Announcing fmt library soversion bump

[Vitaly Zaitsev]

Hello.

fmt 9.0.x update will include a soversion bump from .8 to .9. It has 
both API and ABI changes.

Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0

The list of affected packages in Rawhide:
0ad
OpenImageIO
cantera
ceph
dolphin-emu
domoticz
folly
libsonata
mkvtoolnix
sdrpp
spdlog
vcpkg
zswap-cli

I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will need 
proven-packagers assistance.

Please use the f37-build-side-54844 side tag. I will merge it into 
Rawhide later.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Ben Beasley]

I don’t believe this list is nearly complete. Two packages I maintain 
that would be affected (fmidi and giada) are absent from the list.

Using the following command:

     mock -r fedora-rawhide-x86_64 --dnf-cmd -- repoquery --whatrequires 
'libfmt.so.8*'

I found these additional dependent packages that were missing from the 
original list:

      * bear
      * cachelib
      * easyeffects
      * easyrpg-player
      * fb303
      * fbthrift
      * fizz
      * fmidi
      * freeopcua
      * gerbera
      * giada
      * gnuradio
      * gr-funcube
      * libsemigroups
      * luxcorerender
      * mcrouter
      * nheko
      * proxygen
      * rstudio
      * wangle
      * watchman
      * waybar

I will go ahead and rebuild bear, fmidi, and giada into the side tag.

On 7/10/22 13:43, Vitaly Zaitsev via devel wrote:
>  Hello.
>
>  fmt 9.0.x update will include a soversion bump from .8 to .9. It has 
>  both API and ABI changes.
>
>  Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
>
>  The list of affected packages in Rawhide:
>  0ad
>  OpenImageIO
>  cantera
>  ceph
>  dolphin-emu
>  domoticz
>  folly
>  libsonata
>  mkvtoolnix
>  sdrpp
>  spdlog
>  vcpkg
>  zswap-cli
>
>  I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will 
>  need proven-packagers assistance.
>
>  Please use the f37-build-side-54844 side tag. I will merge it into 
>  Rawhide later.
>

[Ben Beasley]

Two of the three packages I tried to rebuild (bear and fmidi) failed 
with errors like:

/usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot format an argument.
To make type T formattable provide a formatter<T>
specialization:https://fmt.dev/latest/api.html#udt
  1733 |       formattable,
       |       ^~~~~~~~~~~

Links to failing scratch builds in the side tag:

    https://koji.fedoraproject.org/koji/taskinfo?taskID=89380682

    https://koji.fedoraproject.org/koji/taskinfo?taskID=89380736

I found that there was already a bug report open upstream for Bear: 
https://github.com/rizsotto/Bear/issues/471

If that FTBFS rate is typical of the full set of affected packages, then 
merging the side tag without resolving the API incompatibilities in 
dependent packages will be pretty disruptive to Rawhide as a whole.

Would you consider setting up a COPR to figure out how many packages are 
affected by the API changes, and filing Bugzillas on the affected 
dependent packages? It might take a while for affected package 
maintainers to become aware of the problem, investigate, and perhaps 
work with upstreams to prepare the necessary patches.

Link to fmt 9.0.0 release notes: 
https://github.com/fmtlib/fmt/releases/tag/9.0.0

– Ben

On 7/11/22 12:21, Ben Beasley wrote:
>  I don’t believe this list is nearly complete. Two packages I maintain 
>  that would be affected (fmidi and giada) are absent from the list.
>
>  Using the following command:
>
>      mock -r fedora-rawhide-x86_64 --dnf-cmd -- repoquery 
>  --whatrequires 'libfmt.so.8*'
>
>  I found these additional dependent packages that were missing from the 
>  original list:
>
>        * bear
>        * cachelib
>        * easyeffects
>        * easyrpg-player
>        * fb303
>        * fbthrift
>        * fizz
>        * fmidi
>        * freeopcua
>        * gerbera
>        * giada
>        * gnuradio
>        * gr-funcube
>        * libsemigroups
>        * luxcorerender
>        * mcrouter
>        * nheko
>        * proxygen
>        * rstudio
>        * wangle
>        * watchman
>        * waybar
>
>  I will go ahead and rebuild bear, fmidi, and giada into the side tag.
>
>  On 7/10/22 13:43, Vitaly Zaitsev via devel wrote:
> > Hello.
> >
> > fmt 9.0.x update will include a soversion bump from .8 to .9. It has 
> > both API and ABI changes.
> >
> > Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
> >
> > The list of affected packages in Rawhide:
> > 0ad
> > OpenImageIO
> > cantera
> > ceph
> > dolphin-emu
> > domoticz
> > folly
> > libsonata
> > mkvtoolnix
> > sdrpp
> > spdlog
> > vcpkg
> > zswap-cli
> >
> > I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will 
> > need proven-packagers assistance.
> >
> > Please use the f37-build-side-54844 side tag. I will merge it into 
> > Rawhide later.
> >
>
>  _______________________________________________
>  devel mailing list --devel(a)lists.fedoraproject.org
>  To unsubscribe send an email todevel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List
> Archives:https://lists.fedoraproject.org/archives/list/[e-mail]...
>  Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure

[Aleksei Bavshin]

On 7/11/22 09:50, Ben Beasley wrote:
>  Two of the three packages I tried to rebuild (bear and fmidi) failed
>
>  with errors like:
>  
>  /usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot format an
> argument. To make type T formattable provide a formatter<T>
> specialization:https://fmt.dev/latest/api.html#udt
>    1733 |       formattable,
>         |       ^~~~~~~~~~~
>  
>  Links to failing scratch builds in the side tag:
>  
>      https://koji.fedoraproject.org/koji/taskinfo?taskID=89380682
>  
>      https://koji.fedoraproject.org/koji/taskinfo?taskID=89380736
>  
>  I found that there was already a bug report open upstream for Bear: 
>  https://github.com/rizsotto/Bear/issues/471
>  
Thanks for bringing attention to that. I suspect the following fmt 
changelog entry:

"Disabled automatic std::ostream insertion operator (operator<<) 
discovery when fmt/ostream.h is included to prevent ODR violations. You 
can get the old behavior by defining FMT_DEPRECATED_OSTREAM but this 
will be removed in the next major release. Use fmt::streamed or 
fmt::ostream_formatter to enable formatting via std::ostream instead."

That is a VERY breaking change and there will be more packages affected. 
It's likely also affecting dependents of spdlog which use external fmt 
and <spdlog/fmt/ostr.h>.

The fix is simple, but not backward-compatible:
    template <> struct fmt::formatter<type> : ostream_formatter {};
(a custom fmt::formatter for the type would work as well).

