#354 Move Workstation Ostree installer to separate variant
Merged 6 years ago by ausil. Opened 6 years ago by lsedlar.
lsedlar/pungi-fedora f27-ostree-inst  into  f27

file modified
+1 -1
@@ -802,7 +802,7 @@ 

              'failable': ['*'],

          }

      }),

-     ("^Workstation$", {

+     ("^WorkstationOstree$", {

          "x86_64": {

              "repo": "Everything",

              "release": None,

file modified
+8
@@ -133,6 +133,14 @@ 

              <package>fedora-productimg-workstation</package>

          </buildinstallpackages>

      </variant>

+     <variant id="WorkstationOstree" name="Workstation Ostree" type="variant" is_empty="true">

+         <arches>

+             <arch>aarch64</arch>

+             <arch>armhfp</arch>

+             <arch>i386</arch>

+             <arch>x86_64</arch>

+         </arches>

+     </variant>

      <variant name="Cloud" id="Cloud" type="variant">

          <arches>

              <arch>aarch64</arch>

Creating the image requires running lorax. There is however already one lorax instance being run in the buildinstall phase. When the second one runs, it overwrites the outputs from first one and people expecting to boot netinst over PXE are instead getting ostree workstation.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1477916

The same change is already applied on master branch.

Does this mean we build another entirely separate install tree just for this image? seems a bit hacky.

No, this just means the outputs of lorax run for the ostree installer will be put into compose/WorkstationOstree/x86_64/os and not overwrite anything. Just the location is changed, not the amount of work.

Pull-Request has been merged by ausil

6 years ago

Is this pull-request going to affect the branched tree of F27 ?
Currently we expect the image to be located in compose/Workstation/x86_64/iso/ .

@jkurik yes it will affect it. we will have to update the snc scripts also

oh, kay. looks fine, then.

gah, I only just noticed this creates a variant (and a subvariant, as the variant name winds up being used as the image's subvariant) with a space in its name. Is this really a good idea? It seems like it could potentially cause all kinds of problems for things parsing variant / subvariant names.

We've used an underscore in (sub)variant names before, e.g. Jam_KDE, Design_suite, Astronomy_KDE, Cloud_Base, etc etc etc...

@adamwill Good catch. That's a bug in Pungi, it should be using ID as a fallback, not name. As a workaround the name should be changed until the fix (pungi#755) is deployed. Changing the value to something else would be fine too, I guess.

Name itself containing spaces is not a problem. The name is meant as a human readable name for the variant, and at Red Hat there are many cases where it contains spaces. If something has a problem with that, it should be fixed in that tool.

Metadata