS discard constructed SAD and MINAO densities for HCORE
693	70	bugctl has no reopen verb: the specified REOPENED transition (reopen, drop resolution::verified,
694	69	SCC-DFTB-SECCM residual odd-MgO and slab state anomalies misattributed to GFN2 in #421
695	68	Wave submitter routes oversized periodic legs to maru, blocking its strict-order local lane
697	67	χ-CCM D129: finite-torus spatial group and occupied cocycle witness
700	66	DLPNO-CCSD: implement the Riplinger-Neese section II C term-by-term PNO-basis residual so the ex
701	65	DLPNO PNO occupation numbers use a pair density that is neither the Riplinger-Neese Eq. 23 MP2 n
702	64	CCM citation selector credits exact Ewald to MSINDO frozen truncated 1-D Madelung sum
703	63	χ-CCM D130: connect full-mesh RHF snapshots to shared native orbital sewing
704	62	χ four-center multi-k HF: half-cell translation Fock covariance is not qualified for pair reduct
705	61	GDF multi-k termination mixes D_used energy with next-Fock occupation entropy
706	60	Multi-k Ewald UHF/UKS mix selected seed densities with stale HCORE orbitals
708	59	BIPOLE UHF Gilat H-atom energy differs from its fixed-point pin by 6.03 microhartree
709	58	GFN2 feature documentation gate still requires retired exact wording
711	57	Multipole invalid-order test still rejects supported L_max=4
712	56	CASSCF backend parity fixture assumes a unique H2O basin despite roundoff-triggered 7 mHa root c
713	55	UKS SH tight-tolerance trajectory stalls from a legal spin-density seed in the unchanged native
714	54	Periodic GDF: RHF and UHF(M=1) energies of the same all-electron cell differ by 2.3 Ha at kpoint
715	53	Periodic RKS/PBE on a pob-TZVP-rev2 Ag cell allocates >80 GB in the XC build (RHF/UHF on the sam
716	52	Periodic SAD guess ignores ECPs: RHF on an ECP cell stops at higher stationary points (AgCl: -53
717	51	Shared symmetry M1: bounded group and space contracts with chi and molecular consumers
718	50	Verification independence should key on run_id (the chat), not actor_id (the brand): two chats s
719	49	M4a: flip the A-line RI correlation reference to reference=”direct” and add parallel neutral rou
720	48	run_periodic_job rejects an invalid initial_guess before the insulator-smearing guard can warn s
721	47	test_bipole_production_guards fold-truncation snapshot is stale since 6da75abe7 added the ‘Exact
722	46	run_krohf_periodic_gdf rejects initial_guess=AUTO before its ROKS functional refusal since bbe7b
723	45	2D UKS multi-k QVF density artifact falls back with a primitive-cell grid integral of 0.025 inst
724	44	#429 stage 2, Family 3: two-electron real-space builders on the pair-complete list, the Ewald nu
727	43	run_periodic_job’s published API reference documents initial_guess’s default as ‘
728	42	test_periodic_gdf_smearing_diis::test_multik_krhf_smeared_diis_occupations_are_fock_selfconsiste
729	41	Derive the analytic nuclear gradient of the COSMO FINE cavity so CFC geometry optimization is po
730	40	vq submit queues a command whose argv[0] is a vq flag: a misplaced –idempotency-key is swallowe
733	39	run_krohf_periodic_gdf rejects its own default initial_guess: test_open_shell_gdf_drivers_report
734	38	IBZ-native symmetry-broken-state guard no longer fires: test_ibz_native_refuses_a_symmetry_broke
736	37	vq submit returns rc=0 with a job id for a scheduler submission that then fails at workspace sta
737	36	run_validation_smoke.py pins EXPECTED_VERSION = 0.15.28 and refuses every release since 2026-07-
738	35	Periodic ECPs run only on the k-point GDF route: RIJCOSX, Ewald open-shell, ROHF/ROKS-on-GDF and
739	34	Periodic ECP gradients, forces and stress are unimplemented: no dV_ECP/dR lattice kernel
740	33	CASCI/MRCI/CASSCF/NEVPT2/CASPT2 and OVGF still refuse an ECP reference
741	32	QVF DOS/PDOS/COOP/COHP payload is skipped on ECP cells: it rebuilds H from bare Z with no V_ECP
742	31	Periodic ECP energies have no cross-code parity pin: add a PySCF.pbc check for the GDF route
743	30	Basis registry follow-ups after the ECP unification: def2-sv, cc-pVnZ-PP, alias resolution
744	29	CFC segment coarsening yields an indefinite CPCM A matrix at ~1 grid spacing in 6
745	28	CPCM gradient hard-codes switching_sigma_bohr=0.5 while SolventModel exposes it
746	27	Assemble the CFC energy gradient: cpcm_gradient still refuses a FINE cavity
748	26	periodic_runner.py:12847 reads _last.delta_e on a dict, so the SCF diagnostic raises AttributeEr
749	25	Roughly forty validation cases request combinations vibe-qc correctly refuses, so the suite repo
751	24	True multi-k periodic GDF cannot return the lattice-cell density required by QVF and density-gri
752	23	EWALD_3D exchange gauge is still chosen by the mesh when exchange_exxdiv is left at its default:
753	22	AO symmetry maps shells by position and accepts incompatible radial contractions
754	21	Libint Boys interpolation reads beyond its table at T=117 and corrupts periodic GDF integrals
755	20	BIPOLE erfc QQR attenuation drops diffuse angular quartets above the screening threshold
757	19	CFC energy is discontinuous: step-6 basis-point assignment moves whole segment areas between seg
758	18	libecpint radial quadrature falsely converges before an off-centre Gaussian projector peak
760	17	Dormant far-field symmetry reconstruction accepts unqualified orbits and returns incorrect Fock
762	16	Dormant far-field native adapter mixes row-major and column-major AO-pair indices
763	15	All-electron POB cells fail initial-guess ECP preflight on charge-only metadata
765	14	Periodic COHP projects bare Hcore instead of the analyzed HF/KS Hamiltonian
767	13	Periodic Mayer bond orders quadruple an unrestricted singlet and assume PS is Hermitian
768	12	periodic_runner.py:12809 picks the SCF trace row with isinstance(_trace, list), but the GPW/GAPW
769	11	CFC energy is not rotationally invariant: segment directions are lab-fixed, skipping the paper’s
770	10	CFC energy has a C1 kink where step-6’s projection flag flips
771	9	CFC step-5 basis-point-to-atom assignment is a hard argmin, leaving a residual energy jump
772	8	Periodic Becke partition changes under atom-image relabelling at fixed periodic density
773	7	Citation entry maurer_ochsenfeld_qqr_2012 is unreachable after the QQR screening removal, so tes
774	6	bugctl land hangs instead of refusing when –project-id names an unregistered project: the allow
775	5	BIPOLE SR screening still calls itself QQR in a user-facing run-log line, in troubleshooting.md
776	4	Periodic XC gradients reload translated custom bases by name and can abort or change the basis
777	3	Molecular GFN2 adenine converges to four states spanning 29.4 mHa under settings that should not
778	2	M4b/M4c: correlation= keyword on the AICCM gamma variants, with the compound citation keys that