I already confirmed that `waybar` is broken and will submit a fix 
upstream when I have time. Setting up a copr repo to find all other 
breakages sounds like a good idea.
>  If that FTBFS rate is typical of the full set of affected packages,
> then 
>  merging the side tag without resolving the API incompatibilities in 
>  dependent packages will be pretty disruptive to Rawhide as a whole.
>  
>  Would you consider setting up a COPR to figure out how many packages are 
>  affected by the API changes, and filing Bugzillas on the affected 
>  dependent packages? It might take a while for affected package 
>  maintainers to become aware of the problem, investigate, and perhaps 
>  work with upstreams to prepare the necessary patches.
>  
>  Link to fmt 9.0.0 release notes: 
>  https://github.com/fmtlib/fmt/releases/tag/9.0.0
>  
>  – Ben
>  
>  On 7/11/22 12:21, Ben Beasley wrote:
> >
> > I don’t believe this list is nearly complete. Two packages I maintain 
> > that would be affected (fmidi and giada) are absent from the list.
> >
> > Using the following command:
> >
> >     mock -r fedora-rawhide-x86_64 --dnf-cmd -- repoquery 
> > --whatrequires 'libfmt.so.8*'
> >
> > I found these additional dependent packages that were missing from the 
> > original list:
> >
> >       * bear
> >       * cachelib
> >       * easyeffects
> >       * easyrpg-player
> >       * fb303
> >       * fbthrift
> >       * fizz
> >       * fmidi
> >       * freeopcua
> >       * gerbera
> >       * giada
> >       * gnuradio
> >       * gr-funcube
> >       * libsemigroups
> >       * luxcorerender
> >       * mcrouter
> >       * nheko
> >       * proxygen
> >       * rstudio
> >       * wangle
> >       * watchman
> >       * waybar
> >
> > I will go ahead and rebuild bear, fmidi, and giada into the side tag.
> >
> > On 7/10/22 13:43, Vitaly Zaitsev via devel wrote:
> >> Hello.
> >>
> >> fmt 9.0.x update will include a soversion bump from .8 to .9. It has 
> >> both API and ABI changes.
> >>
> >> Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
> >>
> >> The list of affected packages in Rawhide:
> >> 0ad
> >> OpenImageIO
> >> cantera
> >> ceph
> >> dolphin-emu
> >> domoticz
> >> folly
> >> libsonata
> >> mkvtoolnix
> >> sdrpp
> >> spdlog
> >> vcpkg
> >> zswap-cli
> >>
> >> I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will 
> >> need proven-packagers assistance.
> >>
> >> Please use the f37-build-side-54844 side tag. I will merge it into 
> >> Rawhide later.
> >>
> >
> > _______________________________________________
> > devel mailing list --devel(a)lists.fedoraproject.org
> > To unsubscribe send an email todevel-leave(a)lists.fedoraproject.org
> > Fedora Code of
> Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> > List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
> > List
> Archives:https://lists.fedoraproject.org/archives/list/[e-mail]...
> > Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
>  
>  _______________________________________________
>  devel mailing list -- devel(a)lists.fedoraproject.org
>  To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List Archives:
> https://lists.fedoraproject.org/archives/list/[e-mail]
>  Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
>

[Frederic Berat]

Hello,

>  Setting up a copr repo to find all other breakages sounds like a good
idea.

I'm actually working on a tool to help with such use cases.
It may be a bit early since I only made an internal alpha release, and only
planned to go public by the end of this month with it, but since that may
help, feel free to give it a try and provide early feedback.

https://copr.fedorainfracloud.org/coprs/fberat/mass-prebuild/

Long story short, the tool aims to create for you mass rebuild around a
given set of packages. It will first build the packages you feed it (like
an SRPM), then try to calculate the direct reverse dependencies, and build
them against your new package version.
If builds are failing, it will create another pristine build for state
comparison, which should help reduce the investigation (even though it
can't be perfect).

All of that is using COPR infrastructure, and COPR python interface.

You have more details in this description:
https://gitlab.com/fberat/mass-prebuild/-/blob/dfcb3b1b5ae1384e963b56e4bc...

Fred.

On Tue, Jul 12, 2022 at 7:28 AM Aleksei Bavshin <alebastr(a)fedoraproject.org&gt;
wrote:

>  On 7/11/22 09:50, Ben Beasley wrote:
>  > Two of the three packages I tried to rebuild (bear and fmidi) failed
>  > with errors like:
>  >
>  > /usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot
>  format an argument. To make type T formattable provide a formatter<T>
>  specialization:https://fmt.dev/latest/api.html#udt
>  >   1733 |       formattable,
>  >        |       ^~~~~~~~~~~
>  >
>  > Links to failing scratch builds in the side tag:
>  >
>  >     https://koji.fedoraproject.org/koji/taskinfo?taskID=89380682
>  >
>  >     https://koji.fedoraproject.org/koji/taskinfo?taskID=89380736
>  >
>  > I found that there was already a bug report open upstream for Bear:
>  > https://github.com/rizsotto/Bear/issues/471
>
>  Thanks for bringing attention to that. I suspect the following fmt
>  changelog entry:
>
>  "Disabled automatic std::ostream insertion operator (operator<<)
>  discovery when fmt/ostream.h is included to prevent ODR violations. You
>  can get the old behavior by defining FMT_DEPRECATED_OSTREAM but this
>  will be removed in the next major release. Use fmt::streamed or
>  fmt::ostream_formatter to enable formatting via std::ostream instead."
>
>  That is a VERY breaking change and there will be more packages affected.
>  It's likely also affecting dependents of spdlog which use external fmt
>  and <spdlog/fmt/ostr.h>.
>
>  The fix is simple, but not backward-compatible:
>      template <> struct fmt::formatter<type> : ostream_formatter {};
>  (a custom fmt::formatter for the type would work as well).
>
>  I already confirmed that `waybar` is broken and will submit a fix
>  upstream when I have time. Setting up a copr repo to find all other
>  breakages sounds like a good idea.
>  > If that FTBFS rate is typical of the full set of affected packages, then
>  > merging the side tag without resolving the API incompatibilities in
>  > dependent packages will be pretty disruptive to Rawhide as a whole.
>  >
>  > Would you consider setting up a COPR to figure out how many packages are
>  > affected by the API changes, and filing Bugzillas on the affected
>  > dependent packages? It might take a while for affected package
>  > maintainers to become aware of the problem, investigate, and perhaps
>  > work with upstreams to prepare the necessary patches.
>  >
>  > Link to fmt 9.0.0 release notes:
>  > https://github.com/fmtlib/fmt/releases/tag/9.0.0
>  >
>  > – Ben
>  >
>  > On 7/11/22 12:21, Ben Beasley wrote:
>  >>
>  >> I don’t believe this list is nearly complete. Two packages I maintain
>  >> that would be affected (fmidi and giada) are absent from the list.
>  >>
>  >> Using the following command:
>  >>
>  >>     mock -r fedora-rawhide-x86_64 --dnf-cmd -- repoquery
>  >> --whatrequires 'libfmt.so.8*'
>  >>
>  >> I found these additional dependent packages that were missing from the
>  >> original list:
>  >>
>  >>       * bear
>  >>       * cachelib
>  >>       * easyeffects
>  >>       * easyrpg-player
>  >>       * fb303
>  >>       * fbthrift
>  >>       * fizz
>  >>       * fmidi
>  >>       * freeopcua
>  >>       * gerbera
>  >>       * giada
>  >>       * gnuradio
>  >>       * gr-funcube
>  >>       * libsemigroups
>  >>       * luxcorerender
>  >>       * mcrouter
>  >>       * nheko
>  >>       * proxygen
>  >>       * rstudio
>  >>       * wangle
>  >>       * watchman
>  >>       * waybar
>  >>
>  >> I will go ahead and rebuild bear, fmidi, and giada into the side tag.
>  >>
>  >> On 7/10/22 13:43, Vitaly Zaitsev via devel wrote:
>  >>> Hello.
>  >>>
>  >>> fmt 9.0.x update will include a soversion bump from .8 to .9. It has
>  >>> both API and ABI changes.
>  >>>
>  >>> Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
>  >>>
>  >>> The list of affected packages in Rawhide:
>  >>> 0ad
>  >>> OpenImageIO
>  >>> cantera
>  >>> ceph
>  >>> dolphin-emu
>  >>> domoticz
>  >>> folly
>  >>> libsonata
>  >>> mkvtoolnix
>  >>> sdrpp
>  >>> spdlog
>  >>> vcpkg
>  >>> zswap-cli
>  >>>
>  >>> I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will
>  >>> need proven-packagers assistance.
>  >>>
>  >>> Please use the f37-build-side-54844 side tag. I will merge it into
>  >>> Rawhide later.
>  >>>
>  >>
>  >> _______________________________________________
>  >> devel mailing list --devel(a)lists.fedoraproject.org
>  >> To unsubscribe send an email todevel-leave(a)lists.fedoraproject.org
>  >> Fedora Code of Conduct:
>  https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  >> List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
>  >> List Archives:
>  https://lists.fedoraproject.org/archives/list/[e-mail]
>  >> Do not reply to spam on the list, report it:
>  https://pagure.io/fedora-infrastructure
>  >
>  > _______________________________________________
>  > devel mailing list -- devel(a)lists.fedoraproject.org
>  > To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
>  > Fedora Code of Conduct:
>  https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>  > List Archives:
>  https://lists.fedoraproject.org/archives/list/[e-mail]
>  > Do not reply to spam on the list, report it:
>  https://pagure.io/fedora-infrastructure
>  _______________________________________________
>  devel mailing list -- devel(a)lists.fedoraproject.org
>  To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct:
>  https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List Archives:
>  https://lists.fedoraproject.org/archives/list/[e-mail]
>  Do not reply to spam on the list, report it:
>  https://pagure.io/fedora-infrastructure
>

