Looks like opencpn pkgs were switched from wxgtk3.1 --> wxgtk3.0 in bug29309, so they're mistakenly rebuilt.

Updated aom-2.0.1-3.5.mga8 is handled in bug 29808.

Updated kicad-6.0.0-1.mga8 and other kicad pkgs are handled in bug 28352.

Updated 0ad-0.0.25b-1.mga8 was pushed earlier (2022-01-02), but I can't find any bug report about the update.

SRPMS:
aegisub-3.2.2-14.git20191006.9.2.mga8
ariamaestosa-1.4.13-11.1.mga8
asc-2.6.1.0-20.1.mga8
audacity-2.4.2-9.1.mga8
boinc-client-7.16.14-2.1.mga8
codeblocks-20.03-4.1.mga8
codelite-14.0-4.1.mga8
cwstudio-0.9.6-8.1.mga8
diff-pdf-0.4.1-4.1.mga8
dvdstyler-3.1.2-3.1.mga8
erlang-23.2.1-3.1.mga8
flamerobin-0.9.3.1-13.1.mga8
freedink-dfarc-3.14-8.1.mga8
freedv-1.4-4.1.mga8
freefilesync-11.3-2.1.mga8
fskbsetting-0.3.5-6.1.mga8
gnudl-1.0.0-1.3.mga8
gnuplot-5.4.1-3.1.mga8
guayadeque-0.4.6-7.git20201222.2.1.mga8
mediainfo-20.09-2.1.mga8
openbabel-2.4.1-14.1.mga8
openmsx-0.15.0-9.1.mga8
openyahtzee-1.9.3-10.1.mga8
playonlinux-4.4-1.1.mga8
plplot-5.15.0-18.1.mga8
python-wxpython4-4.1.1-1.1.mga
radiotray-ng-0.2.7-12.1.mga8
scorched3d-44-14.1.mga8
slade-3.2.0-0.b2.3.1.mga8
spek-0.8.3-15.1.mga8
tintii-2.10.0-12.1.mga8
trustedqsl-2.5.7-2.1.mga8
urbanlightscape-1.4.0-12.1.mga8
vbam-2.1.4-1.1.mga8
veracrypt-1.24u7-7.1.mga8
woeusb-3.3.1-4.1.mga8
wxformbuilder-3.9.0-1.git20201102.2.1.mga8
wxgtk-3.1.5-1.mga8
wxhexeditor-0.24-9.1.mga8
wxmaxima-20.12.1-1.1.mga8
wxsqlite3-4.6.4-1.1.mga8
wxsvg-1.5.22-3.1.mga8
xchm-1.30-5.1.mga8

Updated python-wxpython4-4.1.1-1.1.mga8 is handled in bug 29291.

(In reply to Jani Välimaa from comment #2)
> Looks like opencpn pkgs were switched from wxgtk3.1 --> wxgtk3.0 in
> bug29309, so they're mistakenly rebuilt.

Hi Jani
You are right 
but never mind they won't be modified by this attempt of rebuild :
I modified the spec files so that they remain strictly built upon wxgtk 3.0
and not with a non specified version of wxgtk : that had induced crashes when they had been rebuilt with the prerelease of wxgtk 3.1

The next version of opencpn (2.6.0) will be compatible with wxgtk 3.1.5 but it needs still some work...

Regards

Besides this I will test guayadeque ... 
but not on my old i586 computer ;-) the Mageia's Build Flags are not good for this :
https://bugs.mageia.org/show_bug.cgi?id=29742#c33

CC: (none) => philippedidier

For other programs crashing the BuildRequire must be rewritten to explicitely require the last version of wxgtk working inside Mageia8 !!!

Inside Cauldron there will be some advises to send upstream for some programs (unless their developers have already modified their code to comply with wxgtk 3.1.5)

Better to open a separate bug about guayadeque.

Please also update all wxgtk libs from core/udpates_testing before trying it as it's running fine on mga8 virtualbox and cauldron real HW.

Thank for all your work

Shouldn't the updated applications require the updated wxgtk?

When I select i.e the updated freefilesync in drakrpm, and only have the old wxgtk, it do neither protest nor suggest the new wxgtk.

CC: (none) => fri

(In reply to Morgan Leijström from comment #14)
> Shouldn't the updated applications require the updated wxgtk?

