Issue & PR triage
Loading...
Showing 83 of 83 issues
| Issue | Title | Author | Age | Action | Impact ↓ | Actionability | Quick Win | Summary |
|---|---|---|---|---|---|---|---|---|
|
craft-providers issue #430 |
Improve integration tests | mr-cal | 1105d | keep open | 72 | 68 | 11 | Integration tests remain disk-bound: LXD images consume nearly all GitHub runner storage, and the proposed Spread migration or pytest cleanup/parametrization overhaul is still unresolved. |
|
craft-providers issue #584 |
Catch `FileNotFound` for all `subprocess.run()` calls | mr-cal | 861d | keep open | 70 | 80 | 28 | Enhancement to catch FileNotFoundError from subprocess.run() calls across craft-providers and raise ProviderError for better user-facing error messages. Still open. |
|
craft-providers issue #508 |
LXD Provider requires a local installation of LXD on macOS | jat-canonical | 993d | keep open | 68 | 70 | 10 | Remote-LXD builds remain unavailable on macOS and Windows because provider checks still require a local LXD install, despite acknowledged demand and partial low-level remote support. |
|
craft-providers issue #895 |
Add a LXD VM provider | lengau | 275d | needs triage | 66 | 58 | 13 | Add a VM-backed LXD provider for Imagecraft loop-device support; the requested capability remains open and is absent from the current provider tree. |
|
craft-providers issue #828 |
Allow getting the apt proxy from the user's apt settings | lengau | 408d | keep open | 66 | 78 | 22 | Add support for propagating host APT proxies, including Proxy-Auto-Detect results refreshed during each machine warmup; the enhancement remains open and unimplemented. |
|
craft-providers issue #665 |
Add a callback hook to allow users to configure a base on warmup before... | lengau | 764d | close resolved | 65 | 91 | 34 | Merged PR #788 adds a prepare_instance callback before base setup, enabling early apt and proxy configuration; the callback is not literally inside warmup, but the requested capability is implemented. |
|
craft-providers issue #259 |
Cannot launch lxd instance with high uids | mr-cal | 1282d | close resolved | 65 | 95 | 40 | LXD launches failed for host UIDs above 65,535; commit 5717e2c and PR #780 now configure separate UID and GID raw.idmap entries, resolving the reported failure. |
|
craft-providers pr #948 |
test: failing reproducer - snap developer-id not asserted | lengau | 190d | needs review | 62 | 58 | 22 | Parses developer-id from snap-revision and asserts the developer account in _get_assertions_file (fix for #445); fix added by copilot-swe-agent. CI green, 0 reviewers, 1 unresolved comment, inactive 34 days. |
|
craft-providers issue #830 |
Provide an asynchronous API for starting and handling instances | lengau | 407d | keep open | 62 | 40 | 5 | Triaged enhancement to add a non-blocking API for launching/handling provider instances, unblocking parallel multi-base builds. No async code has landed; core design (true async vs non-blocking) still undecided. |
|
craft-providers issue #531 |
Provide a way to set Multipass CPU and memory | mr-cal | 960d | keep open | 60 | 80 | 34 | Enhancement request to make Multipass CPU and memory configurable instead of hardcoded 2 CPU/2GB, which currently causes slow or failed builds. |
|
craft-providers issue #381 |
Improve integration tests | mr-cal | 1170d | keep open | 58 | 74 | 10 | Integration-test overhaul is partially implemented: weekly macOS fast/slow jobs and pytest timing exist, but base-instance behavior tests and full base/provider coverage remain absent. |
|
craft-providers issue #684 |
Create a generic craft LXD profile | dilyn-corner | 749d | needs triage | 58 | 50 | 13 | A dedicated craft LXD profile is proposed to isolate build containers from user-edited defaults; the open enhancement still needs a decision on shared versus per-user profiles. |
|
craft-providers issue #880 |
Support LXD backend on non-Linux hosts | lengau | 318d | keep open | 57 | 64 | 12 | Maintainer-authored enhancement to use the LXD provider on non-Linux hosts through a configured remote or VM. Remote arguments exist, but availability checks still require local LXD. |
|
craft-providers issue #454 |
Using default compatibility tag can cause DNS name conflicts | mr-cal | 1084d | keep open | 56 | 89 | 28 | Missing compatibility tags still produce identical LXD base-instance names across projects, causing DNS conflicts and failed launches; the proposed warning and API redesign remain unimplemented. |