[Michel Alexandre Salim]

Hi Vitaly,

Will try and take care of folly ASAP - thanks!

-- 
Michel

[Ben Beasley]

Both bear and fmidi are now rebuilt into the side tag. I defined the 
preprocessor macro FMT_DEPRECATED_OSTREAM as a temporary workaround 
(noting that this will not be available in fmt 10.0). I filed an issue 
upstream with fmidi: https://github.com/jpcima/fmidi/issues/5

On 7/11/22 12:50, Ben Beasley wrote:
>  Two of the three packages I tried to rebuild (bear and fmidi) failed 
>  with errors like:
>
>  /usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot format an
> argument. To make type T formattable provide a formatter<T>
> specialization:https://fmt.dev/latest/api.html#udt
>    1733 |       formattable,
>         |       ^~~~~~~~~~~
>
>  Links to failing scratch builds in the side tag:
>
>      https://koji.fedoraproject.org/koji/taskinfo?taskID=89380682
>
>      https://koji.fedoraproject.org/koji/taskinfo?taskID=89380736
>
>  I found that there was already a bug report open upstream for Bear: 
>  https://github.com/rizsotto/Bear/issues/471
>
>  If that FTBFS rate is typical of the full set of affected packages, 
>  then merging the side tag without resolving the API incompatibilities 
>  in dependent packages will be pretty disruptive to Rawhide as a whole.
>
>  Would you consider setting up a COPR to figure out how many packages 
>  are affected by the API changes, and filing Bugzillas on the affected 
>  dependent packages? It might take a while for affected package 
>  maintainers to become aware of the problem, investigate, and perhaps 
>  work with upstreams to prepare the necessary patches.
>
>  Link to fmt 9.0.0 release notes: 
>  https://github.com/fmtlib/fmt/releases/tag/9.0.0
>
>  – Ben
>
>  On 7/11/22 12:21, Ben Beasley wrote:
> >
> > I don’t believe this list is nearly complete. Two packages I maintain 
> > that would be affected (fmidi and giada) are absent from the list.
> >
> > Using the following command:
> >
> >     mock -r fedora-rawhide-x86_64 --dnf-cmd -- repoquery 
> > --whatrequires 'libfmt.so.8*'
> >
> > I found these additional dependent packages that were missing from 
> > the original list:
> >
> >       * bear
> >       * cachelib
> >       * easyeffects
> >       * easyrpg-player
> >       * fb303
> >       * fbthrift
> >       * fizz
> >       * fmidi
> >       * freeopcua
> >       * gerbera
> >       * giada
> >       * gnuradio
> >       * gr-funcube
> >       * libsemigroups
> >       * luxcorerender
> >       * mcrouter
> >       * nheko
> >       * proxygen
> >       * rstudio
> >       * wangle
> >       * watchman
> >       * waybar
> >
> > I will go ahead and rebuild bear, fmidi, and giada into the side tag.
> >
> > On 7/10/22 13:43, Vitaly Zaitsev via devel wrote:
> >> Hello.
> >>
> >> fmt 9.0.x update will include a soversion bump from .8 to .9. It has 
> >> both API and ABI changes.
> >>
> >> Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
> >>
> >> The list of affected packages in Rawhide:
> >> 0ad
> >> OpenImageIO
> >> cantera
> >> ceph
> >> dolphin-emu
> >> domoticz
> >> folly
> >> libsonata
> >> mkvtoolnix
> >> sdrpp
> >> spdlog
> >> vcpkg
> >> zswap-cli
> >>
> >> I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will 
> >> need proven-packagers assistance.
> >>
> >> Please use the f37-build-side-54844 side tag. I will merge it into 
> >> Rawhide later.
> >>
> >
> > _______________________________________________
> > devel mailing list --devel(a)lists.fedoraproject.org
> > To unsubscribe send an email todevel-leave(a)lists.fedoraproject.org
> > Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> > List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
> > List
> Archives:https://lists.fedoraproject.org/archives/list/[e-mail]...
> > Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
>
>  _______________________________________________
>  devel mailing list --devel(a)lists.fedoraproject.org
>  To unsubscribe send an email todevel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List
> Archives:https://lists.fedoraproject.org/archives/list/[e-mail]...
>  Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure

[Dominik 'Rathann' Mierzejewski]

On Sunday, 10 July 2022 at 19:43, Vitaly Zaitsev via devel wrote:
>  Hello.
>  
>  fmt 9.0.x update will include a soversion bump from .8 to .9. It has both
>  API and ABI changes. 
[...]
>  mkvtoolnix 
mkvtoolnix updated and built in the side tag:
https://koji.fedoraproject.org/koji/taskinfo?taskID=89457215
Coincidentally, upstream added support for fmt 9.0 in this latest
release.

Regards,
Dominik
-- 
Fedora   https://getfedora.org  |  RPM Fusion  http://rpmfusion.org
There should be a science of discontent. People need hard times and
oppression to develop psychic muscles.
        -- from "Collected Sayings of Muad'Dib" by the Princess Irulan

[Vitaly Zaitsev]

On 11/07/2022 18:21, Ben Beasley wrote:
>  I don’t believe this list is nearly complete. Two packages I maintain
>
>  that would be affected (fmidi and giada) are absent from the list. 
It's very strange:

$ dnf repoquery -q  --releasever=rawhide --disablerepo="*" 
--qf="%{name}" --enablerepo=fedora-source --enablerepo=updates-source 
--enablerepo=updates-testing-source --archlist=src 
--whatrequires="fmt-devel"
0ad
OpenImageIO
cantera
ceph
dolphin-emu
domoticz
folly
libsonata
mkvtoolnix
sdrpp
spdlog
vcpkg
zswap-cli

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Vitaly Zaitsev]

On 11/07/2022 18:50, Ben Beasley wrote:
>  If that FTBFS rate is typical of the full set of affected packages,
> then 
>  merging the side tag without resolving the API incompatibilities in 
>  dependent packages will be pretty disruptive to Rawhide as a whole. 
I want to merge everything before the Fedora 37 branching and mass rebuild.

>  Would you consider setting up a COPR to figure out how many packages
> are affected by the API changes, and filing Bugzillas on the affected dependent packages?
I will create FTBFS RHBZ tickets.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Vitaly Zaitsev]

On 12/07/2022 22:13, Ben Beasley wrote:
>  Both bear and fmidi are now rebuilt into the side tag. I defined the
>
>  preprocessor macro FMT_DEPRECATED_OSTREAM as a temporary workaround 
>  (noting that this will not be available in fmt 10.0). 
It was fixed by upstream[1]. I think maintainer should backport the 
upstream patch[2] instead of using FMT_DEPRECATED_OSTREAM definition.

