Skip to content

[tools] fix proot static binary build - #1

Merged
luisguzman-adfa merged 1 commit into
appdevforall:mainfrom
iiab:main
Jun 5, 2026
Merged

[tools] fix proot static binary build#1
luisguzman-adfa merged 1 commit into
appdevforall:mainfrom
iiab:main

Conversation

@luisguzman-adfa

@luisguzman-adfa luisguzman-adfa commented Jun 5, 2026

Copy link
Copy Markdown
Collaborator

cherry-pick fix for static binary builds

@luisguzman-adfa
luisguzman-adfa merged commit d7e1e59 into appdevforall:main Jun 5, 2026
1 check passed
luisguzman-adfa added a commit that referenced this pull request Jul 28, 2026
…er, tests

- #1 Guard overlayTitle in onDone (was an NPE if the overlay was hidden first).
- #3 Add the K2GoFQR bridge only on the /maps/ page (prepareForUrl, called before each load
  from PortalActivity's initial load + shouldOverrideUrlLoading) and remove it elsewhere,
  shrinking the JS surface (methods still guard on 'active').
- #4 Connect the delete MutationObserver ONLY while the trash tool is on (extract fires on
  Next now, so the observer is delete-only) -> no observer churn during normal map use.
- #6 Make isMapsPage pure-string (no android.net.Uri) and add package-private validName/
  validBox + boundsOf; unit-test them (FqrControllerTest, plain JVM + real org.json).
luisguzman-adfa added a commit that referenced this pull request Jul 29, 2026
…d log scrape

Review nits:
- #1: SetupProgressActivity launchMode=singleTask. Reopening (notification -> LibraryActivity
  -> open index) now brings the SINGLE existing index forward (resumes, state preserved) instead
  of stacking a second one on top of the backgrounded instance.
- #3: remove logShowsCleanFinish() and its logCleanFinish log field. It was computed only for the
  K2Go-Recover line (not the verdict) and scanned ~400 log lines each recovery — deadweight and
  unreliable (intermediate Ansible recaps). The K2Go-Recover line keeps marker/reachable/verdict.
luisguzman-adfa added a commit that referenced this pull request Jul 30, 2026
…n, lifecycle guard

- #1: mount the feedback FAB on LibraryActivity only at boot-gate dismiss (onServerReady), so it
  never floats over the boot animation / install progress; and remove it from SetupProgressActivity
  (the install index is itself a progress/gate surface). Wizard + Get More keep it (normal screens).
- #2: installOn(activity, screenTag, bottomMarginDp) overload — LibraryActivity passes 88dp to clear
  the bottom nav; nav-less screens use the default 16dp (flush).
- #4: guard the async screenshot callback with isFinishing()/isDestroyed() before using the activity.
luisguzman-adfa added a commit that referenced this pull request Jul 31, 2026
#1 Cancel: the notification Cancel now runs the full terminal cleanup (release EnvironmentLock,
post terminal) via a done-guard, instead of a bare teardown that stranded the environment stopped
and the lock held. Cancel is offered ONLY for backup (read-only); restore has no Cancel action
(destructive, hard gate). A clean restore clears InstallGuard; a FAILED restore leaves it set so
next-launch recovery repairs the torn rootfs.

#2 Boot handoff: the service no longer boots the environment. Mirroring InstallService (whose server
restart is owned by the install index), DeepOpService leaves it stopped and the hosting Activity boots
via serverController.startEnvironment() on return / next launch — so the service and ServerController
never both own the proot container. Removes sBootEngine and the now-unused EnvironmentControl.start().
luisguzman-adfa added a commit that referenced this pull request Jul 31, 2026
#1 CloneShareService: on a START_STICKY null-intent restart with no active clone (session + daemon
died with the process), stopSelf instead of resurrecting a stale 'transferring' notification / locks.
#2 CloneSendSession.begin: set the volatile active flag last so a reader that sees it also sees the
rest of the fields.
luisguzman-adfa added a commit that referenced this pull request Jul 31, 2026
#1 SyncStateViewModel.getTransport(): double-checked lazy init on a volatile field, so the now
process-global transport is never built twice by the main thread + a probe's IO thread (a stray
RsyncManager/rsync child a later stop() couldn't reach).
#2 SyncFragment.onDestroyView() (old UI): don't stop the shared transport / drop the network binding
while a clone is in flight — that old-UI teardown would otherwise kill the clone daemon and strand its
lock + keep-alive. Follow-ups noted in the PR: daemon liveness check on re-bind; CONFIRM-phase network
binding on onCleared.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants