robigo: progress

GNU R 4.5.0 runtime in Rust · last commit 8a33af5 (2026-11-08) · repository · raw data

Every merge to main is measured by CI (see docs/verification.md). Speed counts instructions with callgrind, which is exact and repeatable on a shared runner; a program whose output differs from GNU R's does not count.

Speed

Compatibility

Top CRAN packages at `9b86fc7`

99.6% of 2,264 test blocks behave as under GNU R, and 3 of 20 packages are fully green. Packages with over 900,000 downloads a month and over 500 packages using them, at PPM snapshot 2026-11-01. Their own test suites run nightly under real testthat and tinytest. A block counts when its outcome matches GNU R 4.5.0's recording, and a package is green when every block does (see docs/verification.md).

Some shards did not report: this run is incomplete and sets no score.

8 tests regressed against the ratchet.

package	matching	scored	share	time vs GNU R
R6 2.6.1	73	73	100.0%	0.7×	green
RcppEigen 0.3.4.0.2	131	131	100.0%	0.8×	green
stringr 1.6.0	212	212	100.0%	0.7×	green
tidyselect 1.2.1	218	219	99.5%	1.3×
Rcpp 1.1.2	1,621	1,629	99.5%	0.8×

What fails most

tests	packages	first failure
3	Rcpp	Mean relative difference: 0.5
2	Rcpp	Modes: character, NULL Lengths: 2, 0 target is character, current is NULL
2	Rcpp	Names: 4 string mismatches 4 string mismatches
1	tidyselect	<getvarError/missingArgError/error/condition> Error in `is_negated(x)`: argument "expr" is missing, with no default
1	Rcpp	Mean relative difference: 1

Package scripts as users run them

8 of 8 smoke scripts produce GNU R 4.5.0's exact stdout, stderr and exit status on the same pinned library. They run on every pull request in minutes, where the suites above are nightly; a script counts only when the whole transcript matches, so this is the "can I just use it" number. Cases that have never passed are pending gaps, not regressions (see docs/verification.md).

script	state
ggplot-build	pass
lubridate-dates	pass
own-dagriculture	absent
purrr-map	pass
rlang-abort	pass
stringr-glue	pass
testthat-file	pass
tibble-print	pass
tidyverse-verbs	pass

Benchmarks at `8a33af5`

benchmark	vs GNU R	robigo instr	GNU R instr
rbench/awf/Bounce	1.42×	580,535,818	825,850,699
rbench/awf/Bounce_nonames	2.61×	338,461,804	884,026,419
rbench/awf/Bounce_nonames_simple	2.58×	309,809,132	799,475,912
rbench/awf/Mandelbrot	2.48×	200,775,496	498,340,202
rbench/awf/Storage	1.40×	533,865,112	748,769,330
rbench/real/convolution	6.23×	250,686,184	1,561,606,373
rbench/real/convolution_slow	6.37×	282,586,443	1,799,669,605
rbench/real/convolution_v	2.70×	271,982,077	734,420,116
rbench/real/flexclust	0.98×	5,272,741,333	5,165,702,664
rbench/real/flexclust_kcca	1.41×	32,033,094,795	45,127,257,262
rbench/real/flexclust_no_s4	1.02×	5,692,020,630	5,779,152,745
rbench/real/volcano	2.75×	193,434,394	531,637,940
rbench/shootout/binarytrees	2.86×	240,329,965	687,745,905
rbench/shootout/binarytrees_2	2.69×	248,278,728	669,082,797
rbench/shootout/binarytrees_naive	2.69×	247,011,971	664,870,058
rbench/shootout/fannkuchredux	2.71×	269,744,529	730,721,419
rbench/shootout/fannkuchredux_naive	2.38×	274,587,838	652,493,898
rbench/shootout/fasta	2.90×	311,244,279	901,724,058
rbench/shootout/fasta_2	2.64×	367,261,561	968,199,431
rbench/shootout/fasta_3	2.58×	367,166,228	949,023,012
rbench/shootout/fasta_naive	3.89×	247,485,673	962,338,651
rbench/shootout/fasta_naive_2	3.75×	254,077,709	953,412,012
rbench/shootout/fastaredux	3.52×	272,999,028	961,284,066
rbench/shootout/fastaredux_naive	4.30×	265,896,388	1,143,264,128
rbench/shootout/knucleotide	1.71×	3,017,557,813	5,156,471,198
rbench/shootout/knucleotide_brute	0.32×	21,801,512,808	6,967,427,773
rbench/shootout/knucleotide_brute_2	0.33×	23,072,131,028	7,569,803,612
rbench/shootout/knucleotide_brute_3	0.31×	117,015,062,774	36,851,517,420
rbench/shootout/mandelbrot_ascii	1.81×	416,719,304	754,610,828
rbench/shootout/mandelbrot_naive_ascii	3.99×	279,944,605	1,117,485,013
rbench/shootout/mandelbrot_noout	1.71×	496,623,577	848,399,226
rbench/shootout/mandelbrot_noout_naive	1.99×	409,993,366	816,269,867
rbench/shootout/nbody	1.48×	691,463,011	1,022,080,922
rbench/shootout/nbody_2	2.26×	460,505,107	1,039,946,053
rbench/shootout/nbody_3	2.24×	501,596,647	1,121,762,084
rbench/shootout/nbody_naive	3.55×	381,821,880	1,356,249,730
rbench/shootout/nbody_naive_2	2.59×	497,874,442	1,291,129,818
rbench/shootout/pidigits	5.44×	279,323,894	1,519,264,130
rbench/shootout/regexdna	12.1×	1,364,996,763	16,574,483,619
rbench/shootout/reversecomplement	1.50×	2,828,604,665	4,253,357,546
rbench/shootout/reversecomplement_2	1.51×	2,649,882,186	4,000,510,640
rbench/shootout/reversecomplement_naive	1.38×	3,161,767,677	4,366,833,747
rbench/shootout/spectralnorm	6.43×	224,530,087	1,444,083,278
rbench/shootout/spectralnorm_alt	2.76×	447,234,393	1,236,253,893
rbench/shootout/spectralnorm_alt_2	3.19×	519,004,529	1,654,296,854
rbench/shootout/spectralnorm_alt_3	7.03×	232,065,679	1,631,896,151
rbench/shootout/spectralnorm_math	1.67×	510,846,428	851,557,080
rbench/shootout/spectralnorm_naive	7.47×	233,137,923	1,741,625,769
rbench/simple/bytecodes	7.08×	200,907,219	1,422,366,465
rbench/simple/emptyFor	7.01×	193,292,580	1,354,477,136
rbench/simple/emptyWhile	8.19×	193,156,800	1,582,358,996
rbench/simple/lapply	3.11×	256,062,360	796,868,179
rbench/simple/lapplyDots	3.08×	257,958,889	793,562,251
rbench/simple/listFor	3.39×	280,157,740	950,345,795
rbench/simple/listWhile	3.25×	383,876,433	1,248,265,260
rbench/simple/matrixFor	3.50×	218,709,098	765,463,071
rbench/simple/profiler-microbenchmark	5.93×	198,313,826	1,175,333,732
rbench/simple/profiler-rsa	3.95×	279,884,890	1,104,508,019
rbench/simple/profiler-shared	6.17×	1,123,558,363	6,936,145,619
rbench/simple/scalarFor	6.23×	193,494,513	1,205,818,363
rbench/simple/scalarWhile	8.54×	193,364,109	1,650,918,675
rbench/simple/superWhile	6.84×	193,187,309	1,322,071,849
rbench/simple/vectorFor	6.17×	223,199,753	1,377,408,702
rbench/simple/vectorWhile	6.49×	233,808,894	1,516,925,907
robigo/apply	3.82×	526,488,979	2,009,732,460
robigo/binarytrees	4.05×	340,570,340	1,379,173,141
robigo/closures	1.42×	885,819,885	1,255,604,183
robigo/connections	0.87×	2,857,102,012	2,472,473,361
robigo/csv	1.20×	1,279,773,184	1,538,878,803
robigo/dataframe	2.64×	588,118,702	1,553,774,970
robigo/fannkuch	3.31×	608,419,744	2,016,403,679
robigo/fib	3.04×	355,160,579	1,080,149,261
robigo/lm	1.44×	651,856,788	936,276,563
robigo/mandelbrot	4.34×	226,968,792	984,131,761
robigo/matrix	2.17×	650,616,661	1,410,934,800
robigo/na-entry-alternation	9.51×	249,422,855	2,371,289,848
robigo/nbody	8.71×	302,452,153	2,633,784,429
robigo/queens	2.16×	235,632,745	508,432,881
robigo/s4-generic	1.79×	531,047,565	950,209,243
robigo/s4-new	1.70×	562,164,327	953,479,547
robigo/s4-portfolio	1.87×	516,300,462	963,126,844
robigo/s4-slot-read	2.00×	401,561,870	803,886,488
robigo/s4-slot-write	1.93×	436,494,301	844,319,652
robigo/s4-valid	1.69×	567,455,663	960,495,162
robigo/sortorder	1.19×	835,286,973	992,747,366
robigo/spectralnorm	10.8×	245,186,035	2,641,629,229
robigo/strings	2.29×	693,259,802	1,589,665,821
robigo/vecarith	1.25×	4,313,859,245	5,409,871,372
robigo/vecops	1.69×	2,379,515,026	4,016,388,127