No, they just require a (any) version of the wxgtk packages.
 
> When I select i.e the updated freefilesync in drakrpm, and only have the old
> wxgtk, it do neither protest nor suggest the new wxgtk.

When/if they are moved to the updates repo they will be installed as a normal
part of updates.

When testing the updates, the updated wxgtk packages should be manually selected
too.

CC: (none) => davidwhodgins

Hi Jani
I updated all the wxgtk packages to final release 
from Mageia8-Core-Updates-testing

Then tested the only programs that I use and Know

1) - updated audacity OK
except this warning when launche from a console :
(Audacity:16827): Gtk-WARNING **: 12:48:12.909: gtk_disable_setlocale() must be called before gtk_init()
(Audacity:16827): Gtk-WARNING **: 12:48:12.911: A floating object was finalized. This means that someone
called g_object_unref() on an object that had only a floating
reference; the initial floating reference is not owned by anyone
and must be removed with g_object_ref_sink().

NB that doesn't prevent it to work

2) - updated guayadeque OK too
There's only a warning when launched from console :
(guayadeque:14372): Gtk-CRITICAL **: 12:29:37.428: gtk_widget_set_size_request: assertion 'height >= -1' failed

NB that doesn't prevent it to work

3) - updated freefilesync OK

NB freefilesync informs that there is a new version 11.16 it may be a good occasion to update it to this new version when rebuilding it to wxgtk 3.1.5 final


4) BUT
- updated dvdstyler crashes (core dumped) with wxgtk 3.1.5 final when it worked perfectly with the pre release version of wxgtk

(In reply to Jani Välimaa from comment #13)
> Better to open a separate bug about guayadeque.
> 
> Please also update all wxgtk libs from core/udpates_testing before trying it
> as it's running fine on mga8 virtualbox and cauldron real HW.

Sorry I didn't do that when I wrote this comment
https://bugs.mageia.org/show_bug.cgi?id=29848#c11  which is useless now !

(In reply to Dave Hodgins from comment #15)
> (In reply to Morgan Leijström from comment #14)
> > Shouldn't the updated applications require the updated wxgtk?
> 
> No, they just require a (any) version of the wxgtk packages.

Practical test disagree:

Kicad 5.1.9-2 works *only* with old wxgtk.

boincmgr, freefilesync, urbanlightscape works *only* with the updated.

Kicad 6.0.0-1 works with new wxgtk, did not test it on old.

I have not tested more, yet.

Running on 64 bit plasma.

All tests in previous comment, i see, was with
python3-wxpython4-4.1.1-1.1.mga8.x86_64
pulled during install of Kicad 5.1.9-2

(In reply to Philippe Didier from comment #16)
> 4) BUT
> - updated dvdstyler crashes (core dumped) with wxgtk 3.1.5 final when it
> worked perfectly with the pre release version of wxgtk
dvdstyler needs also updated wxsvg. dvdstyler starts with updated wxsvg, but emits a msg:
dvdstyler: Symbol `_ZTV9wxSVGCtrl' has different size in shared object, consider re-linking.

I'll push another rebuild of dvdstyler with rebuilt wxsvg.

All SRPMS and associated RPMS. Updated dvdstyler and added wxgtk rpms. Comment 10 is now obsolete.

SRPM:
wxgtk-3.1.5-1.mga8
RPMS:
wxgtk3.1-3.1.5-1.mga8
lib(64)wx_baseu3.1_5-3.1.5-1.mga8
lib(64)wx_baseu_net3.1_5-3.1.5-1.mga8
lib(64)wx_baseu_xml3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_adv3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_aui3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_core3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_gl3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_html3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_media3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_propgrid3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_qa3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_ribbon3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_richtext3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_stc3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_xrc3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_webview3.1_5-3.1.5-1.mga8
lib(64)wxgtku3.1-devel-3.1.5-1.mga8

SRPM:
aegisub-3.2.2-14.git20191006.9.2.mga8
RPMS:
aegisub-3.2.2-14.git20191006.9.2.mga8

SRPM:
ariamaestosa-1.4.13-11.1.mga8
RPMS:
ariamaestosa-1.4.13-11.1.mga8

SRPM:
asc-2.6.1.0-20.1.mga8
RPMS:
asc-data-2.6.1.0-20.1.mga8
asc-2.6.1.0-20.1.mga8


