Yeah, I updated my f33 test box today, and ~80 new packages got pulled in, including some big ones (scala with 22M alone, plus four different variants of openblas for some reason, adding up to 20M, vtk at 17M, opencv itself with 10M ...)

While I like that my Java packages reach a broader audience (/sarcasm) this is probably going to explode the Workstation image size (and others) by at least 100M or so :(


Giving negative karma so this doesn't go to stable, 1.46.0-2.fc32 should supersede the build in this update.

For the record: 1.46.0 breaks compilation of doctests in release+LTO mode (which is the default for RPM builds).

Is any human being reading the comments on this update?

Please either waive the tests (if they're false positive failures) or unpush the update (if they're really broken).

Not sure if it was a good idea to push a major version update to stable fedora. It messes up my "database" for some reason, removing all old entries and re-inserting them with new IDs.

I was told to include this build with the libhandy1 0.90.0 update, and I can't do that if it's already part of another update.

This update has been unpushed.

This update has been unpushed.

Either this update should be pulled because of the broken tests, or they should be waived ... right now, it's just sitting there with lots of positive karma but stuck in testing because of failed gating tests.


Seems to work fine (linux/x86_64 guests on x86_64 via virt-manager).

Works fine.

Well, this update actually fixes the broken postun scriptlet, but upon upgrade, the broken one from the old package version is run (which is unfortunate, but can't really be fixed now ...)

BZ#1853731 Garbage in scriptlet

Works fine (mounting, deleting subvolume, scrubbing, balancing).

BZ#1832911 btrfs-progs-5.6.1 is available
Test Case Mount btrfs

Is this update supposed to go to stable now? It was opened three years ago, so maybe that's not even a good idea anymore.


Seems to work fine.


Crashes on start (looks like it expects rpm headers as bytes but they are strs).

Works fine, and regression and performance tests pass on all tested systems: - AMD ryzen 1700X / NVidia GTX 1070 w/ negativo17 drivers - dell XPS 13 9370 (2018) - libvirt/KVM based fedora 32 VMs



python3-molecule from this update is not installable:

 Problem: cannot install the best candidate for the job
  - nothing provides pre-commit needed by python3-molecule-3.0.2-6.fc30.noarch

Also, is this built from the master branch? Because the f30 branch doesn't have these changes ...


Fixes the LoadError issue with did_you_mean.

BZ#1817178 "bundle install" fails to install nio4r gem by "did_you_mean (LoadError)".