Corpora at `8a33af5`

corpus	passing	cases	share
FastR test snippets (fastr)	31,336	31,374	99.9%
GNU R's regression tests (gnur-tests)	36	59	61.0%
Help-page examples (gnur-examples)	1,113	1,153	96.5%
GNU R's demos (gnur-demos)	15	16	93.8%
An Introduction to R (r-intro)	12	13	92.3%
Pure-R CRAN packages (cran-purer)	39	41	95.1%
C API (capi)	66	68	97.1%
Conformance scripts (conformance)	454	466	97.4%

Open gaps at `8a33af5`

Cases robigo still fails, per corpus with at most a few hundred. A known gap is a conformance script or corpus case with GNU R's recorded output and no ratchet marker yet: CI tracks it without failing on it.

FastR test snippets (fastr): 38 failing

S4/TestS4/58ca4e722adb4a25
builtins/TestBuiltin_D/af1f7c2f6a77afbf
builtins/TestBuiltin_adist/a293491ce8a97832
builtins/TestBuiltin_aregexec/2eb4a38a1eeb9221
builtins/TestBuiltin_aregexec/788b8caf4efe45bd
builtins/TestBuiltin_bincode/f5bcacdc34439f42
builtins/TestBuiltin_cbind/61d7e05c8f316038
builtins/TestBuiltin_classassign/b2b4215a7276f9f7
builtins/TestBuiltin_copyDFattr/99de167ec19f7eb8
builtins/TestBuiltin_dotdotdotlength/4a46c09fe393f84d
builtins/TestBuiltin_dqr/f1ed717e57091d3a
builtins/TestBuiltin_extract_replace/2d3f74f6c25fd37a
builtins/TestBuiltin_function/cc1c05aea449da65
builtins/TestBuiltin_iconv/9a3d0cfc73aec861
builtins/TestBuiltin_iconv/ea6e70a8017468fc
builtins/TestBuiltin_icuSetCollate/ea39cd4fcba58898
builtins/TestBuiltin_intToUtf8/edca96cef69b2cc4
builtins/TestBuiltin_isrecursive/27b06b37744e8f87
builtins/TestBuiltin_isunsorted/18b5d8b101543441
builtins/TestBuiltin_loadInfoFromConn2/45d0e1f485981858
builtins/TestBuiltin_memDecompress/4eb6e51417bc29f4
builtins/TestBuiltin_pmatch/c33bceb53c742b6a
builtins/TestBuiltin_primUntrace/de220d3354cb3fda
builtins/TestBuiltin_quit/580fcff62a60f310
builtins/TestBuiltin_rbind/e2d15825d89e98c7
builtins/TestBuiltin_rcond/922ef201ac8bf67d
builtins/TestBuiltin_row/f1b327f07159646e
builtins/TestBuiltin_serializeInfoFromConn/e9f4596b36ecaa1f
builtins/TestBuiltin_unzip/75846dbd22993069
builtins/TestBuiltin_zzfile/32ed8707688a4d62
builtins/TestBuiltin_zzfile/e0ae267b8228a365
builtins/TestStackBuiltins/10b015faafd62069
builtins/TestStackBuiltins/8924b12e8f314e96
functions/TestFunctions/b2cc3802485c2767
library/base/TestConnections/9a4e82d4b5d3fa18
library/base/TestSimpleArithmetic/f7fa6afb895ab94a
library/base/TestSimpleAttributes/e7475d933064e4bf
library/utils/TestInteractiveDebug/203ed9d6af40af44