SRPM:
boinc-client-7.16.14-2.1.mga8
RPMS:
boinc-client-static-7.16.14-2.1.mga8
boinc-manager-7.16.14-2.1.mga8
boinc-client-7.16.14-2.1.mga8
boinc-client-devel-7.16.14-2.1.mga8

SRPM:
cwstudio-0.9.6-8.1.mga8
RPMS:
cwstudio-0.9.6-8.1.mga8

SRPM:
diff-pdf-0.4.1-4.1.mga8
RPMS:
diff-pdf-0.4.1-4.1.mga8

SRPM:
audacity-2.4.2-9.1.mga8
RPMS:
audacity-2.4.2-9.1.mga8

SRPM:
dvdstyler-3.1.2-3.2.mga8
RPMS:
dvdstyler-3.1.2-3.2.mga8

SRPM:
flamerobin-0.9.3.1-13.1.mga8
RPMS:
flamerobin-0.9.3.1-13.1.mga8

SRPM:
freedv-1.4-4.1.mga8
RPMS:
freedv-1.4-4.1.mga8

SRPM:
freedink-dfarc-3.14-8.1.mga8
RPMS:
freedink-dfarc-3.14-8.1.mga8

SRPM:
codeblocks-20.03-4.1.mga8
RPMS:
lib(64)codeblocks-devel-20.03-4.1.mga8
codeblocks-20.03-4.1.mga8
lib(64)codeblocks0-20.03-4.1.mga8

SRPM:
fskbsetting-0.3.5-6.1.mga8
RPMS:
fskbsetting-0.3.5-6.1.mga8

SRPM:
freefilesync-11.3-2.1.mga8
RPMS:
freefilesync-11.3-2.1.mga8

SRPM:
codelite-14.0-4.1.mga8
RPMS:
codelite-14.0-4.1.mga8

SRPM:
gnudl-1.0.0-1.3.mga8
RPMS:
gnudl-1.0.0-1.3.mga8

SRPM:
erlang-23.2.1-3.1.mga8
RPMS:
erlang-23.2.1-3.1.mga8
erlang-stdlib-23.2.1-3.1.mga8
erlang-parsetools-23.2.1-3.1.mga8
erlang-debugger-23.2.1-3.1.mga8
erlang-tools-23.2.1-3.1.mga8
erlang-examples-23.2.1-3.1.mga8
erlang-odbc-23.2.1-3.1.mga8
erlang-common_test-23.2.1-3.1.mga8
erlang-sasl-23.2.1-3.1.mga8
erlang-tftp-23.2.1-3.1.mga8
erlang-ssl-23.2.1-3.1.mga8
erlang-os_mon-23.2.1-3.1.mga8
erlang-hipe-23.2.1-3.1.mga8
erlang-dialyzer-23.2.1-3.1.mga8
erlang-observer-23.2.1-3.1.mga8
erlang-mnesia-23.2.1-3.1.mga8
erlang-crypto-23.2.1-3.1.mga8
erlang-erl_interface-23.2.1-3.1.mga8
erlang-runtime_tools-23.2.1-3.1.mga8
erlang-jinterface-23.2.1-3.1.mga8
erlang-public_key-23.2.1-3.1.mga8
erlang-ftp-23.2.1-3.1.mga8
erlang-erl_docgen-23.2.1-3.1.mga8
erlang-reltool-23.2.1-3.1.mga8
erlang-snmp-23.2.1-3.1.mga8
erlang-inets-23.2.1-3.1.mga8
erlang-ssh-23.2.1-3.1.mga8
erlang-edoc-23.2.1-3.1.mga8
erlang-doc-23.2.1-3.1.mga8
erlang-wx-23.2.1-3.1.mga8
erlang-eunit-23.2.1-3.1.mga8
erlang-xmerl-23.2.1-3.1.mga8
erlang-kernel-23.2.1-3.1.mga8
erlang-diameter-23.2.1-3.1.mga8
erlang-syntax_tools-23.2.1-3.1.mga8
erlang-compiler-23.2.1-3.1.mga8
erlang-erts-23.2.1-3.1.mga8
erlang-eldap-23.2.1-3.1.mga8
erlang-asn1-23.2.1-3.1.mga8
erlang-et-23.2.1-3.1.mga8
erlang-megaco-23.2.1-3.1.mga8