[1]: 
https://github.com/jpcima/fmidi/commit/20916421656e89a1303a85e76e89db8bd5...
[2]: 
https://github.com/jpcima/fmidi/commit/20916421656e89a1303a85e76e89db8bd5...


-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Maxwell G]

On 22/07/14 07:32PM, Vitaly Zaitsev via devel wrote:
>  On 11/07/2022 18:21, Ben Beasley wrote:
>  > I don’t believe this list is nearly complete. Two packages I maintain
>  > that would be affected (fmidi and giada) are absent from the list.
>  
>  It's very strange:
>  
>  $ dnf repoquery -q  --releasever=rawhide --disablerepo="*"
> --qf="%{name}"
>  --enablerepo=fedora-source --enablerepo=updates-source
>  --enablerepo=updates-testing-source --archlist=src
>  --whatrequires="fmt-devel" 
Use the command from
https://lists.fedoraproject.org/archives/list/[e-mail].o....
If you only enable the *-source repositories, it won't include
fmt-devel's virtual Provides.

```
$ diff -u <(echo "0ad
OpenImageIO
cantera
ceph
dolphin-emu
domoticz
folly
libsonata
mkvtoolnix
sdrpp
spdlog
vcpkg
zswap-cli" | sort) <(sudo dnf repoquery --repo=rawhide{,-source} -q --whatrequires
fmt-devel | grep '\.src' | pkgname | sort | uniq)
--- /proc/self/fd/12    2022-07-14 12:51:17.366920423 -0500
+++ /proc/self/fd/13    2022-07-14 12:51:17.366920423 -0500
@@ -1,13 +1,27 @@
 0ad
+bear
 cantera
 ceph
 dolphin-emu
 domoticz
+easyeffects
+easyrpg-player
+fcitx5
+fcitx5-chinese-addons
+fcitx5-gtk
+fcitx5-m17n
+fcitx5-zhuyin
+fmidi
 folly
+giada
+libsemigroups
 libsonata
 mkvtoolnix
+ntfs2btrfs
 OpenImageIO
 sdrpp
 spdlog
 vcpkg
+waybar
 zswap-cli
+zxing-cpp

$ sudo dnf repoquery --repo=rawhide-source -q --requires bear | grep fmt
cmake(fmt)
```

-- 
Thanks,

Maxwell G (@gotmax23)
Pronouns: He/Him/His

[Davide Cavalca]

On Thu, 2022-07-14 at 12:58 -0500, Maxwell G via devel wrote:
>  +ntfs2btrfs
>  
I just rebuilt this one in the sidetag.

Cheers
Davide

[Ben Beasley]

Yes, fmidi added support upstream after I reported the issue (and after I applied the
FMT_DEPRECATED_OSTREAM workaround in Fedora). I adjusted the Fedora package to track
upstream as soon as a patch became available[1].

[1]
https://src.fedoraproject.org/rpms/fmidi/c/5d985a532037b7b366d33c478be4f3...

On Thu, Jul 14, 2022, at 1:39 PM, Vitaly Zaitsev via devel wrote:
> On 12/07/2022 22:13, Ben Beasley wrote:
>> Both bear and fmidi are now rebuilt into the side tag. I defined the 
>> preprocessor macro FMT_DEPRECATED_OSTREAM as a temporary workaround 
>> (noting that this will not be available in fmt 10.0).
>
> It was fixed by upstream[1]. I think maintainer should backport the 
> upstream patch[2] instead of using FMT_DEPRECATED_OSTREAM definition.
>
> [1]: 
> https://github.com/jpcima/fmidi/commit/20916421656e89a1303a85e76e89db8bd5...
> [2]: 
>
https://github.com/jpcima/fmidi/commit/20916421656e89a1303a85e76e89db8bd5...
>
>
> -- 
> Sincerely,
>    Vitaly Zaitsev (vitaly(a)easycoding.org)
> _______________________________________________
> devel mailing list -- devel(a)lists.fedoraproject.org
> To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
> Fedora Code of Conduct: 
> https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives: 
> https://lists.fedoraproject.org/archives/list/[e-mail]
> Do not reply to spam on the list, report it: 
> https://pagure.io/fedora-infrastructure

[Mamoru TASAKA]

Ben Beasley wrote on 2022/07/12 1:50:
>  Two of the three packages I tried to rebuild (bear and fmidi) failed
> with errors like:
>  
>  /usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot format an
> argument. To make type T formattable provide a formatter<T>
> specialization:https://fmt.dev/latest/api.html#udt
>    1733 |       formattable,
>         |       ^~~~~~~~~~~
>  
>  Links to failing scratch builds in the side tag:
>  
>      https://koji.fedoraproject.org/koji/taskinfo?taskID=89380682
>  
>      https://koji.fedoraproject.org/koji/taskinfo?taskID=89380736
>  
>  I found that there was already a bug report open upstream for Bear:
> https://github.com/rizsotto/Bear/issues/471
>  
>  If that FTBFS rate is typical of the full set of affected packages, then merging the side
> tag without resolving the API incompatibilities in dependent packages will be pretty
> disruptive to Rawhide as a whole.
>  
>  Would you consider setting up a COPR to figure out how many packages are affected by the
> API changes, and filing Bugzillas on the affected dependent packages? It might take a
> while for affected package maintainers to become aware of the problem, investigate, and
> perhaps work with upstreams to prepare the necessary patches.
>  
>  Link to fmt 9.0.0 release notes: https://github.com/fmtlib/fmt/releases/tag/9.0.0 

So instead I tried copr build: currently as far as I am correct
currently 29 packages needs rebuilding:

      1	0ad-0.0.25b-2.fc36.src.rpm
      2	cachelib-16^20220314gitbd22b0e-1.fc37.src.rpm
      3	cantera-2.6.0-26.fc37.src.rpm
      4	ceph-17.2.1-4.fc37.src.rpm
      5	coeurl-0.2.0-2.fc37.src.rpm
      6	cryfs-0.11.2-3.fc37.src.rpm
      7	dolphin-emu-5.0.16380-2.fc37.src.rpm
      8	easyeffects-6.2.6-1.fc37.src.rpm
      9	easyrpg-player-0.7.0-2.fc36.src.rpm
     10	fb303-2022.03.14.00-1.fc37.src.rpm
     11	fbthrift-2022.03.14.00-1.fc37.src.rpm
     12	fizz-2022.03.14.00-1.fc37.src.rpm
     13	folly-2022.03.14.00-1.fc37.src.rpm
     14	freeopcua-0-0.23.20200131.da2b76f.fc37.src.rpm
     15	gerbera-1.11.0-1.fc37.src.rpm
     16	gnuradio-3.10.3.0-1.fc37.src.rpm
     17	gr-funcube-1.0.0-14.20220130gitbf71b979.fc37.src.rpm
     18	libsemigroups-2.2.0-1.fc37.src.rpm
     19	luxcorerender-2.6-1.fc37.src.rpm
     20	mangohud-0.6.7.1-2.fc37.src.rpm
     21	mcrouter-0.41.0.20220314-1.fc37.src.rpm
     22	nheko-0.9.3-2.fc37.src.rpm
     23	proxygen-2022.03.14.00-2.fc37.src.rpm
     24	rstudio-2022.07.0+548-1.fc37.src.rpm
     25	sdrpp-1.0.4-6.fc37.src.rpm
     26	vcpkg-2022.06.15-1.fc37.src.rpm
     27	wangle-2022.03.14.00-1.fc37.src.rpm
     28	watchman-2021.05.10.00-13.fc37.src.rpm
     29	waybar-0.9.13-1.fc37.src.rpm

and the first try for rebuilding theese pkgs for fmt-9 is ....
rather sad: only 7 pkgs succeeded.
(Note that I have not checked the reason of rebuild failure
  for each pkg yet.)

https://copr.fedorainfracloud.org/coprs/mtasaka/fmt9-test/builds/

So what should we do? Now the timing is rather critical because
mass rebuild for F37 is planned at 2022-07-20, which will rebuild fmt-9
on rawhide tree anyway, which will cause rawhide tree breakage.

