Horton
Menu

Re: Heads Up: cirrus-ci is shutting down June 1st

Henson Choi <assam258@gmail.com>

View thread
From: Henson Choi <assam258@gmail.com>
To: Nazir Bilal Yavuz <byavuz81@gmail.com>
Cc: Jacob Champion <jacob.champion@enterprisedb.com>, Andres Freund <andres@anarazel.de>, Peter Eisentraut <peter@eisentraut.org>, Jelte Fennema-Nio <postgres@jeltef.nl>, Thomas Munro <thomas.munro@gmail.com>, pgsql-hackers@postgresql.org, Zsolt Parragi <zsolt.parragi@percona.com>
Date: 2026-06-29T00:25:51Z
Lists: pgsql-hackers 
Hi,

Since cfbot keeps coming up here, a quick bug report.

cfbot treats a .tgz attachment as a patch even when the archive has no
patch inside.  It then applies nothing, tests plain master, and reports
green -- so a real patch posted earlier drops out of testing and the CF
entry shows a false green.

Live example: CF 5802 (Unicode Normalization).  The last real patchset
is v11.  My next message attached coverage.tgz (a gcov HTML report, not
a patch); cfbot took that as the latest submission and has been testing
master ever since.  That's the reproducer -- and the reason 5802 looks
"stale/merged" on the dashboard when it isn't.

Suggested fix: only treat an attachment as a patch if it's a
.patch/.diff, or an archive containing one; otherwise ignore it and fall
back to the last attachment set that had patches.

Regards,
Henson

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. ci: Add GitHub Actions based CI

  2. ci: Remove support for cirrus-ci based CI