SRPM:
gnuplot-5.4.1-3.1.mga8
RPMS:
gnuplot-mode-5.4.1-3.1.mga8
gnuplot-5.4.1-3.1.mga8
gnuplot-nox-5.4.1-3.1.mga8
gnuplot-qt-5.4.1-3.1.mga8

SRPM:
guayadeque-0.4.6-7.git20201222.2.1.mga8
RPMS:				
guayadeque-0.4.6-7.git20201222.2.1.mga8

SRPM:
openbabel-2.4.1-14.1.mga8
RPMS:
lib(64)openbabel5-2.4.1-14.1.mga8
lib(64)openbabel-devel-2.4.1-14.1.mga8
openbabel-2.4.1-14.1.mga8
perl-openbabel-2.4.1-14.1.mga8
python3-openbabel-2.4.1-14.1.mga8
ruby-openbabel-2.4.1-14.1.mga8

SRPM:
openyahtzee-1.9.3-10.1.mga8
RPMS:
openyahtzee-1.9.3-10.1.mga8

SRPM:
mediainfo-20.09-2.1.mga8
RPMS:
mediainfo-gui-common-20.09-2.1.mga8
mediainfo-gui-wx-20.09-2.1.mga8
mediainfo-20.09-2.1.mga8
mediainfo-gui-qt-20.09-2.1.mga8

SRPM:
radiotray-ng-0.2.7-12.1.mga8
RPMS:
radiotray-ng-0.2.7-12.1.mga8

SRPM:
plplot-5.15.0-18.1.mga8
RPMS:
ocaml-plplot-5.15.0-18.1.mga8
python3-plplot-qt5-5.15.0-18.1.mga8
lib(64)qsastime0-5.15.0-18.1.mga8
lib(64)plplot-devel-5.15.0-18.1.mga8
plplot-5.15.0-18.1.mga8
lib(64)plplottcltk_main1-5.15.0-18.1.mga8
lib(64)plplotada4-5.15.0-18.1.mga8
python3-plplot-5.15.0-18.1.mga8
lib(64)plplot17-5.15.0-18.1.mga8
java-plplot-5.15.0-18.1.mga8
lib(64)tclmatrix10-5.15.0-18.1.mga8
lib(64)plplotqt2-5.15.0-18.1.mga8
lib(64)plplotcxx15-5.15.0-18.1.mga8
lib(64)plplotfortran0-5.15.0-18.1.mga8
d-plplot-5.15.0-18.1.mga8
lib(64)plplotwxwidgets1-5.15.0-18.1.mga8
lib(64)csiro0-5.15.0-18.1.mga8
lua-plplot-5.15.0-18.1.mga8
lib(64)plplottcltk14-5.15.0-18.1.mga8

SRPM:
spek-0.8.3-15.1.mga8
RPMS:
spek-0.8.3-15.1.mga8

SRPM:
scorched3d-44-14.1.mga8
RPMS:
scorched3d-44-14.1.mga8

SRPM:
slade-3.2.0-0.b2.3.1.mga8
RPMS:
slade-3.2.0-0.b2.3.1.mga8

SRPM:
tintii-2.10.0-12.1.mga8
RPMS:
tintii-2.10.0-12.1.mga8

SRPM:
trustedqsl-2.5.7-2.1.mga8
RPMS:
trustedqsl-2.5.7-2.1.mga8

SRPM:
urbanlightscape-1.4.0-12.1.mga8
RPMS:
urbanlightscape-1.4.0-12.1.mga8

SRPM:
vbam-2.1.4-1.1.mga8
RPMS:
vbam-2.1.4-1.1.mga8

SRPM:
veracrypt-1.24u7-7.1.mga8
RPMS:
veracrypt-1.24u7-7.1.mga8

SRPM:
openmsx-0.15.0-9.1.mga8
RPMS:
openmsx-catapult-0.15.0-9.1.mga8
openmsx-0.15.0-9.1.mga8