To avoid this, should fmt rawhide branch be reverted for now, or
"noautobuild" file should be added to avoid rebuild for fmt
(if noautobuild file still works)?


Note: some pkgs depends on folly, which cannot be even installed on rawhide,
and FTBFS even for rawhide, and now I am looking at this first.

Regards,
Mamoru

[Vitaly Zaitsev]

On 16/07/2022 06:07, Mamoru TASAKA wrote:
>  So what should we do? Now the timing is rather critical because
>  mass rebuild for F37 is planned at 2022-07-20, which will rebuild fmt-9
>  on rawhide tree anyway, which will cause rawhide tree breakage. 
FTBFS issues will be created.

>  To avoid this, should fmt rawhide branch be reverted for now, or
>  "noautobuild" file should be added to avoid rebuild for fmt
>  (if noautobuild file still works)?  
I will merge the side tag before 2022-07-20 (today or tomorrow). All 
packages will be rebuilt automatically.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Vitaly Zaitsev]

On 16/07/2022 06:07, Mamoru TASAKA wrote:
>  So what should we do? Now the timing is rather critical because
>  mass rebuild for F37 is planned at 2022-07-20, which will rebuild fmt-9
>  on rawhide tree anyway, which will cause rawhide tree breakage. 
Just merged[1] the side tag to Rawhide.

The following packages still needs rebuild:

     0ad
     cachelib
     cantera
     ceph
     cryfs
     dolphin-emu
     easyrpg-player
     fb303
     fbthrift
     fizz
     folly
     freeopcua
     gerbera
     gnuradio
     gr-funcube
     libsemigroups
     luxcorerender
     mangohud
     mcrouter
     proxygen
     rstudio
     sdrpp
     wangle
     watchman
     waybar

Maintainers can build their packages manually for Rawhide, or on 
2022-07-20 mass rebuild  will do it automatically.

FTI and FTBFS issues will be created during mass rebuild.

[1]: https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Richard Shaw]

On Sun, Jul 17, 2022 at 2:32 AM Vitaly Zaitsev via devel <
devel(a)lists.fedoraproject.org&gt; wrote:

>  On 16/07/2022 06:07, Mamoru TASAKA wrote:
>  > So what should we do? Now the timing is rather critical because
>  > mass rebuild for F37 is planned at 2022-07-20, which will rebuild fmt-9
>  > on rawhide tree anyway, which will cause rawhide tree breakage.
>
>  Just merged[1] the side tag to Rawhide.
>
>  The following packages still needs rebuild:
>
>       gnuradio
>       sdrpp 

I'm working on rebuilding gnuradio, sdrpp, and sdrangel (depends on
gnuradio) for a new codec2 release and ran into this issue.

sdrpp has been rebuilt in my side tag but currently gnuradio still fails to
build[1] with the error:

/usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot
format an argument. To make type T formattable provide a formatter<T>
specialization: https://fmt.dev/latest/api.html#udt

Thanks,
Richard

[1] https://kojipkgs.fedoraproject.org//work/tasks/8546/89608546/build.log

On 7/17/22 3:54 PM, Richard Shaw <hobbes1069(a)gmail.com&gt; wrote:
>  On Sun, Jul 17, 2022 at 2:32 AM Vitaly Zaitsev via devel
> <devel(a)lists.fedoraproject.org 
>  <mailto:[e-mail]>> wrote:
>  
>      On 16/07/2022 06:07, Mamoru TASAKA wrote:
>       > So what should we do? Now the timing is rather critical because
>       > mass rebuild for F37 is planned at 2022-07-20, which will rebuild fmt-9
>       > on rawhide tree anyway, which will cause rawhide tree breakage.
>  
>      Just merged[1] the side tag to Rawhide.
>  
>      The following packages still needs rebuild:
>  
>            gnuradio
>            sdrpp
>  
>  
>  I'm working on rebuilding gnuradio, sdrpp, and sdrangel (depends on gnuradio) for a
> new 
>  codec2 release and ran into this issue.
>  
>  sdrpp has been rebuilt in my side tag but currently gnuradio still fails to build[1] with
>
>  the error:
>  
>  /usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot format an
> argument. 
>  To make type T formattable provide a formatter<T> specialization: 
>  https://fmt.dev/latest/api.html#udt <https://fmt.dev/latest/api.html#udt>
>  
>  Thanks,
>  Richard
>   
Same for ceph. I tried unsucessfully to rebuild it. The bug is not yet upstream,
only in Debian yet https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1014549

[Vitaly Zaitsev]

On 17/07/2022 15:54, Richard Shaw wrote:
>  /usr/include/fmt/core.h:1733:7: error: static assertion failed:
> Cannot 
>  format an argument. To make type T formattable provide a formatter<T> 
>  specialization: https://fmt.dev/latest/api.html#udt 
>  <https://fmt.dev/latest/api.html#udt> 
Can you try the following workaround:

%global optflags %(echo %{optflags} -DFMT_DEPRECATED_OSTREAM)

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Richard Shaw]

On Sun, Jul 17, 2022 at 3:18 PM Vitaly Zaitsev via devel <
devel(a)lists.fedoraproject.org&gt; wrote:

>  On 17/07/2022 15:54, Richard Shaw wrote:
>  > /usr/include/fmt/core.h:1733:7: error: static assertion failed: Cannot
>  > format an argument. To make type T formattable provide a formatter<T>
>  > specialization: https://fmt.dev/latest/api.html#udt
>  > <https://fmt.dev/latest/api.html#udt>
>
>  Can you try the following workaround:
>
>  %global optflags %(echo %{optflags} -DFMT_DEPRECATED_OSTREAM)
>  
That worked for now. What's the long term fix? All upstreams need to update
to the new API?

Thanks,
Richard

[Vitaly Zaitsev]

On 18/07/2022 18:18, Richard Shaw wrote:
>  That worked for now. What's the long term fix? 
It will work until the next major release - 10.0.0.

>  All upstreams need to update to the new API? 
Yes.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Vitaly Zaitsev]

On 17/07/2022 21:55, zebob.m(a)gmail.com wrote:
>  Same for ceph. I tried unsucessfully to rebuild it. 
Can you try this workaround:

%global optflags %(echo %{optflags} -DFMT_DEPRECATED_OSTREAM)

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Kaleb Keithley]

On Sun, Jul 17, 2022 at 3:56 PM <zebob.m(a)gmail.com&gt; wrote:

>  Same for ceph. I tried unsucessfully to rebuild it. The bug is not yet
>  upstream, only in Debian yet
>  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1014549
>  
Do you have a link to the failed build?