GNU R's regression tests (gnur-tests): 23 failing

Help-page examples (gnur-examples): 40 failing

GNU R's demos (gnur-demos): 1 failing

base/error.catching

An Introduction to R (r-intro): 1 failing

12-packages

Pure-R CRAN packages (cran-purer): 2 failing

C API (capi): 2 failing

Conformance scripts (conformance): 12 failing

Every commit

271 commits

commit	date	speed	fastr	gnur-tests	gnur-examples	gnur-demos	r-intro	cran-purer	packages	capi	conformance	builtins
`8a33af5` #491: CI: run only the benchmarks a PR's changed paths can a	2026-11-08	2.74×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.4%	702/760
`0d1f098` #495: gc: core collections walk the runtime graph only while	2026-11-08	2.74×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.4%	702/760
`a900546` #493: perf(jit): cut scalar-complex Mandelbrot instructions	2026-11-08	2.74×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.4%	702/760
`f57b71c` #492: council: spare-cores designs reviewed (#373-#381)	2026-11-08	2.71×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.4%	702/760
`3669e86` #494: Spare-cores measurements: loop-compile stalls 8-18% on	2026-11-08	2.71×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.4%	702/760
`49bda63` #496: Avoid repeated NA twin copies and close failed-spawn l	2026-11-08	2.71×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.4%	702/760
`9b86fc7` #487: Dogfooding: tidyverse loads without natives, smoke tie	2026-11-08	2.71×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.1%	97.2%	702/760
`2a5ff0b` #485: perf(jit): remove redundant integer NA checks from nat	2026-11-07	2.71×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		97.0%	97.4%	702/760
`2fb18b4` #484: package merge train 2: #445, #478, #453, #469, #398	2026-11-07	2.84×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%	93.4%	97.0%	97.1%	702/760
`49f2829` #483: ci: remove the self-hosted runners, whose privileged d	2026-11-07	2.83×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		98.4%	96.9%	701/760
`5495ef2` #300: io: gzcon(), the gzip connection wrapper	2026-11-06	2.83×	99.9%	61.0%	96.5%	93.8%	92.3%	95.1%		98.4%	97.1%	701/760
`a0982fc` #482: merge train 5: #312, #342, #317, #323, #333, #338, #34	2026-11-06	2.83×	99.9%	61.0%	96.4%	93.8%	92.3%	95.1%		98.4%	97.1%	700/760
`1f9a03d` #297: io: url() connections through libcurl, and file:// thr	2026-11-06	2.51×	99.9%	61.0%	96.4%	93.8%	92.3%	95.1%		98.4%	97.4%	700/760
`f964eca` #272: conformance(fastr): tracemem reports copies of traced	2026-11-06	2.51×	99.9%	61.0%	96.4%	93.8%	92.3%	95.1%		98.4%	97.4%	700/760
`4835c4b` #295: conditions: unhandled errors invoke the abort restart	2026-11-06	2.51×	99.9%	61.0%	96.4%	93.8%	92.3%	95.1%		98.4%	97.4%	700/760
`30afc6b` #290: linpack: port dtrco, kappa's exact triangular conditio	2026-11-06	2.51×	99.9%	61.0%	96.4%	93.8%	92.3%	95.1%		98.4%	97.3%	700/760
`1a8edaa` #481: package merge train 1: #465, #458, #470, #477	2026-11-06	2.51×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%	85.9%	98.4%	97.1%	700/760
`0c06832` #435: capi: R_new_custom_connection, so curl() connections w	2026-11-06	2.51×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.3%	700/760
`003bff0` #448: suspendInterrupts(): an interrupt raised from C waits	2026-11-06	2.51×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.3%	700/760
`2ec8a12` #473: merge train 3+4: #316, #318, #324, #326, #329, #336, #	2026-11-06	2.51×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.1%	700/760
`2789dbc` #480: ci: fork PRs stay off self-hosted runners even when th	2026-11-06	2.29×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`23db57d` #479: ci: run build and corpus jobs on our own Docker runner	2026-11-06	2.29×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`f03f611` #298: interp: in-place hand-off for complex assignment over	2026-11-05	2.29×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`6802969` #442: gc: one mark over both heaps, so cycles through C and	2026-11-05	2.29×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.7%	700/760
`b900075` #463: bridge: convert symbols by identity, not by hashing th	2026-11-05	2.29×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`12bd424` #456: ledger(packages): round 3 preview, 12 of 20 green	2026-11-05	2.29×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`1de9e8f` #475: build-natives: retry the R sources download from CRAN	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`4842ba0` #454: dotcode: .Fortran passes the first string of a charact	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`e0229b2` #436: capi: C keeps one live copy of each value it looks up	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`2079b4c` #461: capi corpus: ggplot2's test-scales segfault is #167's	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`2f27634` #474: corpora: one GNU R recording file per capi case, so PR	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`9d2ba3d` #471: rjit: installed code is what was unserialized as bytec	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.7%	700/760
`c587fce` #467: envir: active bindings are called as GNU R calls them	2026-11-05	2.30×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`fe27484` #451: Promise constant arguments where GNU R's AST interpret	2026-11-05	2.30×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`23c6015` #462: bridge: hash address-keyed tables with multiply-and-fo	2026-11-05	2.30×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	700/760
`853a2a4` #425: compiler: cmpfun() and compile() work in the default b	2026-11-05	2.30×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.7%	700/760
`d493028` #444: bridge: an object's attributes come back only when C w	2026-11-05	2.30×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.7%	694/760
`71ee616` #427: conditions: a missing argument read by compiled code i	2026-11-05	2.30×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.7%	694/760
`468c9d0` #452: interp: sys.call()/sys.calls() report statement srcref	2026-11-05	2.31×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.7%	694/760
