Repository navigation
Hotfix Beta Release: 11/20/26 - #3375
Conversation
|
Channel deleted. |
Test Results 73 files 572 suites 0s ⏱️ Results for commit 7105b7d. ♻️ This comment has been updated with latest results. |
|
zigbee-lock_coverage.xml
Minimum allowed coverage is Generated by 🐒 cobertura-action against 7105b7d |
|
zwave-lock_coverage.xml
Minimum allowed coverage is Generated by 🐒 cobertura-action against 7105b7d |
ctowns
left a comment
There was a problem hiding this comment.
Looks good!
Just a small note: commit 8854e87 has a name that implied it is a merge commit, but it actually seems to have been squashed and merged directly as a single commit via the squash and merge function. Just noting this for later since it somewhat breaks the naming convention
ctowns
left a comment
There was a problem hiding this comment.
Actually nevermind, shouldn't this be the commit we are pulling forward? ed92d83
Not the merge commit but the actual commit? I'm not sure why it still shows the changes in the diff but in the past we have cherry-picked the actual change commits, not the merge commits.
|
I don't think it matters really? It's the difference between Both take forward the changes. I can take over the other commit if you'd prefer. |
|
Yeah I'd prefer cherry-picking the other functional commit because that is how we have handled cherry-picks in this repo, and also it makes the git history cleaner and provides a more descriptive commit name. Even yesterday as a part of the prod release, we can see that actual commits were cherry-picked rather than squashed merge commit with names like "Merge ___ from ___", so at least the name should be updated to match what is actually being cherry-picked (not the merge commit) |
|
I think there is an argument to be made that links to a PR with the full context of the change are more clear. I will make the change. |
84a04df to
979f3ec
Compare
|
Duplicate profile check: Warning - duplicate profiles detected. |
|
Profile category check: ✅ Passed - all profiles have a category defined. |
These locks never support codes, so none of the legacy lockCodes handling applies to them. Promote the subdriver to the top level, along with its configurations.lua, and mark its devices as SLGA_MIGRATED so legacy-handlers does not also handle them. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
A Schlage lock refuses a duplicate PIN with STATUS_NOT_AVAILABLE, which now returns duplicate instead of success. A masked occupied report now confirms the add or update pending on that slot instead of being dropped, so the next add no longer overwrites the PIN. The manufacturer-specific report check is folded into the report handler. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: Pawel <41388251+kainpets@users.noreply.github.com>
979f3ec to
7105b7d
Compare
|
matter-lock_coverage.xml
Minimum allowed coverage is Generated by 🐒 cobertura-action against 7105b7d |
|
zigbee-switch_coverage.xml
Minimum allowed coverage is Generated by 🐒 cobertura-action against 7105b7d |
I wholeheartedly agree - Those links can be added to the commit itself by adding the #PR value, which is something that is done automatically during via the squash and merge UI. I think we should always try squash and merge when we can we can because it 1) adds this PR name automatically, and 2) it ideally reduces the number of cherry-picks needed now that we find ourselves doing cherry-picks very frequently. I think a descriptive commit header with the PR# included is the best option, and this is something that the squash and merge UI provides, but its also something we can add directly to commits as well in the future, so it's a good call out. For now my suggestion was made to try to enforce some of the conventions of the repo, and I agree that adding PR# through either squash and merge or via the commit message in a merged commit is the best option for documentation. |
Description of Change
Rolling up critical Zigbee/Zwave lock changes to beta.