-- 

Kaleb

[Vitaly Zaitsev]

On 18/07/2022 19:12, Kaleb Keithley wrote:
>  Do you have a link to the failed build? 
https://koji.fedoraproject.org/koji/taskinfo?taskID=89606207

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Kaleb Keithley]

On Sun, Jul 17, 2022 at 3:56 PM <zebob.m(a)gmail.com&gt; wrote:

>  Same for ceph. I tried unsucessfully to rebuild it. The bug is not yet
>  upstream, only in Debian yet
>  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1014549 

I opened https://tracker.ceph.com/issues/56610

-- 

Kaleb

[Adam Williamson]

On Mon, 2022-07-18 at 13:47 -0400, Kaleb Keithley wrote:
>  On Sun, Jul 17, 2022 at 3:56 PM <zebob.m(a)gmail.com&gt; wrote:
>  
>  > 
>  > Same for ceph. I tried unsucessfully to rebuild it. The bug is not yet
>  > upstream, only in Debian yet
>  > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1014549
>  
>  
>  I opened https://tracker.ceph.com/issues/56610 
This is a problem, because it breaks the Rawhide compose: gnome-boxes
is part of Workstation, it requires libvirt-daemon-kvm, that requires
libvirt-daemon-driver-storage, that requires libvirt-daemon-driver-
storage-rbd, and that requires librbd1 which is part of ceph.

We could temporarily drop Boxes from Workstation, I guess, especially
since it doesn't work right now due to libsoup hilarity[0]. But we
should really have a proper fix ASAP.
-- 
Adam Williamson
Fedora QA
IRC: adamw | Twitter: adamw_ha
https://www.happyassassin.net

[Adam Williamson]

On Mon, 2022-07-18 at 12:05 -0700, Adam Williamson wrote:
>  On Mon, 2022-07-18 at 13:47 -0400, Kaleb Keithley wrote:
>  > On Sun, Jul 17, 2022 at 3:56 PM <zebob.m(a)gmail.com&gt; wrote:
>  > 
>  > > 
>  > > Same for ceph. I tried unsucessfully to rebuild it. The bug is not yet
>  > > upstream, only in Debian yet
>  > > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1014549
>  > 
>  > 
>  > I opened https://tracker.ceph.com/issues/56610
>  
>  This is a problem, because it breaks the Rawhide compose: gnome-boxes
>  is part of Workstation, it requires libvirt-daemon-kvm, that requires
>  libvirt-daemon-driver-storage, that requires libvirt-daemon-driver-
>  storage-rbd, and that requires librbd1 which is part of ceph.
>  
>  We could temporarily drop Boxes from Workstation, I guess, especially
>  since it doesn't work right now due to libsoup hilarity[0]. But we
>  should really have a proper fix ASAP. 
Sorry, that libsoup hilarity[0] reference:

[0]: https://bugzilla.redhat.com/show_bug.cgi?id=2105116
-- 
Adam Williamson
Fedora QA
IRC: adamw | Twitter: adamw_ha
https://www.happyassassin.net

[Vitaly Zaitsev]

On 18/07/2022 21:05, Adam Williamson wrote:
>  We could temporarily drop Boxes from Workstation, I guess,
> especially
>  since it doesn't work right now due to libsoup hilarity[0]. But we
>  should really have a proper fix ASAP. 
Fixed: https://src.fedoraproject.org/rpms/ceph/pull-request/5

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Adam Williamson]

On Mon, 2022-07-18 at 22:15 +0200, Vitaly Zaitsev via devel wrote:
>  On 18/07/2022 21:05, Adam Williamson wrote:
>  > We could temporarily drop Boxes from Workstation, I guess, especially
>  > since it doesn't work right now due to libsoup hilarity[0]. But we
>  > should really have a proper fix ASAP.
>  
>  Fixed: https://src.fedoraproject.org/rpms/ceph/pull-request/5 
Thanks! But apparently not completely fixed yet:
https://koji.fedoraproject.org/koji/taskinfo?taskID=89669510

Looks like it's failing on some warnings as it's being built with -
Wall...
-- 
Adam Williamson
Fedora QA
IRC: adamw | Twitter: adamw_ha
https://www.happyassassin.net

[Vitaly Zaitsev]

On 18/07/2022 23:28, Adam Williamson wrote:
>  Looks like it's failing on some warnings as it's being built
> with -
>  Wall... 
Another issue:

/builddir/build/BUILD/ceph-17.2.1/src/mon/LogMonitor.cc:418:21: error: 
no matching function for call to 'format_to(fmt::v9::memory_buffer&, 
const char [4], const LogEntry&)'
   418 |       fmt::format_to(file_log_buffer, "{}\n", le);
       |       ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from /usr/include/fmt/format.h:48,
                  from 
/builddir/build/BUILD/ceph-17.2.1/src/mon/LogMonitor.h:22:
/usr/include/fmt/core.h:3190:17: note: candidate: 'template<class 
OutputIt, class ... T, typename 
std::enable_if<fmt::v9::detail::is_output_iterator<OutputIt, 
char>::value, int>::type <anonymous> > OutputIt 
fmt::v9::format_to(OutputIt, format_string<T ...>, T&& ...)'
  3190 | FMT_INLINE auto format_to(OutputIt out, format_string<T...> 
fmt, T&&... args)
       |                 ^~~~~~~~~
/usr/include/fmt/core.h:3190:17: note:   template argument 
deduction/substitution failed:
/usr/include/fmt/core.h:3189:11: error: no type named 'type' in 'struct 
std::enable_if<false, int>'
  3189 |           FMT_ENABLE_IF(detail::is_output_iterator<OutputIt, 
char>::value)>
       |           ^~~~~~~~~~~~~

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Miro Hrončok]

On 10. 07. 22 19:43, Vitaly Zaitsev via devel wrote:
>  Hello.
>  
>  fmt 9.0.x update will include a soversion bump from .8 to .9. It has both API 
>  and ABI changes.
>  
>  Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
>  
>  The list of affected packages in Rawhide:
>  0ad
>  OpenImageIO
>  cantera
>  ceph
>  dolphin-emu
>  domoticz
>  folly
>  libsonata
>  mkvtoolnix
>  sdrpp
>  spdlog
>  vcpkg
>  zswap-cli
>  
>  I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will need 
>  proven-packagers assistance.
>  
>  Please use the f37-build-side-54844 side tag. I will merge it into Rawhide later. 
I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61 was pushed 
with just a handful of builds, while many others are broken. What happened here?

cachelib
https://bugzilla.redhat.com/show_bug.cgi?id=2108349

cantera
https://bugzilla.redhat.com/show_bug.cgi?id=2108350

ceph
https://bugzilla.redhat.com/show_bug.cgi?id=2108351

dolphin
https://bugzilla.redhat.com/show_bug.cgi?id=2108353

easyrpg-player
https://bugzilla.redhat.com/show_bug.cgi?id=2108354

fb303
https://bugzilla.redhat.com/show_bug.cgi?id=2108355

fizz
https://bugzilla.redhat.com/show_bug.cgi?id=2108356

freeopcua
https://bugzilla.redhat.com/show_bug.cgi?id=2108357

gerbera
https://bugzilla.redhat.com/show_bug.cgi?id=2108358

mangohud
https://bugzilla.redhat.com/show_bug.cgi?id=2108361

proxygen
https://bugzilla.redhat.com/show_bug.cgi?id=2108362

rstudio
https://bugzilla.redhat.com/show_bug.cgi?id=2108367

sdrpp
https://bugzilla.redhat.com/show_bug.cgi?id=2108368

wangle
https://bugzilla.redhat.com/show_bug.cgi?id=2108369

-- 
Miro Hrončok
-- 
Phone: [phone]
IRC: mhroncok

[Miro Hrončok]

On 19. 07. 22 1:28, Miro Hrončok wrote:
>  dolphin
>  https://bugzilla.redhat.com/show_bug.cgi?id=2108353 
Sorry, dolphin-emu.

-- 
Miro Hrončok
-- 
Phone: [phone]
IRC: mhroncok

[Richard Shaw]

On Mon, Jul 18, 2022 at 6:28 PM Miro Hrončok <mhroncok(a)redhat.com&gt; wrote:

>  sdrpp
>  https://bugzilla.redhat.com/show_bug.cgi?id=2108368 

I closed this one. It was "fixed" when I submitted my side tag for codec2
soname bump.

