#152 Move arch specific package blacklist to their correct groups
Merged by dustymabe. Opened by siosm.
Unknown source master

Download 152.patch

Fixes: Blacklist x86_64 specific packages and make them arch specific

I both like this change and don't like this change. For one it puts things in more natural categories, but it also loses the comment # Blacklist x86_64 specific packages, they will be added back by fedora-common-ostree.yaml, which I think is an important/subtle detail. We need to blacklist them because they manually get added back later. See the last paragraph of the description of https://pagure.io/workstation-ostree-config/issue/150 for more info.

To be clear, this PR doesn't actually fix any issues, just re-orders things a bit, correct?

If you run comps-sync without this PR, you will get:

Syncing packages common to all desktops:
  - All manifest packages are already listed in comps.
  - 9 packages not in manifest:
    hyperv-daemons (mandatory, groups: guest-desktop-agents)
    mcelog (default, groups: workstation-product)
    microcode_ctl (mandatory, groups: workstation-product)
    open-vm-tools-desktop (mandatory, groups: guest-desktop-agents)
    virtualbox-guest-additions (mandatory, groups: guest-desktop-agents)
    xorg-x11-drv-intel (mandatory, groups: base-x)
    xorg-x11-drv-openchrome (mandatory, groups: base-x)
    xorg-x11-drv-vesa (mandatory, groups: base-x)
    xorg-x11-drv-vmware (mandatory, groups: base-x)

as the packages are not in the correct blacklist groups. This is not cosmetic. I understand that this is unexpected. Maybe we should change the script logic to take only one big blacklist which would make it "simpler".

I have updated the PR with the previous comments.

rebased onto 4925cc15946be49b99e5d1b99ca2464e28108e13

as the packages are not in the correct blacklist groups. This is not cosmetic.

ahh.. well I'm glad I asked :)

I have updated the PR with the previous comments.

thanks for adding the comments back - very useful for future versions of me and you reading back and trying to understand

LGTM

cc @mclasen

Pull-Request has been merged by dustymabe

Metadata