|
craft-providers issue #789 |
Checking the compatibility tag should be more robust | bepri | 444d | keep open | 55 | 68 | 23 | Open maintainer enhancement: compatibility checks still treat missing, None, or empty tags as compatible instead of failing or requiring investigation. |
|
craft-providers issue #391 |
LXD issue when build multiple charms in parallel | beliaev-maksim | 1154d | close resolved | 55 | 78 | 33 | Parallel LXD charm builds failed with a snapd 'daemon is stopping' error during `snap watch`. Fixed by retrying snap watch in commit e0f6bc7 (PR #899), released in 3.3.0; issue still open awaiting closure. |
|
craft-providers issue #907 |
Rename base instances to template instances | lengau | 260d | keep open | 55 | 90 | 21 | Planned API-wide rename of “base instance” to “template instance,” including provider methods, hooks, launch behavior, and on-disk names; current code still uses the base terminology. |
|
craft-providers pr #1010 |
fix: improve robustness of compatibility tag | cmatsuoka | 61d | needs review | 52 | 74 | 22 | Rejects instances with missing compatibility tags, centralizes config loading and error handling, and updates tests; lint and fast checks pass while five slow CI jobs remain pending. |
|
craft-providers issue #765 |
Set up weekly tests for tests that don't need to run on each PR | lengau | 486d | close resolved | 52 | 90 | 29 | Weekly Linux LXD integration coverage now exists, but the issue remains open despite its implementation in PR #857. |
|
craft-providers issue #445 |
`developer-id` of the `snap-revision` is not asserted | mr-cal | 1086d | keep open | 52 | 94 | 30 | Host snap injection still omits the snap-revision developer account, causing transferred snaps such as Rockcraft 1194 to fail assertion; open PR #948 proposes a tested fix. |
|
craft-providers issue #544 |
Managed instance names don't take the base into account | tigarmo | 980d | keep open | 52 | 90 | 29 | Managed instances still do not persist and compare build-base, allowing a reused Ubuntu devel instance to bypass version validation; the triaged fix is to store it in InstanceConfiguration. |
|
craft-providers issue #625 |
Drop Centos 7 support | lengau | 804d | keep open | 51 | 88 | 26 | CentOS 7 remains exposed by craft-providers despite Charmcraft 4 removing its use; this valid enhancement still needs the provider base, alias, remote image, and tests removed. |
|
craft-providers issue #452 |
macOS CI is failing in PRs | mr-cal | 1084d | keep open | 50 | 55 | 7 | macOS PR CI fails without logs; qa.yaml multipass/smoketest jobs commented out as a temporary workaround (#982), but root cause (runner timeout/disconnect) unfixed and weekly macOS jobs still fail per #467. Reopened by bepri. |
|
craft-providers issue #884 |
Enable Ubuntu minimal builds | lengau | 317d | keep open | 48 | 65 | 18 | Maintainer-filed enhancement to let the LXD backend use Ubuntu minimal images; LXD still hard-maps Ubuntu aliases to coreXX images, with no implementation evidence. |
|
craft-providers issue #517 |
Multipass: Use native mounts where available | lengau | 973d | keep open | 48 | 90 | 29 | Multipass mounts still use classic SSHFS because the provider does not detect the active driver or pass `--type native`; this open enhancement targets faster mounts on Hyper-V, QEMU, and LXD. |
|
craft-providers issue #211 |
multipass: better error handling when images are unavailable | mr-cal | 1360d | keep open | 48 | 78 | 22 | Multipass launch gives generic errors when Snapcraft images are unreachable or unsupported on arm64; this acknowledged enhancement remains open with a downstream report and internal ticket. |
|
craft-providers issue #399 |
Loading the instance config should be more robust | mr-cal | 1140d | keep open | 48 | 68 | 22 | InstanceConfiguration still rejects unknown YAML fields, preventing compatibility-tag cleanup when a config was written by a newer Charmcraft version; no final design has been chosen. |
|
craft-providers issue #646 |
Consider switching from using the LXD CLI to pylxd | lengau | 790d | keep open | 48 | 46 | 6 | Partially underway: pylxd handles some interactions, but most provider operations still use the LXC CLI; maintainers deferred reassessment until the LXD 6.x migration. |
|
craft-providers issue #794 |