Thanks,
Richard

[Vitaly Zaitsev]

On 19/07/2022 01:28, Miro Hrončok wrote:
>  I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61
> was 
>  pushed with just a handful of builds, while many others are broken. What 
>  happened here? 
Maintainers had 1 week to rebuild all their packages. Most did nothing. 
Blame them, not me.

I did my best: announced it in ML, waited 1 week, asked for the 
provenpackager assistance multiple times both in ML and IRC/Matrix. No 
one volunteered to help me.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Iñaki Ucar]

On Tue, 19 Jul 2022 at 11:38, Vitaly Zaitsev via devel
<devel(a)lists.fedoraproject.org&gt; wrote:
>  On 19/07/2022 01:28, Miro Hrončok wrote:
>  > I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61 was
>  > pushed with just a handful of builds, while many others are broken. What
>  > happened here?
>
>  Maintainers had 1 week to rebuild all their packages. Most did nothing.
>  Blame them, not me.
>
>  I did my best: announced it in ML, waited 1 week, asked for the
>  provenpackager assistance multiple times both in ML and IRC/Matrix. No
>  one volunteered to help me. 
As a suggestion for the next time: if you are not going to or cannot
rebuild the dependent packages, please add the maintainers to BCC. I
simply cannot read every single thread in the mailing list, nor recall
every single dependency of my packages.

-- 
Iñaki Úcar

[Vitaly Zaitsev]

On 16/07/2022 06:07, Mamoru TASAKA wrote:
>  So what should we do? Now the timing is rather critical because
>  mass rebuild for F37 is planned at 2022-07-20, which will rebuild fmt-9
>  on rawhide tree anyway, which will cause rawhide tree breakage. 
I'm very sorry for this breakage.

I've checked 6/10 of failed builds. They can't be fixed easily due to 
removed deprecated API usage. They should be fixed by upstreams, but it 
will take time.

Now we have 2 options:
1. We can postpone fmt 9.x release to Fedora 38 and restore fmt 8.1.1 
with an Epoch bump.
2. We can introduce a new fmt8 compatibility package.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Mamoru TASAKA]

Vitaly Zaitsev via devel wrote on 2022/07/19 18:31:
>  On 19/07/2022 01:28, Miro Hrončok wrote:
> > I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61 was pushed with
> just a handful of builds, while many others are broken. What happened here?
>  
>  Maintainers had 1 week to rebuild all their packages. Most did nothing. Blame them, not
> me.
>  
>  I did my best: announced it in ML, waited 1 week, asked for the provenpackager assistance
> multiple times both in ML and IRC/Matrix. No one volunteered to help me.
>   
So far I am doing my best to fix this breakage. Now I've fixed 5 of these.
Currently as far as I am correct 15 pkgs are still broken - I am slowly looking at these.

Mamoru

[Richard Shaw]

On Tue, Jul 19, 2022 at 4:32 AM Vitaly Zaitsev via devel <
devel(a)lists.fedoraproject.org&gt; wrote:

>  On 19/07/2022 01:28, Miro Hrončok wrote:
>  > I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61
>  was
>  > pushed with just a handful of builds, while many others are broken. What
>  > happened here?
>
>  Maintainers had 1 week to rebuild all their packages. Most did nothing.
>  Blame them, not me.
>
>  I did my best: announced it in ML, waited 1 week, asked for the
>  provenpackager assistance multiple times both in ML and IRC/Matrix. No
>  one volunteered to help me.
>  
Unfortunately it's very difficult to "coordinate" things like this with a
worldwide community. Someone / somewhere is always "holiday" or "summer
vacation", or in my case last week was my son's summer camp. In reality, a
week really isn't enough time but we can't have stuff like this going on
for months either so if you maintain a package that has a large number of
dependencies it may be time to work on becoming a proven packager so you
can at least rebuild other packages or apply "easy" fixes.

If the packages are broken for other reasons, then you can use direct
emails, <package&gt;-maintainers(a)fedoraproject.org, or file BZ tickets.

Thanks,
Richard

[Vitaly Zaitsev]

On 19/07/2022 13:53, Mamoru TASAKA wrote:
>  So far I am doing my best to fix this breakage. Now I've fixed 5
> of these. 
Thank you.

>  Currently as far as I am correct 15 pkgs are still broken - I am
> slowly 
>  looking at these. 
We are working on creating fmt8{,-devel} to mitigate all other breakages 
due to the major API changes in fmt.

Broken packages can be fixed easily with BuildRequires: fmt8-devel 
instead of BuildRequires: fmt-devel.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Vitaly Zaitsev]

On 19/07/2022 14:04, Richard Shaw wrote:
>  Unfortunately it's very difficult to "coordinate"
> things like this with 
>  a worldwide community. 
True. Nearly impossible without provenpackager rights, I'd say.

>  In reality, a week really isn't enough time but we can't have
> stuff like this going on for months either so if you maintain a package that has a large
> number of dependencies 
It was my fault that I didn't provide a compatibility package due 
massive breaking changes in fmt's API.

Now it can be easily mitigated by using fmt8-devel.

Rawhide compose will be fixed automatically soon, as fmt8 provides 
missing libfmt.so.8 shared library.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Kalev Lember]

On Tue, Jul 19, 2022 at 2:23 PM Vitaly Zaitsev via devel <
devel(a)lists.fedoraproject.org&gt; wrote:

>  On 19/07/2022 14:04, Richard Shaw wrote:
>  > Unfortunately it's very difficult to "coordinate" things like this
> with
>  > a worldwide community.
>
>  True. Nearly impossible without provenpackager rights, I'd say.
>
>  > In reality, a week really isn't enough time but we can't have stuff like
>  this going on for months either so if you maintain a package that has a
>  large number of dependencies
>
>  It was my fault that I didn't provide a compatibility package due
>  massive breaking changes in fmt's API.
>
>  Now it can be easily mitigated by using fmt8-devel.
>
>  Rawhide compose will be fixed automatically soon, as fmt8 provides
>  missing libfmt.so.8 shared library.
>  
Adding the fmt8 package sounds like a good plan to me. Thanks!

If you do these kinds of changes in the future, I think it makes sense to
prioritize getting rebuilds done for packages that are included on Fedora's
release blocking images (Workstation, KDE, etc). It's more OK to push out a
change that breaks an odd leaf package somewhere, but what's most important
is that rawhide composes keep happening and for that we need the package
set that's on release blocking images to not have dependency issues.

It sounds like you really need provenpackager access as well to be able to
drive fmt changes in the future. I see you've already applied so it looks
like you are a step ahead here. :)

Keep up the good work!

-- 
Kalev

[Maxwell G]

On 22/07/19 02:22PM, Vitaly Zaitsev via devel wrote:
>  Now it can be easily mitigated by using fmt8-devel.
>  
>  Rawhide compose will be fixed automatically soon, as fmt8 provides missing
>  libfmt.so.8 shared library. 
Yes, the *compose*, but those packages will still need to be fixed to
not FTBFS.

-- 
Thanks,

Maxwell G (@gotmax23)
Pronouns: He/Him/His

[Mamoru TASAKA]

Vitaly Zaitsev via devel wrote on 2022/07/11 2:43:
>  Hello.
>  
>  fmt 9.0.x update will include a soversion bump from .8 to .9. It has both API and ABI
> changes.
>  
>  Changelog: https://github.com/fmtlib/fmt/releases/tag/9.0.0
>  
>  The list of affected packages in Rawhide:
>  0ad
>  OpenImageIO
>  cantera
>  ceph
>  dolphin-emu
>  domoticz
>  folly
>  libsonata
>  mkvtoolnix
>  sdrpp
>  spdlog
>  vcpkg
>  zswap-cli
>  
>  I can rebuild only spdlog, vcpkg and zswap-cli. For others, I will need proven-packagers
> assistance.
>  
>  Please use the f37-build-side-54844 side tag. I will merge it into Rawhide later.
>   
So current status update for fmt-9 change on rawhide tree:

33 pkgs are now using fmt-9 (built successfully with some modification) on rawhide tree
(cryfs is still building, however it will succeed as scratch build was already okay).

