Hi,
last month's status: https://lists.opensuse.org/archives/list/factory@lists.opensuse.org/thread/Z...
Last months' reproducible builds project updates (including my work): https://reproducible-builds.org/reports/2021-04/
This time I spent most time on rebuilding and fixing issues in 15.3 https://rb.zq1.de/leap/15.3/ has some stats.
It was a bit more difficult than 15.2, because older SLE binaries do not get rebuilt, and even Leap is split between Backports and Leap project. I also ran my scripts on https://build.opensuse.org/project/show/openSUSE:Step:FrontRunner to check what stops building there and found many issues.
I uploaded https://rb.zq1.de/compare.factory-20210531/ today and rbstats are: total-packages: 14026 (+71) build-tried: 13989 (+44) build-failed: 58 (-5) build-n-a: 164 (-9) build-succeeded: 13767 (+58) build-official-failed+na: 398 (+136) build-compare-failed: 365 (-1) build-compare-succeeded: 13402 (+59) verify-failed: 468 (-27) verified-semi-reproducible: 12588 (-372) bit-by-bit-identical: 13126 (-65) not-bit-by-bit-identical: 641 (+106)
https://rb.zq1.de/compare.factory-20210531/graph.png shows the change over time
https://rb.zq1.de/compare.factory-20210531/unreproduciblerings.txt lists very unreproducible core packages (bootstrap+DVD)
Of the badly unreproducible packages, 3 were in ring0 46 were in ring1
That makes it 49/3247 => 1.51 % which is below the overall average of which is below the overall average of 365/13767 => 2.65 %
641/13767 => 4.66 % of packages are not perfectly reproducible
newly unreproducible core packages: kopete is again/still unreproducible from https://bugreports.qt.io/browse/QTBUG-83186
python-pandas suffers from the usual .pyc non-determinism
javassist has filesys-order-related variations in javadoc index-all.html
Ciao Bernhard M.
Hi,
last month's status: https://lists.opensuse.org/archives/list/factory@lists.opensuse.org/thread/I...
Last months' reproducible builds project updates (including my work): https://reproducible-builds.org/reports/2021-05/
This month we improved reproducibility of our kernel packages via https://bugzilla.suse.com/show_bug.cgi?id=1187167 though I still need to test the fix for the embedded random signing key.
Most of the bugreports and fixes I did this month are about being able to build in the year 2036 and with -j1
https://bugzilla.opensuse.org/show_bug.cgi?id=1185637 could still use some love in that area. I tried to extend the certs, but for some reason this did not make tests pass.
Additionally, I pushed several fixes towards 15.3
I uploaded https://rb.zq1.de/compare.factory-20210630/ today and rbstats are: total-packages: 14082 (+56) build-tried: 14053 (+64) build-failed: 33 (-25) build-n-a: 163 (-1) build-succeeded: 13857 (+90) build-official-failed+na: 248 (-150) build-compare-failed: 359 (-6) build-compare-succeeded: 13498 (+96) verify-failed: 475 (+7) verified-semi-reproducible: 13079 (+491) bit-by-bit-identical: 13206 (+80) not-bit-by-bit-identical: 640 (-1)
https://rb.zq1.de/compare.factory-20210630/graph.png shows the change over time
https://rb.zq1.de/compare.factory-20210630/unreproduciblerings.txt lists very unreproducible core packages (bootstrap+DVD)
Of the badly unreproducible packages, 3 were in ring0 41 were in ring1
That makes it 44/3247 => 1.36 % which is below the overall average of 359/13857 => 2.59 %
640/13857 => 4.62 % of packages are not perfectly reproducible
Ciao Bernhard M.
Hi,
last month's status: https://lists.opensuse.org/archives/list/factory@lists.opensuse.org/message/...
Last months' reproducible builds project updates (including my work): https://reproducible-builds.org/reports/2021-06/
Not much exciting happened this month. I'm still seeing and reporting many unreliable builds, especially in python modules.