Warn if requested specs don't match the existing instance | bepri | 443d | needs triage | 48 | 46 | 11 | Existing instances are reused without validating requested hardware settings; a maintainer asked for elaboration, but the mismatch policy and covered specifications remain unspecified. |
|
craft-providers issue #821 |
Unmount nonexistent directories on start | lengau | 415d | keep open | 48 | 58 | 13 | Open, triaged enhancement to remove stale host-directory mounts before starting reused instances; missing-path validation is proposed, but the full decision rule remains unspecified. |
|
craft-providers issue #714 |
craft-providers error: Timed out waiting for networking to be ready. | nteodosio | 657d | needs triage | 48 | 45 | 8 | Open Type: Bug report: Snapcraft 8.5.1 LXD builds time out while getent checks DNS for snapcraft.io; host/network context and a part YAML are missing, leaving the root cause unknown. |
|
craft-providers pr #932 |
Work/ci linux multipass | lengau | 224d | keep open | 45 | 22 | 4 | Maintainer WIP enabling Multipass integration tests on Linux CI (uncommenting test-multipass in qa.yaml); +293/-178 over 5 files. Idle 196 days, 0 reviews/comments; policy and macOS slow-integration checks failing. Tracks open issue #754. |
|
craft-providers issue #796 |
Improve approach to waiting for Multipass start | bepri | 443d | keep open | 45 | 75 | 24 | Enhancement: replace hardcoded 20s sleep after Multipass install with polling until multipassd is ready. TODO still in installer.py, triaged, but may be blocked by upstream canonical/multipass#1995. |
|
craft-providers pr #900 |
feat(base): cache uv directory | carlcsaposs-canonical | 265d | needs review | 45 | 72 | 24 | Adds uv cache directory mounting to Base._mount_shared_cache_dirs (mirroring merged pip-cache PR #394), fixing charmcraft#2461. Approved, lint fixed via merged #910, slow tests still pending. |
|
craft-providers issue #281 |
multipass: do not always use sudo to execute commands | mr-cal | 1255d | keep open | 45 | 88 | 29 | MultipassInstance still always rootifies execute_run() and execute_popen() through _rootify_multipass_command; the requested per-call use_sudo option remains absent. |
|
craft-providers issue #946 |
Install snapd from a user-defined channel | mr-cal | 191d | keep open | 45 | 77 | 19 | Triaged enhancement to honor CRAFT_SNAPD_CHANNEL by refreshing and tracking snapd; current code still apt-installs snapd, and snapd injection was suggested. |
|
craft-providers issue #449 |
Handle SIGINT | mr-cal | 1086d | keep open | 43 | 76 | 18 | Maintainer-labeled enhancement remains open: Ctrl-C during LXD base setup can leave a non-ephemeral base running; current code stops many setup failures but does not guarantee removal. |
|
craft-providers issue #304 |
Base _setup_wait_for_system_ready should use systemctl's --wait parameter | lengau | 1238d | keep open | 42 | 74 | 19 | Enhancement to replace polling of systemctl is-system-running with native --wait; it remains unimplemented and needs CentOS 7 handling plus a Xenial compatibility decision. |
|
craft-providers pr #723 |
feat(bases): make BaseName compatible with craft-platforms | lengau | 631d | close superseded | 42 | 10 | 2 | Adds BaseName convenience methods and DistroBase-compatible API types. Open for 603 days without reviews or reported CI and apparently superseded by craft-platforms#44. |
|
craft-providers pr #876 |
ci: add weekly tests on a variety of distros | lengau | 320d | keep open | 42 | 12 | 3 | Adds a weekly LXD integration-test matrix for Debian, Fedora, Rocky Linux, CentOS Stream, AlmaLinux, openSUSE, and Mint; every new distro job is failing and the unreviewed PR has had no activity for 293 days. |
|
craft-providers issue #212 |
Incorrectly handling SNAPCRAFT_MAX_PARALLEL_BUILD_COUNT | stolk | 1360d | keep open | 42 | 64 | 10 | SNAPCRAFT_MAX_PARALLEL_BUILD_COUNT not honored and Multipass instances hardcoded to 2 CPUs. Maintainer triaged; still unfixed in craft-providers (cpus=2 at multipass_provider.py:300), provider-side work split out to #531, snapcraft-side partly fixed in 8.4.4. |
|
craft-providers pr #824 |
feat: add a `get_base` factory function | lengau | 414d | needs review | 42 | 76 | 17 | Adds a public get_base factory with documentation and tests; all 13 CI checks pass, but after 386 days it remains unreviewed with no comments. |
|
craft-providers issue #119 |
