I recently added this because there's a need for a development-focused
sync chat instance and the CNCF Slack makes sense alongside
other CNCF projects.
Signed-off-by: Colin Walters <walters@verbum.org>
Literally just find and replace github.com/containers/bootc with
github.com/bootc-dev/bootc
We have the redirect in place so none of this is really important, but
I figure removing as many instances of the old repo and replacing them
with the current can't hurt for things like search engine
optimization. Plus some non-zero number of people might assume one is
a fork of the other or something.
Signed-off-by: John Eckersberg <jeckersb@redhat.com>
Useful when running automatic reports and analyses
where all issues need to be visible at once without truncation.
To do this we needed to add a LintExecutionConfig to
propagate configuration through the lint system.
Also, refactor how we print things so there's a
shared helper functions for consistent formatting with and without truncation.
Closes: https://github.com/bootc-dev/bootc/issues/1260
Signed-off-by: Colin Walters <walters@verbum.org>
Prep for lint truncation improvement.
- Kill the handling of the "max 0" case, it's pathological; this
requires passing NonZeroUsize
- Drop the requirement for the iterator to be clonable as it
improves ergonomics, and return a vector of the collected elements
instead
Signed-off-by: Colin Walters <walters@verbum.org>
Especially fix the intro and note that bootc's APIs are stable.
Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Colin Walters <walters@verbum.org>
For historical reasons the ostree sepolicy API can exist as a
no-op even if it didn't find a policy, one has to query `.csum()`
or `.name()` to verify it's present.
In our code just map that case to None.
Followup to 99d30dfd55
to ensure we consistently handle this case.
Signed-off-by: Colin Walters <walters@verbum.org>
Part of #904
Displays pinned deployments as part of "bootc status".
Includes unit tests to ensure correct parsing of the
pinned deployments, and that they are displayed in
human readable formats correctly.
Signed-off-by: Robert Sturla <robertsturla@outlook.com>
Main motivation: I was looking at making more changes here
- Use an idiom I'd like to standardize more of copy context to `FROM scratch` image
which is then mounted and consumed in other phases by mounting. This helps
avoid polluting later containers with intermediate copied files.
- Change `build.sh` to handle being run from any directory
- Drop the `dev-rootfs` stuff as it's weird and awkward; instead we should
encourage multi-step builds deriving from this image
- Don't make `bootc.tar.zst` only to immediately untar it; just use `COPY`
from the build container
- Use heredocs to condense multiple `RUN` invocations to avoid pointless
small layers
Signed-off-by: Colin Walters <walters@verbum.org>
Check for the policy csum to make sure there is actually a policy
available to lookup when trying to label the bootc container storage.
Fixes #1303
Signed-off-by: ckyrouac <ckyrouac@redhat.com>
As we head towards factory reset work, this will be of critical
importance.
For now this only shows up in the API, not in human readable
status.
Signed-off-by: Colin Walters <walters@verbum.org>
In some cases rust-analyzer seems to be building without the `bootc`
feature which triggers some unused variable warnings here.
Signed-off-by: Colin Walters <walters@verbum.org>
This uses findmnt to locate filesystem mounts that are on the same
source as the root mount. If any are found, the user is warned these
filesystems will persist unmounted in the bootc system. The user must
hit <enter> to proceed.
This does the same for logical volumes in the same group as root.
It also adds a generic warning to help the user understand what will
happen after rebooting into the bootc system.
Signed-off-by: ckyrouac <ckyrouac@redhat.com>