I uploaded https://rb.zq1.de/compare.factory-20210730/ today and rbstats are: total-packages: 14070 (-12) build-tried: 14042 (-11) build-failed: 36 (+3) build-n-a: 198 (+35) build-succeeded: 13808 (-49) build-official-failed+na: 29 (-219) build-compare-failed: 372 (+13) build-compare-succeeded: 13436 (-62) verify-failed: 481 (+6) verified-semi-reproducible: 13096 (+17) bit-by-bit-identical: 13161 (-45) not-bit-by-bit-identical: 637 (-3)
https://rb.zq1.de/compare.factory-20210730/graph.png shows the change over time
https://rb.zq1.de/compare.factory-20210730/unreproduciblerings.txt lists very unreproducible core packages (bootstrap+DVD)
Of the badly unreproducible packages, 3 were in ring0 43 were in ring1
That makes it 46/3247 => 1.42 % which is below the overall average of 372/13808 => 2.69 %
637/13808 => 4.61 % of packages are not perfectly reproducible
Ciao Bernhard M.
Hi,
this was an uneventful month
last month's status: https://lists.opensuse.org/archives/list/factory@lists.opensuse.org/message/...
Last months' reproducible builds project updates (including my work): https://reproducible-builds.org/reports/2021-07/
I uploaded https://rb.zq1.de/compare.factory-20210830/ today and rbstats are: total-packages: 14127 (+57) build-tried: 14098 (+56) build-failed: 25 (-11) build-n-a: 215 (+17) build-succeeded: 13858 (+50) build-official-failed+na: 115 (+86) build-compare-failed: 373 (+1) build-compare-succeeded: 13485 (+49) verify-failed: 469 (-12) verified-semi-reproducible: 13126 (+30) bit-by-bit-identical: 13221 (+60) not-bit-by-bit-identical: 628 (-9)
https://rb.zq1.de/compare.factory-20210830/graph.png shows the change over time
https://rb.zq1.de/compare.factory-20210830/unreproduciblerings.txt lists very unreproducible core packages (bootstrap+DVD)
Of the badly unreproducible packages, 4 were in ring0 40 were in ring1
That makes it 44/3247 => 1.36 % which is below the overall average of 373/13858 => 2.69 %
628/13858 => 4.53 % of packages are not perfectly reproducible
newly unreproducible: grep: parallelism via PGO introduced via https://build.opensuse.org/request/show/912393
Ciao Bernhard M.
Hi,
last month's status: https://lists.opensuse.org/archives/list/factory@lists.opensuse.org/message/...
Last months' reproducible builds project updates (including my work): https://reproducible-builds.org/reports/2021-08/
I uploaded https://rb.zq1.de/compare.factory-20210930/ today and rbstats are: total-packages: 14153 (+26) build-tried: 14123 (+25) build-failed: 39 (+14) build-n-a: 292 (+77) build-succeeded: 13792 (-66) build-official-failed+na: 167 (+52) build-compare-failed: 386 (+13) build-compare-succeeded: 13406 (-79) verify-failed: 462 (-7) verified-semi-reproducible: 13194 (+68) bit-by-bit-identical: 13294 (+73) not-bit-by-bit-identical: 500 (-128) not-bit-by-bit-identicalcheck: 498 (-139)
https://rb.zq1.de/compare.factory-20210930/graph.png shows the change over time
https://rb.zq1.de/compare.factory-20210930/unreproduciblerings.txt lists very unreproducible core packages (bootstrap+DVD)
Of the badly unreproducible packages, 4 were in ring0 38 were in ring1
That makes it 42/3247 => 1.29 % which is below the overall average of 386/13792 => 2.80 %
500/13792 => 3.63 % of packages are not perfectly reproducible
I have been busy with other work, so not much progress on openSUSE reproducible builds.
Newly unreproducible core packages:
* openssl-1_0_0 * librsvg
Ciao Bernhard M.
-
Bernhard M. Wiedemann