`8e66785` #447: conditions: C errors and warnings reach R handlers thr	2026-11-05	2.31×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`e8110a5` #440: capi: plotmath's Greek letters render as Greek in SVG:	2026-11-05	2.31×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`64e0f22` #446: capi: a native routine's result outside the heap is an	2026-11-05	2.31×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`f16e4dc` #449: attributes: a call's argument tags are not its attribu	2026-11-05	2.31×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`1ff0177` #424: io: one method per GNU R Rconn pointer, so a new conne	2026-11-05	2.31×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`7a0d27b` #439: dotcode: .C/.Fortran refuse NA and non-finite argument	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`3d99bae` #441: regular expressions: GNU R's own TRE for the construct	2026-11-05	2.33×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`05ffd4c` #429: capi fixture: keep the cloned value protected across d	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`57e9bf8` #434: ledger(packages): round 2, 11 of 20 green	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`f74ea99` #428: build-natives: reconfigure once per newly installed pa	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`4c19fc8` #430: capi: a `...` C binds keeps its identity through gc(),	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`88bef9b` #426: repl: no prompt under --no-echo and the prompt option	2026-11-05	2.28×	99.9%	59.3%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`9d9dff7` #421: capi: export Rstrdup and Rf_utf8ToLatin1AdobeSymbol2ut	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`d7ecc43` #417	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`802f242` #420: build-natives: configure again once pango is installed	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`e4ab973` #418: ledger: package campaign round 1 — 9 of 20 green, 84.5	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.9%	694/760
`4b43af4` #408: vm: the dispatch loop keeps only hot paths inline, so	2026-11-05	2.28×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`dfc5412` #407: rjit: local() in a loop works with the VM off, so the	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`8dc0ff6` #412: bind: lengths() of a plain list reads its elements in	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`078fdc4` #404: capi corpus: envattr_eval pins that R code C evaluates	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`47ccdf6` #360: io: stderr() follows sink(type = "message"), so captur	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`249212e` #394: is.nan: character and raw vectors are FALSE, not an er	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.3%	694/760
`6abc95f` #411: interp: closure returns skip run_on_exit when there is	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`67cc57e` #392	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`e0f608c` #402: core: collect the C heap when GNU R would (R_NSize/R_V	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`6058221` #395: grDevices: build and load GNU R's cairo module, so png	2026-11-05	2.27×	99.9%	61.0%	96.3%	93.8%	92.3%	95.1%		100.0%	97.6%	694/760
`6222a23` #253: strings: port adist's full path so utils::adist works	2026-11-05	2.25×	99.9%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.6%	694/760
`c0d96b7` #345	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.6%	693/760
`ce9c115` #369: packages: httr goes green once skips its own code alwa	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.6%	693/760
`8f64679` #358	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.6%	693/760
`6e45e52` #361: capi: R_Busy is a no-op as in GNU R, so bit64's sorts	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.3%	693/760
`3a8cc0b` #356: vm: the simple-.Internal memos hold callers weakly, so	2026-11-04	2.25×	99.8%	59.3%	96.3%	93.8%	92.3%	92.7%		100.0%	97.6%	693/760
`d5a4c83` #346: packages: curl's tests run against a pinned local http	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.3%	693/760
`93d45d4` #363: dispatch: S3 methods of primitives run in GNU R's disp	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.3%	693/760
`6206e55` #359: eval: run on.exit code registered in eval's own contex	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	92.7%		100.0%	97.3%	693/760
`eac45a9` #348: dispatch: internal generics find their S4 methods, tur	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	90.2%		100.0%	97.3%	693/760
`dbe74cd` #339: dotcode: .External2 keeps its routine's visibility, so	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`1e28554` #334: jit: nested arithmetic arguments run natively, binaryt	2026-11-04	2.25×	99.8%	61.0%	96.3%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`b014f33` #314: %*%: vector times matrix 3x faster, spectralnorm_alt a	2026-11-04	2.24×	99.8%	61.0%	96.3%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`bf7332e` #301: packages: score the top 20 CRAN packages' test suites	2026-11-04	2.22×	99.8%	61.0%	96.3%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`5ce8d76` #309: memory: gctorture() really collects at every allocatio	2026-11-04	2.22×	99.8%	61.0%	96.2%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`e7de574` #308: dispatch: Ops asks chooseOpsMethod() on conflicting me	2026-11-04	2.22×	99.8%	61.0%	96.3%	93.8%	92.3%	90.2%	64.3%	100.0%	97.3%	693/760
`c2d0108` #307: capi: port R_check_class_etc and R_data_class so Matri	2026-11-04	2.22×	99.8%	61.0%	96.2%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`884aa54` #287: io: writeBin/readBin size 16 is C's long double	2026-11-03	2.22×	99.8%	61.0%	96.2%	93.8%	92.3%	90.2%		100.0%	97.5%	693/760
`e1d893a` #289: conformance(fastr): is.* predicates and time limits	2026-11-03	2.22×	99.8%	61.0%	96.1%	93.8%	92.3%	90.2%		100.0%	97.2%	693/760
`732a179` #291: internet: libcurlVersion over the dlopen'd libcurl	2026-11-03	2.22×	99.8%	61.0%	96.1%	93.8%	92.3%	90.2%		100.0%	97.5%	692/760
`a7cca4f` #273: envir: env.profile over the hash-table model	2026-11-03	2.22×	99.8%	61.0%	96.0%	93.8%	92.3%	90.2%		100.0%	97.5%	691/760
`3f7429b` #265: capi: C-assembled promises bind as runtime promises	2026-11-03	2.15×	99.8%	61.0%	95.9%	93.8%	92.3%	90.2%		100.0%	97.8%	691/760
`4fa3579` #271: bind: deparse.level=2 tags use deparse1line's 10-byte	2026-11-03	2.15×	99.8%	61.0%	95.9%	93.8%	92.3%	90.2%		100.0%	97.8%	691/760
`7c7f8b2` #275: conformance(fastr): envir/diag argument validation and	2026-11-03	2.15×	99.8%	61.0%	95.8%	93.8%	92.3%	90.2%		100.0%	97.7%	691/760
