composetree: allow url in manifest package list #1847
Conversation
|
☔ The latest upstream changes (presumably 79dfcea) made this pull request unmergeable. Please resolve the merge conflicts. |
|
Ahh looks like #1745 conflicted. Mind rebasing? |
This allows URLs to be used in the manifest file when doing a treecompose only in unified-core mode. This is a convenient way of adding one off packages for testing purposes but should not be the preferred method of composing a tree. The import local rpm functions were also moved to rpmostree-importer to facilitate usage.
mike-nguyen
force-pushed
the
cached_package_url
branch
from
4f5dc90
to
bee767b
Compare
|
Rebased |
| } | ||
|
|
||
| gboolean | ||
| rpmostree_importer_import_many_local_rpms (OstreeRepo *repo, |
There was a problem hiding this comment.
I think given that this in in the importer now, maybe just rpmostree_importer_import_rpms ?
There was a problem hiding this comment.
Let's also make this code move a separate prep commit? I think that'll clean up the history nicely make it easier to review the critical changes.
There was a problem hiding this comment.
I think given that this in in the importer now, maybe just
rpmostree_importer_import_rpms?
Sounds good. I was trying to think of a better but I am terrible with naming.
Let's also make this code move a separate prep commit? I think that'll clean up the history nicely make it easier to review the critical changes.
Will do.
| /* We download URI packages and import them into the pkgcache first so we can | ||
| add it to the local packages to install */ | ||
| g_autoptr(GUnixFDList) fd_list = g_unix_fd_list_new (); | ||
| for (char **it = pkgnames; it && *it; it++) |
There was a problem hiding this comment.
Hmm, were you able to try out the cached-packages approach we discussed? Should mostly be about taking this diff hunk here and putting it on the compose side, and then feeding the NEVRAs that the importer returns into the cached-packages field.
There was a problem hiding this comment.
Hmm, were you able to try out the
cached-packagesapproach we discussed? Should mostly be about taking this diff hunk here and putting it on the compose side,
@jlebon Can you clarify on what you mean by putting it on the compose side?
and then feeding the NEVRAs that the importer returns into the
cached-packagesfield.
Should I create another cached-packages array and add the NEVRAS in the loop at https://github.com/projectatomic/rpm-ostree/pull/1847/files#diff-07963b1a5922690400970963652cbe3bR1967 ?
|
@mike-nguyen: PR needs rebase. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
@mike-nguyen: The following tests failed, say
Full PR test history. Your PR dashboard. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here. |
This allows URLs to be used in the manifest file when doing a
treecompose only in unified-core mode. This is a convenient way of
adding one off packages for testing purposes but should not be the
preferred method of composing a tree.
The import local rpm functions were also moved to rpmostree-importer
to facilitate usage.