Skip to content

Record your screen, ship a demo. Free and open-source, GPU-accelerated, no watermarks, no subscriptions. Windows, macOS, Linux. Actively maintained.

ci(nix): require every sources attempt to report a display

EtienneLescot9bc6e1d47a76
diff --git a/.github/workflows/nix-build.yml b/.github/workflows/nix-build.yml
index 5749ade40..e4a6a1e9e 100644
--- a/.github/workflows/nix-build.yml
+++ b/.github/workflows/nix-build.yml
@@ -436,6 +436,14 @@ jobs:
# xvfb-run folds the command's stderr into stdout, so the stderr
# capture is empty and every milestone lands in the other file.
grep -a milestone "/tmp/sources.$i.out" || echo " (none at all)"
+ elif [ "$RC" -eq 0 ] && ! python3 -c "import json,sys; sys.exit(0 if json.load(open(sys.argv[1]))['displays'] else 1)" "/tmp/sources.$i.json"; then
+ # Exit 0 is not an answer on its own. Electron 43 settles the
+ # combined screen+window call after its 3s ready timer with the
+ # screen dropped, so `displays: []` is #462's other face, and
+ # Xvfb always has one screen to report.
+ FAILED=$((FAILED + 1))
+ echo "returned no display (#462). Last milestone:"
+ grep -a milestone "/tmp/sources.$i.out" | tail -1 || echo " (none)"
elif [ "$RC" -eq 0 ]; then
OK=$((OK + 1))
LAST_OK="/tmp/sources.$i.json"
@@ -470,10 +478,8 @@ jobs:
PY
fi
- # Noted, not acted on. Enumeration is flaky here and the export
- # assertion below is the more
- # important question; a flaky attempt costs an annotation rather than
- # the answer. Verdict at the end, once both have had their say.
+ # Annotated here, decided at the end: export runs either way, so a
+ # failed enumeration cannot hide whether the addon works.
if [ "$HUNG" -gt 0 ]; then
echo "::warning::openscreen sources had to be killed on $HUNG of $ATTEMPTS attempts. Check the milestones above for how far it got before the outer bound fired."
fi
@@ -600,21 +606,18 @@ jobs:
# flaky must not hide whether export works, which is the whole point of
# having packaged the compositor addon.
#
- # The gate is "did enumeration ever answer" and "does export work", not
- # "did all five attempts pass". Requiring FAILED -eq 0 made the job red
- # by construction: the standing numbers on this runner are 1/5, 3/5 and
- # 4/5 ok, so a run where export is perfect and four enumerations succeed
- # still failed. A post-merge check that is red by design is one that gets
- # muted, which costs more than the flakiness it was reporting. The
- # per-attempt warnings above keep that flakiness visible without letting
- # it decide the build; tighten this to $ATTEMPTS once the capture failure
- # is understood and fixed.
+ # Every enumeration must answer with a display, and export must work.
+ # The gate used to be "did enumeration ever answer", because the call
+ # failed on 1 to 4 attempts in 5 for a reason nobody had found. #462
+ # found it: asking for screens and windows in one getSources call is
+ # what fails under X11, and the get-sources handler now asks one type
+ # at a time (under Xvfb: 0 failures in 32 runs, against 7 in 16
+ # combined). A flake left after that is a regression, not runner luck.
#
- # Export, on the other hand, is a hard gate on every trigger again. It no
- # longer depends on a capability this host does not have, so there is
- # nothing left to excuse: if it fails now, the package is broken.
+ # Export is a hard gate on every trigger. It does not depend on a
+ # capability this host lacks, so if it fails, the package is broken.
echo "=== verdict: enumeration $OK/$ATTEMPTS ok, export $EXPORT_OK, $RUN_CLI_N run_cli invocations ==="
- if [ "$EXPORT_OK" -ne 1 ] || [ "$OK" -eq 0 ]; then
+ if [ "$EXPORT_OK" -ne 1 ] || [ "$OK" -ne "$ATTEMPTS" ]; then
exit 1
fi