`0ec42f9` #246: conformance: honour options(OutDec=) in as.character()	2026-11-03	2.15×	99.8%	61.0%	95.8%	93.8%	92.3%	90.2%		100.0%	97.7%	691/760
`8c1b28a` #270: scripts/rivals.py: measure robigo against every other	2026-11-03	2.15×	99.8%	61.0%	95.8%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`3920a71` #274: jit: run hot closures as native code, fib(27) 53 ms ->	2026-11-03	2.17×	99.8%	61.0%	95.8%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`ac9d0c3` #267: serialize: write ASCII streams to connections (seriali	2026-11-03	2.09×	99.8%	61.0%	95.8%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`d15362e` #269: aperm: a non-vector perm raises R's LENGTH error inste	2026-11-03	2.09×	99.8%	61.0%	95.7%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`10182a7` #264: conformance(fastr): runtime installs reject the empty	2026-11-03	2.09×	99.8%	61.0%	95.7%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`52a4edc` #266: jit: closures created by installed code are bytecode;	2026-11-03	2.09×	99.8%	61.0%	95.7%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`63d68b2` #263: conformance(fastr): terms() takes a NULL response; ape	2026-11-02	2.09×	99.8%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`de202cd` #262: CI: large PRs get a full OpenCodeReview instead of a 3	2026-11-02	2.09×	99.8%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`9714ecc` #261: conformance(fastr): matrix column labels use the escap	2026-11-02	2.09×	99.8%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`22d63fb` #258: perf(vm): nested pure-call inlining (fib -51%)	2026-11-02	2.09×	99.8%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`7ec6b3b` #260: conformance(fastr): keep pairlist attributes when prin	2026-11-02	2.08×	99.8%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`d05db01` #249: connections: PSOCK clusters work: sockets, and unseria	2026-11-02	2.08×	99.8%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	691/760
`3cb1fac` #257: io: memCompress/memDecompress (connections.c's in-memo	2026-11-02	2.08×	99.7%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	686/760
`ab05731` #256: io: bzfile connections (libbz2), and gzfile/file magic	2026-11-02	2.08×	99.7%	61.0%	95.5%	93.8%	92.3%	90.2%		100.0%	98.0%	684/760
`7ba97f8` #228: build(deps): bump fancy-regex from 0.16.2 to 0.20.0	2026-11-02	2.08×	99.7%	61.0%	95.4%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`015a18d` #226: build(deps): bump pprof from 0.14.1 to 0.15.0	2026-11-02	2.08×	99.7%	61.0%	95.4%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`e8afe19` #242: namespace: tibbles print again: importIntoEnv keeps ac	2026-11-02	2.08×	99.7%	61.0%	95.4%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`a38821d` #252: conformance: trace()/.primTrace, and S4 rep keeps its	2026-11-02	2.08×	99.7%	61.0%	95.4%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`750fa6d` #229: build(deps): bump the actions group across 1 directory	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`7ddaeb7` #255: ratchet: mark calls-promise-wrapper{,-force} passing	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`963fc9c` #222: vm: promise thunks call simple-.Internal wrappers thro	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`e9046bd` #250: unsafe: fix the memory bugs an audit found, and deny u	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`505d3a5` #238: capi: rlang errors print again: keep env attributes an	2026-11-02	2.08×	99.7%	59.3%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`63b7d64` #247: iconv: cli's ASCII output works: the C library's conve	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`d570f67` #231: capi: R_NewEnv with a NULL enclosure and FRAME/ENCLOS	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`1055fa6` #251: conformance(fastr): port qr_coef_cmplx from Lapack.c	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	97.9%	684/760
`b8de1d3` #244: conformance(gnur-tests): dispatch as.double, as.call a	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	98.2%	683/760
`5eb76b2` #248: conformance(fastr): open gzipped files through the gz	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	98.2%	683/760
`8b3894e` #243: conformance(fastr): port mget() argument validation an	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	98.5%	683/760
`466d375` #245: conformance: Recall promises, NA attribute names, as.d	2026-11-02	2.08×	99.7%	61.0%	95.1%	93.8%	92.3%	90.2%		100.0%	98.5%	683/760
`b73b21a` #241: natives: a fresh checkout builds grid, parallel and fr	2026-11-02	2.08×	99.7%	59.3%	94.6%	93.8%	92.3%	90.2%		100.0%	98.5%	683/760
`1b8fbb8` #240: conformance(gnur-tests): ls(sorted=) orders names with	2026-11-02	2.08×	99.7%	59.3%	94.6%	93.8%	92.3%	90.2%		100.0%	98.5%	683/760
`21de310` #239: conformance(fastr): align active bindings, slot names	2026-11-02	2.08×	99.7%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	98.5%	683/760
`2c5ac38` #235: sprintf: tidy the %d fast path after #212's review	2026-11-02	2.08×	99.6%	55.9%	94.5%	93.8%	84.6%	90.2%		100.0%	98.5%	683/760
`3399fe4` #215: tests: track the CRAN-gate gaps found on the popular p	2026-11-02	2.08×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	98.5%	683/760
`5922aeb` #159: track(conformance-fastr): align parent.env builtins wi	2026-11-02	2.08×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	99.4%	683/760
`16427cc` #212: perf(sort,sprintf): sorted-input fast paths in the rad	2026-11-02	2.08×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	99.4%	683/760
`073a8a2` #230: nightly: make the binaries build everywhere, and 14% f	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	99.4%	683/760
`14c187d` #227: lint: run clippy in CI, and fix what it finds	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	99.4%	683/760
`18545b2` #225: fix: tryCatch catches huge allocations, and bad input	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	99.4%	683/760