Provide fixtures / some helpers to test an application using craft-providers | jugmac00 | 1641d | needs triage | 42 | 48 | 8 | Enhancement request for reusable public test fixtures/helpers that simplify downstream mocking of elaborate LXD JSON responses; remains open, uncommented, and not maintainer-triaged. |
|
craft-providers issue #873 |
Switch from external tmate to the Canonical tmate action | lengau | 321d | keep open | 42 | 86 | 33 | Triaged refactor: replace mxschmitt/action-tmate@v3 with canonical/action-tmate in 3 workflow steps (check-renovate.yaml; tests-weekly.yaml x2). Still unmodified at HEAD; prerequisite for self-hosted runners. |
|
craft-providers issue #583 |
Documentation for network connectivity | mr-cal | 861d | keep open | 40 | 80 | 26 | Open documentation request to add network connectivity troubleshooting guides for LXD and Multipass. The how-to docs are empty; only a partial Docker/iptables note exists in explanation/. |
|
craft-providers pr #1026 |
fix: include unknown Ubuntu version in compatibility warnings | Shahzeb-DataEngineer | 14d | needs review | 40 | 35 | 10 | Improves compatibility warnings to include unknown Ubuntu versions instead of raising errors. Has 3 unresolved review comments and no reviewers yet. |
|
craft-providers issue #703 |
Allow pointing to daily or release buildd images for LXD | mr-cal | 728d | keep open | 40 | 76 | 19 | Enhancement to let LXD select Ubuntu buildd daily or release images, including through app configuration, remains open; the current remote is hard-coded per Ubuntu alias. |
|
craft-providers pr #998 |
ci: enable and stabilize multipass tests in QA | lengau | 77d | needs review | 40 | 48 | 11 | Maintainer CI PR enabling multipass integration tests in qa.yaml on a large self-hosted runner (LXD disabled), wiring setup-tests to install-multipass, waiting on multipassd with diagnostics, and fixing bytes-vs-string and cache-mount test assertions. All listed checks pass; 5 review comments unresolved. |
|
craft-providers issue #754 |
Run Multipass tests on Linux CI | lengau | 513d | keep open | 40 | 70 | 11 | Enhancement to run Multipass integration tests on Linux CI (nested virt now available). Still open: qa.yaml's test-multipass job remains commented out and WIP PR #932 is inactive with failing checks. |
|
craft-providers issue #766 |
Switch centos and almalinux containers to use images.lxd.canonical.com | lengau | 486d | keep open | 40 | 82 | 26 | Maintainer-filed enhancement (lengau, labelled Type: Enhancement) asking to point CentOS 7 and AlmaLinux 9 LXD remotes at images.lxd.canonical.com; code at HEAD still uses images.linuxcontainers.org. Unimplemented, no discussion. |
|
craft-providers issue #383 |
bumping the compatibility tag should be more robust | mr-cal | 1169d | keep open | 38 | 84 | 22 | Tests still hardcode buildd-base-v7 while production derives it from Base.compatibility_tag, leaving compatibility-tag bumps vulnerable to inconsistent test updates. |
|
craft-providers pr #1011 |
chore: reduce verbosity of timer logging | cmatsuoka | 61d | needs review | 38 | 45 | 14 | Changes the LXD instance heartbeat from 3 to 30 seconds and removes per-update timer logging; one slow test is failing and three review comments remain unresolved. |
|
craft-providers issue #858 |
Error when setting up almalinux test | lengau | 327d | close resolved | 38 | 87 | 19 | AlmaLinux unit setup exposed a leaked InstanceTimer thread exception; cleanup was later fixed by #996, leaving this open issue apparently resolved. |
|
craft-providers issue #213 |
Use alternate lxd profiles | cmatsuoka | 1353d | keep open | 38 | 45 | 9 | Enhancement request (by maintainer cmatsuoka) to allow alternate LXD profiles, e.g. CRAFT_PROVIDERS_LXD_PROFILE(S), for devs with multiple pools. mr-cal replied with design questions (plural list, global vs per-app) still unanswered; no code implements profiles. |
|
craft-providers issue #969 |
BuilddBaseAlias enum lookup raises bare ValueError — should raise a typed error | tonyandrewmeyer | 149d | close resolved | 38 | 94 | 28 | Resolved: compatibility checks now catch unknown host and guest Ubuntu VERSION_IDs and skip validation instead of leaking BuilddBaseAlias ValueError; no typed error was added. |
|
craft-providers issue #795 |
Verify present addresses when adding a remote | bepri | 443d | keep open | 38 | 68 | 16 | Enhancement remains open: RemoteImage.add_remote checks only remote_name, so an existing address under another LXC remote name is not detected; address-level handling and tests are missing. |