Only 1 package still uses fmt8 - 0ad .
0ad FTBFS on f37 due to different issue from fmt change - scratch build for F-37 shows
virtualenv related
issue - perhaps due to python3.11 changes, and scratch build for F-36 shows some rust(?)
related error,
which is beyond my knowledge currently.
Note that rawhide 0ad currently linked against boost1.76 - while rawhide boost is already
1.78,
so 0ad is FTI even if fmt8 package is introduced anyway currently.

Regards,
Mamoru

[Demi Marie Obenour]

On 7/19/22 08:04, Richard Shaw wrote:
>  On Tue, Jul 19, 2022 at 4:32 AM Vitaly Zaitsev via devel <
>  devel(a)lists.fedoraproject.org&gt; wrote:
>  
> > On 19/07/2022 01:28, Miro Hrončok wrote:
> >> I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61
> > was
> >> pushed with just a handful of builds, while many others are broken. What
> >> happened here?
> >
> > Maintainers had 1 week to rebuild all their packages. Most did nothing.
> > Blame them, not me.
> >
> > I did my best: announced it in ML, waited 1 week, asked for the
> > provenpackager assistance multiple times both in ML and IRC/Matrix. No
> > one volunteered to help me.
> >
>  
>  Unfortunately it's very difficult to "coordinate" things like this with a
>  worldwide community. Someone / somewhere is always "holiday" or "summer
>  vacation", or in my case last week was my son's summer camp. In reality, a
>  week really isn't enough time but we can't have stuff like this going on
>  for months either so if you maintain a package that has a large number of
>  dependencies it may be time to work on becoming a proven packager so you
>  can at least rebuild other packages or apply "easy" fixes. 
I wonder if Fedora should have a way of saying “this release breaks
ABI, rebuild all dependent packages automatically”.  This would allow
dynamic linking of Rust and Haskell programs, which have no stable
ABI at all.  Bodhi or koji would rebuild all dependent packages and
push them out (following the usual testing process) without the need
for any intervention by those packages’ maintainers.
-- 
Sincerely,
Demi Marie Obenour (she/her/hers)

[Fabio Valentini]

On Wed, Jul 20, 2022 at 5:38 PM Demi Marie Obenour
<demiobenour(a)gmail.com&gt; wrote:
>  On 7/19/22 08:04, Richard Shaw wrote:
>  > On Tue, Jul 19, 2022 at 4:32 AM Vitaly Zaitsev via devel <
>  > devel(a)lists.fedoraproject.org&gt; wrote:
>  >
>  >> On 19/07/2022 01:28, Miro Hrončok wrote:
>  >>> I see https://bodhi.fedoraproject.org/updates/FEDORA-2022-ee990d1f61
>  >> was
>  >>> pushed with just a handful of builds, while many others are broken. What
>  >>> happened here?
>  >>
>  >> Maintainers had 1 week to rebuild all their packages. Most did nothing.
>  >> Blame them, not me.
>  >>
>  >> I did my best: announced it in ML, waited 1 week, asked for the
>  >> provenpackager assistance multiple times both in ML and IRC/Matrix. No
>  >> one volunteered to help me.
>  >>
>  >
>  > Unfortunately it's very difficult to "coordinate" things like this
> with a
>  > worldwide community. Someone / somewhere is always "holiday" or
> "summer
>  > vacation", or in my case last week was my son's summer camp. In reality, a
>  > week really isn't enough time but we can't have stuff like this going on
>  > for months either so if you maintain a package that has a large number of
>  > dependencies it may be time to work on becoming a proven packager so you
>  > can at least rebuild other packages or apply "easy" fixes.
>
>  I wonder if Fedora should have a way of saying “this release breaks
>  ABI, rebuild all dependent packages automatically”.  This would allow
>  dynamic linking of Rust and Haskell programs, which have no stable
>  ABI at all.  Bodhi or koji would rebuild all dependent packages and
>  push them out (following the usual testing process) without the need
>  for any intervention by those packages’ maintainers. 
Speaking for Rust: This might seem like a good idea at first, but no,
this won't work at all.
1) Building a shared Rust library ("dylib" crate type, as opposed to
the "cdylib" crate type, which can be used to produce libraries with a
C ABI) is an opt-in feature *at the crate / library level* that I have
never seen used in our 2000+ Rust packages in Fedora. So it's
basically completely untested.
2) Our packaging tools don't support this use case at all. Built Rust
packages are currently noarch and source-only.
3) Rust crates can be built with different feature sets or options
enabled (i.e. behaviour that's controlled by conditional compilation,
not by runtime switches). We'd need to provide a shared library for
every combination of enabled / disabled features, and for crates with
a dozen features, that would mean building and shipping 2^12 (i.e.
4096) different libraries. For a crate with 1491 features (yes, we
have one of those), that would mean building 2 ^ 1491 slightly
different shared libraries (~ 10^450). That's bigger than "number of
particles in the known universe" type stuff.
4) I'm pushing about a dozen updates for various Rust crates to Fedora
*almost every day*. You'd need to rebuild binaries for all those
library changes *and* any Rust compiler changes. Rebuilding all Rust
applications that are packaged for Fedora multiple times a day just
for those dependency changes would cripple Fedora build
infrastructure.

It *might* work for C/C++ libraries in *some* cases (i.e. where
there's only ABI breaks but no API changes).
So in this case, automatic rebuilds wouldn't have helped either,
because of non-trivial-to-fix breaking API changes with fmt 9.0.0.

Fabio

[Vitaly Zaitsev]

On 20/07/2022 16:22, Mamoru TASAKA wrote:
>  33 pkgs are now using fmt-9 (built successfully with some
> modification) 
>  on rawhide tree
>  Only 1 package still uses fmt8 - 0ad . 
I think I should retire fmt8 before the F37 is branched.

-- 
Sincerely,
   Vitaly Zaitsev (vitaly(a)easycoding.org)

[Frantisek Zatloukal]

On Wed, Jul 20, 2022 at 4:24 PM Mamoru TASAKA <mtasaka(a)fedoraproject.org&gt;
wrote:

>  Vitaly Zaitsev via devel wrote on 2022/07/11 2:43:
>  0ad FTBFS on f37 due to different issue from fmt change - scratch build
>  for F-37 shows virtualenv related
>  issue - perhaps due to python3.11 changes, and scratch build for F-36
>  shows some rust(?) related error,
>  which is beyond my knowledge currently.
>  
The python 3.11 compat should work with
https://src.fedoraproject.org/rpms/mozjs78/blob/rawhide/f/0001-Python-Bui...
which I'd made for mozjs78 which is what 0ad is currently using.


>  Note that rawhide 0ad currently linked against boost1.76 - while
> rawhide
>  boost is already 1.78,
>  so 0ad is FTI even if fmt8 package is introduced anyway currently.
>
>  Regards,
>  Mamoru
>  _______________________________________________
>  devel mailing list -- devel(a)lists.fedoraproject.org
>  To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct:
>  https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List Archives:
>  https://lists.fedoraproject.org/archives/list/[e-mail]
>  Do not reply to spam on the list, report it:
>  https://pagure.io/fedora-infrastructure
>  

-- 

Best regards / S pozdravem,

František Zatloukal
Senior Quality Engineer
Red Hat