`686a187` #221: build-natives: self-provision libcurl when configure w	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`a437bef` #223: ci: PRs stop waiting on release-mode test builds, and	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`d75a565` #216: fix: stop .Call corrupting running R code, and remove	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`c4e71be` #211: errors: run options(error=) with the stack live, and s	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`4956dbc` #209: cran-gate: stop a hung package from blocking CI for ho	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`01ca454` #210: docs: define when Arrow buffers can be zero-copy, and	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`954a807` #208: interp: honor options(expressions=), and make stack ov	2026-11-02	2.10×	99.6%		94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`fb87481` #207: progress page: record the conformance scripts and list	2026-11-02	2.09×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%	100.0%	683/760
`bd4cddb` #205: interp: return with no target context raises the catch	2026-11-02	2.09×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`292ebb5` #206: perf(vm): compile constant and symbol closure bodies;	2026-11-02	2.09×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`46fc129` #204: parsing: parse(text=) coerces its argument like coerce	2026-11-02	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`675cc1c` #196: envir: topenv honors matchThisEnv (envir.c:topenv)	2026-11-01	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`7a89cd0` #195: serialize: force lazy-load database promises in their	2026-11-01	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`21d63d4` #187: docs: bayesfam dogfooding ledger	2026-11-01	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`776872a` #175: cran: gate harness for CRAN packages with compiled cod	2026-11-01	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`ca47129` #176: conformance: deparse non-finite complex as complex(rea	2026-11-01	2.10×	99.6%	57.6%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`0506bd4` #179: lapack: port La_qr, qr_coef_real and qr_qy_real (Lapac	2026-11-01	2.10×	99.6%	55.9%	94.5%	93.8%	84.6%	90.2%		100.0%		683/760
`fb002dc` #183: perf(apply): direct elements and inline FUN in mapply	2026-11-01	2.10×	99.6%	55.9%	94.4%	93.8%	84.6%	90.2%		100.0%		680/760
`2943450` #188: conditions: honor show.error.messages when printing un	2026-11-01	2.10×	99.6%	55.9%	94.4%	93.8%	84.6%	90.2%		100.0%		680/760
`4582066` #184: native: tell which native build is missing when a rout	2026-11-01	2.10×	99.6%	55.9%	94.4%	93.8%	84.6%	90.2%		100.0%		680/760
`9a54ca5` #185: capi: bridge C-assembled closures (from_sexp of raw CL	2026-11-01	2.10×	99.6%	55.9%	94.4%	93.8%	84.6%	90.2%		100.0%		680/760
`c2db7c8` sysutils: port .Internal(Sys.umask) and warn on attributes o	2026-11-01	2.10×	99.6%	55.9%	94.3%	93.8%	84.6%	90.2%		100.0%		680/760
`f2f04bc` conformance(fastr): report parse errors with GNU R's three R	2026-11-01	2.10×	99.6%	55.9%	94.1%	93.8%	84.6%	90.2%		100.0%		679/760
`6b917a7` #177: capi: fix numericDeriv's zero gradient by keeping C's	2026-11-01	2.10×	99.5%	55.9%	94.1%	93.8%	84.6%	90.2%		100.0%		679/760
`e6fe408` conformance: print deferred warnings with R's 'In addition:'	2026-11-01	2.10×	99.5%	55.9%	92.5%	93.8%	76.9%	90.2%		100.0%		679/760
`532eef0` #172: task: port top-level task callbacks (main.c)	2026-11-01	2.10×	99.5%	54.2%	92.5%	93.8%	76.9%	90.2%		100.0%		679/760
`6c89483` capi: implement substitute, qsort, streams and envir predica	2026-11-01	2.10×	99.5%	54.2%	92.2%	93.8%	76.9%	90.2%		100.0%		679/760
`10654ab` #169: grDevices colors()/col2rgb(); capi pairlist coerceVect	2026-11-01	2.10×	99.5%	52.5%	92.2%	93.8%	76.9%	90.2%		100.0%		679/760
`7516f17` #171: capi corpus: a promise case for R_PromiseExpr	2026-11-01	2.10×	99.5%	54.2%	92.2%	93.8%	76.9%	90.2%		100.0%		679/760
`cd51fcf` sysutils: port .Internal(file.show()) and .Internal(file.sym	2026-11-01	2.10×	99.5%	54.2%	92.2%	93.8%	76.9%	90.2%		100.0%		679/760
`7b61f07` conformance: is.na() on expressions warns and returns xlengt	2026-11-01	2.10×	99.5%	54.2%	91.7%	93.8%	69.2%	90.2%		100.0%		677/760
`5fe0672` perf(apply): unbox inline closure calls and fast-path scalar	2026-11-01	2.10×	99.5%	54.2%	91.7%	87.5%	69.2%	90.2%		100.0%		677/760
`6849338` #156: capi: port R_IsPackageEnv and R_EnvironmentIsLocked	2026-11-01	2.10×	99.5%	54.2%	91.7%	87.5%	69.2%	90.2%		100.0%		677/760
`9a4d858` conformance: implement pcre_config, out-of-context nargs, an	2026-11-01	2.10×	99.5%	52.5%	91.7%	87.5%	69.2%	90.2%		100.0%		677/760
`29f8d4a` conformance(fastr): match GNU R inherits() argument validati	2026-11-01	2.09×	99.5%	54.2%	91.5%	87.5%	69.2%	87.8%		100.0%		676/760
`674f277` capi: implement R_mkClosure/R_PromiseExpr, preserve CHARSXP,	2026-11-01	2.09×	99.5%	54.2%	91.5%	87.5%	69.2%	87.8%		100.0%		676/760
`0b48677` conformance(fastr): align do_vector argument checking and mo	2026-11-01	2.09×	99.5%	54.2%	91.5%	87.5%	69.2%	87.8%		90.9%		676/760
`f523d79` #154: base: port small missing internals and their error pat	2026-10-31	2.09×	99.5%	54.2%	91.5%	87.5%	69.2%	87.8%		90.9%		676/760
`7eb4e44` #153: cli: match R's shutdown order and stop TLS destructor	2026-10-31	2.09×	99.4%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`2300ec5` #127: perf: benchmark hill-climb round 1 (hunt biggest absol	2026-10-31	2.09×	99.4%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`55befe4` #118: Nightly releases, crates.io groundwork, install/usage	2026-10-31	2.03×	99.4%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`1511a11` #152: lang: returnValue() reports the pending return value i	2026-10-31	2.03×	99.4%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`59deee1` #150: base: bitwise type rules, all.names counting, length<-	2026-10-31	2.03×	99.4%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`b35ba27` #151: perf: measure like-for-like at source and re-pair only	2026-10-31	2.03×	99.3%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`ebbbd21` #142: perf: cache GNU R oracle output in perf-main to measur	2026-10-31	2.24×	99.3%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`13fda5c` #141: complex: implement polyroot and make Re/Im/Mod/Arg/Con	2026-10-31	2.24×	99.3%	54.2%	91.4%	87.5%	69.2%	87.8%		90.9%		675/760
`10e1a12` #131: ci: pin HOME to the euid home so rustup can install	2026-10-31	2.24×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.9%		675/760
`8c77948` #130: vm: unboxed scalar promises and generalized compiled-w	2026-10-31	2.24×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.9%		675/760
