Package: protobuf
Version: 3.21.12-10+b1
severity: wishlist

Hi,

Recently, I've noticed that auto-protobuf transition was recorded,
but it is not proceeded on-going yet. [1]

mozc package depends to protobuf and for upgrading to newer Mozc, 
it requires more newer version of protobuf such as protobuf 3.25.4-1
in experimental.

The discussion about upgrading newer version of protobuf is on-going at
Bug#1034668 [2], not to bother discussion, I'll file another bug
to track current situation.

It would be nice abseil and protobuf transition goes smoothly.

Here is the summary of FTBFS about auto-protobuf (amd64).


Summary:

* FTBFS occurs when only protobuf in experimental was shipped in
  most cases.
* A few FTBFS occurs when both of protobuf and abseil in experimental 
were shipped. (The issue a. was resolved in many cases.)
* If already FTBFS occurs in sid, it also fails with experimental
  version of protobuf and abseil too.

Legends:

* a. sid: build with protobuf 3.21.12-10 and libabsl-dev 20230802.1-4
  in unstable.
* b. exp-protobuf: build with protobuf 3.25.4-1 in experimental
  and libabsl-dev 20230802.1-4 in unstable.
* c. exp-absl-protobuf: build with protobuf 3.25.4-1 in experimental
  and libabsl-dev 20240722.0-1 in experimental.
* OK means built successfully, NG means FTBFS or broken dependency.
  -- means not checked yet.

Round1:

  android-platform-tools    a: OK b: NG c: OK
  astroidmail               a: OK b: OK c: OK
  autofdo                   a: OK b: NG c: OK
  berkeley-express          a: OK b: NG c: OK
  clementine                a: NG b: NG c: NG
  cubemap                   a: OK b: NG c: OK
  gnss-sdr                  a: OK b: NG c: OK
  grpc                      a: NG b: NG c: NG
  harvest-tools             a: OK b: NG c: OK
  kosmindoormap             a: NG b: NG c: NG
  libarcus                  a: OK b: OK c: OK
  libcompizconfig           a: OK b: OK c: OK
  libphonenumber            a: OK b: NG c: OK
  marble                    a: OK b: OK c: OK
  merkaartor                a: OK b: OK c: OK
  mixxx                     a: OK b: OK c: OK
  mosh                      a: OK b: NG c: OK
  mozc                      a: OK b: NG c: NG
  mumble                    a: OK b: OK c: OK
  nageru                    a: OK b: OK c: OK
  ola                       a: NG b: NG c: NG
  onnx                      a: OK b: OK c: OK
  opencv                    a: OK b: NG c: OK
  orthanc                   a: OK b: OK c: OK
  osmpbf                    a: OK b: NG c: OK
  ostinato                  a: OK b: NG c: OK
  pink-pony                 a: OK b: NG c: OK
  pokerth                   a: OK b: NG c: OK
  protobuf-c                a: OK b: NG c: OK
  qgis                      a: OK b: OK c: OK
  r-cran-rprotobuf          a: OK b: OK c: OK
  spectrum2                 a: OK b: OK c: OK
  target-factory            a: OK b: NG c: OK
  usbguard                  a: OK b: OK c: OK
  vlc                       a: NG b: NG c: NG
  zbackup                   a: OK b: NG c: OK

Round2:

  android-platform-frameworks-base  a: OK b: OK c: OK
  bear                              a: OK b: NG c: OK
  cura-engine                       a: OK b: NG c: OK
  fastnetmon                        a: OK b: NG c: OK
  grpc-java                         a: OK b: NG c: OK
  justbuild                         a: OK b: NG c: OK
  llvm-toolchain-17                 a: OK b: NG c: --
  llvm-toolchain-18                 a: NG b: NG c: --
  llvm-toolchain-19                 a: -- b: -- c: --
  onnxruntime                       a: OK b: NG c: NG
  open-vm-tools                     a: OK b: NG c: OK
  opentelemetry-cpp                 a: OK b: NG c: OK
  orthanc-mysql                     a: OK b: OK c: OK
  orthanc-postgresql                a: OK b: OK c: OK
  qt6-grpc                          a: OK b: NG c: OK

FTBFS will happen if experimental version of protobuf and abseil were
shipped: (except already failed in sid)

* mozc
* onnxruntime

Regards,

[1] https://release.debian.org/transitions/html/auto-protobuf.html
[2] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1034668

Reply via email to