|
craft-providers issue #942 |
Feature Request for custom software mirror | Kxuan | 196d | keep open | 38 | 58 | 13 | Requests a craft-providers env var (like legacy SNAPCRAFT_BUILD_ENVIRONMENT_PRIMARY_MIRROR) to point Ubuntu build instances at a custom APT mirror; unimplemented, 0 comments, no maintainer response, still untriaged. |
|
craft-providers pr #808 |
chore: prepare for Python 3.14 | lengau | 440d | close superseded | 35 | 15 | 3 | Adds Python 3.14 and free-threaded CI coverage, but has 14 failing checks and was tracked as superseded by craft-providers#1509 after PyO3 0.26 support landed. |
|
craft-providers issue #929 |
Multipass is needed for TICS coverage | bepri | 227d | close resolved | 35 | 80 | 21 | TICS workflow needed Multipass installed to run coverage. Resolved by merged PR #991 (6c4b99c), which added an install-multipass Makefile target wired into setup-tics; subsequent TICS run succeeded. Only the flaky multipass_instance test module remains excluded. |
|
craft-providers issue #901 |
Remove mypy and pyright | mr-cal | 265d | keep open | 35 | 88 | 23 | Drop mypy/pyright in favour of ty. Makefile, CI and pyproject already use ty only; common.mk still defines lint-mypy, lint-pyright, install-pyright, and stale .gitignore/type: ignore entries remain. Mirrors snapcraft#6007 and craft-cli#420, both completed. |
|
craft-providers issue #582 |
Add support for Xenial | jugmac00 | 862d | keep open | 35 | 60 | 7 | Enhancement request to add Xenial (16.04) support to craft-providers. Mapping is missing from remotes.py and snapd commands need version abstraction for older snapd on Xenial. Issue acknowledged by maintainer as non-trivial scope. |
|
craft-providers issue #793 |
Warn about unused parameters in subprocess calls | bepri | 443d | keep open | 35 | 82 | 20 | LXC and Multipass still omit timeout and check when exec uses subprocess.Popen, so the requested warning for silently discarded parameters remains unimplemented. |
|
craft-providers issue #467 |
macOS CI is failing weekly tests | mr-cal | 1065d | needs triage | 35 | 25 | 3 | Maintainer-tracked CI issue: macOS runners drop connection after ~1h losing all logs, so flaky macOS integration tests were pushed out of PR CI (#452) into the weekly workflow. Still open, no logs or reproducer captured. |
|
craft-providers issue #657 |
check the storage space before trying to spin container | dariuszd21 | 770d | keep open | 35 | 32 | 6 | Maintainer-filed enhancement to warn (or fail early) when host/LXD storage is nearly full before spinning up a build container; still open, with ownership and a concrete threshold unresolved after lengau questioned whether craft-providers is the right place. |
|
craft-providers issue #450 |
Timer thread logging is overly verbose | mr-cal | 1086d | keep open | 33 | 86 | 22 | Maintainer-filed enhancement to quiet timer-thread output and refresh every 30 seconds instead of 5; the current LXD readiness loop still logs full status and timer details every 3 seconds. |
|
craft-providers issue #711 |
Rebase craft-providers on starbase | mattculler | 660d | close resolved | 32 | 72 | 6 | Maintainer request to rebase craft-providers on canonical/starbase for a unified dev approach; completed via #786/#787 and follow-up starbase merges, and CI has since moved to starflow. |
|
craft-providers pr #931 |
ci: add support for Python 3.14 | lengau | 224d | needs review | 32 | 52 | 13 | Adds Python 3.14 to CI matrices (qa.yaml, weekly macOS workflows, likely lockfile constraints). 18/19 checks pass; integration-slow-macos (3.14, macos-14-large) fails. No reviewers, no comments, idle 196 days. |
|
craft-providers issue #172 |
Check for docker upstream iptables installation for proper error message | sergiusens | 1452d | keep open | 32 | 66 | 11 | Enhancement to detect docker-installed iptables rules that break LXD networking and fail early with a fix suggestion. Tracked as CRAFT-1683; not implemented in source (only docs added). |
|
craft-providers issue #312 |
Integration tests are slow due to fixture ordering | mr-cal | 1233d | keep open | 32 | 62 | 14 | Enhancement: integration tests provision LXD instances before the installed_snap fixture can skip, wasting time; current test_snap_installer.py signatures still list core22_lxd_instance first. Unaddressed. |