`8cd818a` #129: io, serialize, native: buffer ConnKind::File writes an	2026-10-30	2.23×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.9%		675/760
`0abc3e1` #128: methods: implement guarded VM slot-read instruction	2026-10-30	2.22×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.9%		675/760
`ff75903` #126: Port SETCAR/SETCDR/SET_TAG shared cons-cell mutation	2026-10-30	2.22×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.9%		675/760
`c3bb38a` #124: Connection writes skip the vfprintf round-trip for lit	2026-10-30	2.22×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		675/760
`2c46ca3` #125: bytecode: wire putconst in-place mutation, cmpfun layo	2026-10-30	2.22×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		675/760
`fc9f7e1` #123: Match GNU R in 12 orphaned data cases	2026-10-30	2.22×	99.3%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		675/760
`c937758` #122: Port SET_VECTOR_ELT shared-list mutation	2026-10-30	2.22×	99.2%	50.8%	90.6%	87.5%	69.2%	87.8%		90.5%		675/760
`d075864` #121: bytecode: putconst follow-up, cmpfun layout fixtures,	2026-10-30	2.22×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		675/760
`17dafc3` #120: Read and write compiled R files with exact ASCII and X	2026-10-30	2.22×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		675/760
`a32c562` #119: library(stats): skip the discarded bytecode instructio	2026-10-30	2.22×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`758efb7` #108: Direct connection writes and general simple-internal i	2026-10-30	2.21×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`f215c4c` #105: Construct GNU R bytecode and compiled closures through	2026-10-30	2.19×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`2ac8f81` #100: Preserve GNU bytecode behind an opt-in fidelity featur	2026-10-30	2.19×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`a142ca6` #117: Align warmed @ slot reads with GNU R and ratchet 21/21	2026-10-30	2.19×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`ef49317` #116: Reclaim environment cycles and fix enc2 string marks	2026-10-30	2.19×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`0bee889` #115: Constrain S4 metadata invalidation before caching	2026-10-30	2.20×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`83e58b9` #114: Cut S4 dispatch instructions by 2.1%	2026-10-30	2.20×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`5bff0ca` #112: Parse ratchet: one marker file per source	2026-10-30	2.20×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`b5c3373` #113: Activate the FastR methods default profile	2026-10-30	2.20×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`ff78657` #111: Conflict-free ratchets: one marker file per case	2026-10-30	2.20×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`b7ee268` #98: Run environment finalizers across cycles and process ex	2026-10-30	2.20×	99.2%	52.5%	90.6%	87.5%	69.2%	87.8%		90.5%		673/760
`1a591ff` #106: Make eigen, SVD and pivoted Cholesky work through syst	2026-10-30	2.20×	99.2%	52.5%	90.5%	87.5%	69.2%	87.8%		90.5%		673/760
`17768b5` #110: Mark octal/hex escaped string literals native like GNU	2026-10-30	2.21×	99.1%	50.8%	88.7%	87.5%	69.2%	87.8%		90.5%		665/760
`ec88bd1` #107: Prepare FastR methods defaults with an isolated oracle	2026-10-30	2.21×	99.1%	49.2%	88.7%	87.5%	69.2%	87.8%		90.5%		665/760
`532be45` #104: Preserve substitute dots cells without growing ordinar	2026-10-30	2.21×	99.1%	50.8%	88.7%	87.5%	69.2%	87.8%		90.5%		665/760
`eb7a7ae` #103: Match base bytecode is.vector argument contexts	2026-10-30	2.21×	99.1%	50.8%	88.7%	87.5%	69.2%	87.8%		90.5%		665/760
`805e5fb` #71: Attach methods in GNU R corpus suites	2026-10-30	2.21×	99.1%	50.8%	88.7%	87.5%	69.2%	87.8%		90.5%		665/760
`d1c95c5` #99: Build complete startup caches and confirm perf regressi	2026-10-30	2.21×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`d81d29c` #102: Reduce native CSV formatting instructions by 8.5%	2026-10-30	2.21×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`fd618f5` #92: Reduce startup instructions and flat-cache memory use	2026-10-30	2.21×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`fb99c0b` #101: Preserve native writes through the SEXP bridge	2026-10-30	2.19×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`1b4670a` #94: Match GNU R's compiled setter error calls	2026-10-30	2.19×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`f5dae40` #97: Cut S4 generic dispatch instructions by 3.5%	2026-10-30	2.19×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`bb23b54` #96: Preserve parser reductions for getParseData	2026-10-30	2.21×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`d2ce510` #95: Preserve original bytes and encoding marks through icon	2026-10-30	2.21×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`df3d535` #84: Preserve byte strings across the runtime and native bri	2026-10-30	2.21×	99.1%	52.5%	88.6%	87.5%	69.2%	87.8%		90.5%		665/760
`1074f6c` Progress charts: speed and compatibility over time in the RE	2026-10-30	2.20×	99.0%	50.8%	88.3%	81.2%	69.2%	87.8%		90.5%		665/760
`4754027` #93: Preserve closure source references and original source	2026-10-30	2.20×	99.0%	52.5%	88.3%	81.2%	69.2%	87.8%		90.5%		665/760
`f668f60` #90: Cut VM inline-call loop instructions 5%	2026-10-29	2.20×	99.0%	50.8%	88.3%	81.2%	69.2%	87.8%		90.5%		665/760
`29cac51` #91: parse: retain source references through expression subs	2026-10-29	2.20×	99.0%	50.8%	88.3%	81.2%	69.2%	87.8%		90.5%		665/760
`d63d043` #60: Cut naïve mandelbrot instructions 29% with exact VM ari	2026-10-29	2.20×	99.0%	47.5%	87.7%	68.8%	69.2%	87.8%		90.5%		665/760
`dbf6754` #89: Match GNU R closure printing and parse diagnostics	2026-10-29	2.19×	99.0%	47.5%	87.7%	68.8%	69.2%	87.8%		90.5%		665/760
`e25e6cc` #86: Match GNU R NaN payloads in VM scalar arithmetic	2026-10-29	2.19×	99.0%	47.5%	86.6%	68.8%	61.5%	82.9%		90.5%		665/760
`24a5b8e` #85: Match GNU R tracing state for repeated calls and symbol	2026-10-29	2.19×	99.0%	47.5%	86.6%	68.8%	61.5%	82.9%		90.5%		665/760
`3d928e1` #88: startup-perf round 5: re-baseline + serialize reader fa	2026-10-29	2.19×	99.0%	47.5%	86.6%	68.8%	61.5%	82.9%		90.5%		665/760
`bda4730` #87: system(intern=TRUE): inherit child stderr (the reg-pack	2026-10-29	1.95×	99.0%	47.5%	86.6%	68.8%	61.5%	82.9%		90.5%		665/760
`74e8c73` #83: conformance round 3: identical row.names, malformed fac	2026-10-29	1.95×	98.9%	47.5%	86.6%	68.8%	61.5%	82.9%		90.5%		665/760
