--- Begin Message ---
Package: libxdamage1
Version: 1:1.1.6-1+b1
Severity: normal
File: /usr/share/doc/libxdamage1/changelog.gz
I want to install a deb package of amd64 architecture, which depends on libxdamage1, but the version of
libxdamage1:loong64(1:1.1.6-1+b1) and libxdamage1:amd64(1:1.1.6-1+b2) do not match.
I think a binary-only update of libxdamage1:loong64 to 1:1.1.6-1+b2 should fix this issue.
-- System Information:
Debian Release: 13.0
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: loong64 (loongarch64)
Foreign Architectures: amd64
Kernel: Linux 6.12.27-loong64 (SMP w/8 CPU threads)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=zh_CN.UTF-8, LC_CTYPE=zh_CN.UTF-8 (charmap=UTF-8), LANGUAGE=zh_CN:zh
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled
Versions of packages libxdamage1:loong64 depends on:
ii libc6 2.41-7
ii libx11-6 2:1.8.12-1
libxdamage1:loong64 recommends no packages.
libxdamage1:loong64 suggests no packages.
-- no debconf information
--- End Message ---
--- Begin Message ---
- To: <1105850-done@bugs.debian.org>
- Subject: Re: Bug#1105850: /usr/share/doc/libxdamage1/changelog.gz: version of libxdamage1:loong64 doesn't match the one of amd64
- From: Drew Parsons <dparsons@debian.org>
- Date: Fri, 30 May 2025 23:01:04 +0200
- Message-id: <E1uL6qe-0000000CY5n-3i7t@sandy>
Hi Jianfeng, thanks for the bug report, and thanks for caring about package quality.
In this case however, there is no bug.
The +b version tag you're seeing indicates a binNMU [1],
a binary non-maintainer upload.
This means the package is rebuilt with no changes made to the source,
for instance to update the build against new binary-incompatible
versions of libraries it depends on.
A binNMU may be needed for various reasons, and not all architectures
always need the same rebuild. So the +b number may get out of sync
on different architectures.
In this case, loong64 is still not completely stable and is not always
keeping up with package builds alongside the other architectures.
Likely what happened here was that loong64 was not able to build
the package at the time the +b1 binNMU was made on the other architectures.
When it did build, it was the first time it built against the original
source, so it did not need a +b suffix.
Then when another binNMU was needed later, the other architectures
went from +b1 to +b2, while it was the first binNMU for loong64 so it got +b1.
No bug here, so closing the bug.
Thanks again for reporting though.
Drew
[1] https://wiki.debian.org/binNMU
--- End Message ---