|
craft-providers issue #116 |
Better parameter documentation for `execute_run` -> `command` | jugmac00 | 1645d | keep open | 32 | 88 | 21 | execute_run still documents command only as “Command to execute”; users need explicit argument-list, escaping, and shell-feature guidance, leaving this triaged documentation request unresolved. |
|
craft-providers issue #970 |
Test on Python 3.14 and Ubuntu 26.04 (Resolute) | lengau | 148d | needs triage | 30 | 60 | 14 | Request to add Python 3.14 and Ubuntu 26.04 testing to CI. A previous PR (#971) failed and was closed, leaving the tracking issue open without maintainer review or labels. |
|
craft-providers pr #949 |
test: failing reproducer - no error for high UID (>65535) | lengau | 190d | keep open | 30 | 25 | 7 | Adds failing reproducer test test_set_id_map_raises_for_high_uid showing _set_id_map() lacks a uid>65535 check; the actual fix is described but NOT implemented. Red by design on 4 test-basic jobs, 0 reviews, idle 163 days. |
|
craft-providers issue #468 |
Check for a running process should be more fine-grained | mr-cal | 1065d | keep open | 28 | 82 | 18 | Enhancement: the LXD launcher's creator-PID liveness check only tests Path('/proc/<pid>').exists(); suggested fix is to also verify /proc/<pid>/cmdline is a python process so PID reuse triggers immediate cleanup instead of a 60s wait. Still unimplemented. |
|
craft-providers issue #792 |
Create an inner class for LXD variables | bepri | 443d | keep open | 28 | 58 | 8 | Maintainer-filed enhancement to group LXD launcher variables into an inner class/dataclass to shrink launch()'s long parameter lists; still open, approach under discussion. |
|
craft-providers issue #791 |
Move troubleshooting info on explanation index | bepri | 443d | keep open | 28 | 84 | 18 | Documentation cleanup remains open: network troubleshooting is still embedded in the explanation index, and NetworkError still links to that index. |
|
craft-providers pr #881 |
ci: add Linux multipass weekly tests | lengau | 318d | close superseded | 25 | 10 | 2 | Adds Linux Multipass jobs (latest/candidate, latest/edge) to tests-weekly.yaml; both new linux-multipass checks fail, no reviews or activity for 290 days. Superseded by WIP PR #932. |
|
craft-providers issue #894 |
docs: odd Explanation tree/page rendering | steinbro | 279d | keep open | 22 | 82 | 13 | Docs bug: the "Failure to properly execute commands that depend on network access" section lives inline in docs/explanation/index.rst with a _network-error label but is not in the toctree, so it renders in the page body yet not in the sidebar. |
|
craft-providers pr #817 |
ci: enable macos for multipass tests | lengau | 420d | close stale | 15 | 10 | 1 | Adds macOS (macos-14-large) to the multipass slow-test CI matrix (+10/-10 in the old tests.yaml). Inactive 386 days; macos-14-large and minimum-deps jobs fail and it targets a workflow file since deleted. |
|
craft-providers issue #420 |
docs: bring all starter pack documentation checks into tox | lengau | 1119d | close stale | 15 | 90 | 11 | Starter-pack documentation checks now run through the Make/uv docs-lint workflow; tox was removed, making the request to integrate them into tox obsolete. |
|
craft-providers pr #950 |
test: failing reproducer - no warning for default compatibility tag | lengau | 190d | keep open | 10 | 28 | 3 | Adds a 21-line failing reproducer test (test_default_compat_tag_logs_warning) for issue #454: BuilddBase with the default compatibility tag logs no warning. No fix included; fast/min-deps CI red, unreviewed for 163 days. |
|
craft-providers issue #271 |
LXD integration tests incompatible with canonical/setup-lxd | mr-cal | 1261d | close stale | 8 | 20 | 1 | Old request to make LXD integration tests work with canonical/setup-lxd user/group permissions. The tests.yaml workflow that motivated it was deleted (b2b9b81) and CI moved to Starflow, making the request obsolete. |
|
craft-providers pr #859 |
fix: add explicit test IDs to environment parametrizations | Copilot | 327d | close superseded | 5 | 8 | 0 | Adds explicit pytest IDs to three unit-test files; inactive for 300 days and unreviewed. It does not fix the reported timer failure, which was addressed separately by #996. |
|
craft-providers pr #1033 |
fix(lxd): retry instance start on forkstart errors | GreyViperTooth | 2d | ? | ? | ? |