`03600ed` #82: Startup: serve lazy-load databases from a flat pre-infl	2026-10-29	1.95×	98.9%	44.1%	86.5%	68.8%	61.5%	82.9%		90.5%		665/760
`677edf3` #80: JIT: cut matrixFor instructions by 58% with native matr	2026-10-29	1.95×	98.9%	44.1%	86.5%	68.8%	61.5%	82.9%		90.5%		665/760
`4e56951` #77: core: run GNU R utils natives alongside the graphics en	2026-10-29	1.92×	98.9%	44.1%	86.5%	68.8%	61.5%	82.9%		90.5%		665/760
`098bd95` #70: ci: run corpora and perf with GNU R's stats natives bui	2026-10-29	1.95×	98.8%	39.0%	57.8%	12.5%	30.8%	78.0%		90.5%		665/760
`5aea011` #69: core: GNU R's graphics engine in libR — base graphics a	2026-10-29	1.96×	97.9%	35.6%	54.0%	12.5%	30.8%	78.0%		90.5%		665/760
`7952a6d` #78: stats natives load for +0.55M instructions instead of +	2026-10-29	1.96×	97.9%	35.6%	54.0%	12.5%	30.8%	78.0%		90.5%		664/760
`1f1134e` #81: perf: reduce release instruction counts with thin LTO	2026-10-29	1.88×	97.9%	35.6%	54.0%	12.5%	30.8%	78.0%		90.5%		664/760
`7ff8a14` #76: ledger: hand three graphics-example failures to sweep-d	2026-10-29	1.80×	97.9%	35.6%	54.0%	12.5%	30.8%	78.0%		90.5%		664/760
`7663033` #68: cbind/rbind: split language objects into components as	2026-10-29	1.80×	97.9%	35.6%	54.0%	12.5%	30.8%	78.0%		90.5%		664/760
`1d63aef` #75: Objects C code keeps in statics survive across calls (R	2026-10-29	1.80×	97.9%	35.6%	53.9%	12.5%	30.8%	78.0%		90.5%		664/760
`45b4dd2` #47: complex: gcc-exact NaN payload bits for +, -, * and the	2026-10-28	1.73×	97.8%	35.6%	52.5%	12.5%	30.8%			75.0%		664/760
`750a904` #52: methods: library(methods), setClass/new/setGeneric/setM	2026-10-28	1.73×	96.8%	35.6%	52.5%	12.5%	30.8%			75.0%		664/760
`13e6038` #43: Fix NA and NaN propagation in complex matrix products	2026-10-28	1.74×	96.6%	33.9%	48.5%	12.5%	30.8%			75.0%		663/760
`32c6002` #53: core: base-package natives from GNU R's sources by defa	2026-10-28	1.74×	96.6%	33.9%	48.5%	12.5%	30.8%			75.0%		663/760
`d8a097f` #37: core: G2 continued — medium objects, isolates, fork-saf	2026-10-28	1.74×	96.6%	33.9%	48.5%	12.5%	30.8%			75.0%		663/760
`9960a44` #49: perf-vm: for (j in seq(length = n)) as a native range (	2026-10-28	1.80×	96.5%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`b6fc7c8` #46: Restore main's performance after #9/#29 (dataframe −48%	2026-10-28	1.76×	96.5%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`7d8c589` #48: perf-vm: x[[i]] <<- v in place (nbody_naive 1.32x -> 0.	2026-10-28	1.70×	96.5%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`414ae44` #45: perf-vm: <<- and x[[i]] <- v in compiled loops (geomean	2026-10-28	1.68×	96.5%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`1696902` #44: perf-vm: unboxed returns, reused argument vectors (fib	2026-10-28	1.53×	96.5%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`963b3cf` #32: perf-vm: caller-forced promises, %*% row vector (geomea	2026-10-28	1.50×	96.5%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`0e5d94e` #40: Measure pure-R package scripts with a pinned CRAN corpu	2026-10-28	1.49×	96.3%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`ad59f6f` #39: capi: load the libR.so stub from memory, closing a /tmp	2026-10-28	1.49×	96.3%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`abb8355` #38: CI: core and C API tests under gctorture	2026-10-28	1.49×	96.3%	33.9%	47.3%	12.5%	30.8%			70.0%		656/760
`37a5f96` #28: core: robigo-core heap and libR.so — packages' compiled	2026-10-28	1.55×	94.6%	30.5%	45.3%					70.0%		647/760
`1a78d2e` #34: parser: 2x parse throughput and a proper parse_console	2026-10-28	1.56×	94.6%	30.5%	45.3%					0.0%		647/760
`49c071a` #30: Preserve names attributes and global condition handlers	2026-10-28	1.47×	94.6%	30.5%	45.3%					0.0%		647/760
`fa86c67` #33: Make zones.R hermetic with vendored zoneinfo via TZDIR	2026-10-28	1.45×	94.6%	30.5%	45.3%					0.0%		647/760
`bf98504` #27: Preserve class attributes and supplied language argumen	2026-10-28	1.45×	94.6%	30.5%	45.3%					0.0%		647/760
`5b0b240` #23: perf-vm: x[[i, j]] in the VM, index promises forced in	2026-10-28	1.44×	94.5%	30.5%	44.6%					0.0%		647/760
`007110b` #26: conformance-strings: agrep/agrepl, grepl quantifiers	2026-10-28	1.37×	94.5%	30.5%	44.6%					0.0%		647/760
`d2158bd` #24: print/format/deparse: 390 -> 436 of 437 fastr cases	2026-10-28	1.37×	94.4%	30.5%	44.5%					0.0%		645/760
`8f5b7a7` #22: Match GNU R extraction, replacement, and array validati	2026-10-28	1.37×	94.0%	30.5%	44.1%					0.0%		644/760
`e8e2d31` #19: Match GNU R environment lookup semantics	2026-10-28	1.37×	94.0%	30.5%	44.1%					0.0%		644/760
`a4c7c3e` #17: distributions: non-central d/p/q, Tukey, Wilcoxon/signr	2026-10-28	1.36×	93.0%	30.5%	43.9%					0.0%		643/760
`0adb3db` #16: Make reduction precision selectable per interpreter	2026-10-28	1.36×	91.6%	27.1%	43.5%					0.0%		643/760
`10d0a81` docs/briefs: ready-to-paste agent briefs for the free tracks	2026-10-28	1.37×	91.6%	28.8%	43.6%					0.0%		643/760
`7bb554c` #15: startup: lazy-load databases, save/load on connections,	2026-10-27	1.37×	91.6%	28.8%	43.6%					0.0%		643/760
`7ee7200` #13: Design: robigo-core uses the mark-region heap (memory b	2026-10-27	1.38×	90.5%	28.8%	37.8%					0.0%		628/760
`d998e0b` #12: memory: the locality study settles the bake-off — mark-	2026-10-27	1.38×	90.5%	28.8%	37.8%					0.0%		628/760
`5e3b03a` #8: parser: byte-exact with GNU R — 100% on a 33k-source par	2026-10-27	1.38×	90.5%	28.8%	37.8%					0.0%		628/760
`5794f92` #10: corpus.py: keep R's temp files inside each case directo	2026-10-27	1.40×	90.5%	27.1%	37.7%					0.0%		628/760
`9633f1e` #7: memory bake-off: mark-region gc + conservative stack win	2026-10-27	1.40×	90.5%	27.1%	37.7%					0.0%		628/760
`135ed54` #6: cli: run files as R's console, with the echo transcript	2026-10-27	1.40×	90.5%	27.1%	37.7%					0.0%		628/760
`2166f66` #2: capi: the libR census and a compiled C API corpus (G2's	2026-10-27	1.40×	90.3%	0.0%	0.0%					0.0%		628/760
`28f108d` robigo: a GNU R 4.5 runtime in Rust, with verification infra	2026-10-27	1.40×	90.3%	0.0%	0.0%							628/760

robigo: progress

Speed

Compatibility

Top CRAN packages at 9b86fc7

Package scripts as users run them

Benchmarks at 8a33af5

Corpora at 8a33af5

Open gaps at 8a33af5

Every commit

Top CRAN packages at `9b86fc7`

Benchmarks at `8a33af5`

Corpora at `8a33af5`

Open gaps at `8a33af5`