SRPM:
woeusb-3.3.1-4.1.mga8
RPMS:
woeusb-3.3.1-4.1.mga8

SRPM:
wxformbuilder-3.9.0-1.git20201102.2.1.mga8
RPMS:
wxformbuilder-3.9.0-1.git20201102.2.1.mga8

SRPM:
wxhexeditor-0.24-9.1.mga8
RPMS:
wxhexeditor-0.24-9.1.mga8

SRPM:
wxmaxima-20.12.1-1.1.mga8
RPMS:
wxmaxima-20.12.1-1.1.mga8

SRPM:
wxsqlite3-4.6.4-1.1.mga8
RPMS:
lib(64)wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8
lib(64)wx_gtk3u_wxsqlite3_3.1-devel-4.6.4-1.1.mga8

SRPM:
wxsvg-1.5.22-3.1.mga8
RPMS:
lib(64)wxsvg3-1.5.22-3.1.mga8
lib(64)wxsvg-devel-1.5.22-3.1.mga8

SRPM:
xchm-1.30-5.1.mga8
RPMS:
xchm-1.30-5.1.mga8

All packages aplhabetically. Updated dvdstyler. Comment 9 is now obsolete.

aegisub-3.2.2-14.git20191006.9.2.mga8
ariamaestosa-1.4.13-11.1.mga8
asc-2.6.1.0-20.1.mga8
asc-data-2.6.1.0-20.1.mga8
audacity-2.4.2-9.1.mga8
boinc-client-7.16.14-2.1.mga8
boinc-client-devel-7.16.14-2.1.mga8
boinc-client-static-7.16.14-2.1.mga8
boinc-manager-7.16.14-2.1.mga8
codeblocks-20.03-4.1.mga8
codelite-14.0-4.1.mga8
cwstudio-0.9.6-8.1.mga8
diff-pdf-0.4.1-4.1.mga8
d-plplot-5.15.0-18.1.mga8
dvdstyler-3.1.2-3.2.mga8
erlang-23.2.1-3.1.mga8
erlang-asn1-23.2.1-3.1.mga8
erlang-common_test-23.2.1-3.1.mga8
erlang-compiler-23.2.1-3.1.mga8
erlang-crypto-23.2.1-3.1.mga8
erlang-debugger-23.2.1-3.1.mga8
erlang-dialyzer-23.2.1-3.1.mga8
erlang-diameter-23.2.1-3.1.mga8
erlang-doc-23.2.1-3.1.mga8
erlang-edoc-23.2.1-3.1.mga8
erlang-eldap-23.2.1-3.1.mga8
erlang-erl_docgen-23.2.1-3.1.mga8
erlang-erl_interface-23.2.1-3.1.mga8
erlang-erts-23.2.1-3.1.mga8
erlang-et-23.2.1-3.1.mga8
erlang-eunit-23.2.1-3.1.mga8
erlang-examples-23.2.1-3.1.mga8
erlang-ftp-23.2.1-3.1.mga8
erlang-hipe-23.2.1-3.1.mga8
erlang-inets-23.2.1-3.1.mga8
erlang-jinterface-23.2.1-3.1.mga8
erlang-kernel-23.2.1-3.1.mga8
erlang-megaco-23.2.1-3.1.mga8
erlang-mnesia-23.2.1-3.1.mga8
erlang-observer-23.2.1-3.1.mga8
erlang-odbc-23.2.1-3.1.mga8
erlang-os_mon-23.2.1-3.1.mga8
erlang-parsetools-23.2.1-3.1.mga8
erlang-public_key-23.2.1-3.1.mga8
erlang-reltool-23.2.1-3.1.mga8
erlang-runtime_tools-23.2.1-3.1.mga8
erlang-sasl-23.2.1-3.1.mga8
erlang-snmp-23.2.1-3.1.mga8
erlang-ssh-23.2.1-3.1.mga8
erlang-ssl-23.2.1-3.1.mga8
erlang-stdlib-23.2.1-3.1.mga8
erlang-syntax_tools-23.2.1-3.1.mga8
erlang-tftp-23.2.1-3.1.mga8
erlang-tools-23.2.1-3.1.mga8
erlang-wx-23.2.1-3.1.mga8
erlang-xmerl-23.2.1-3.1.mga8
flamerobin-0.9.3.1-13.1.mga8
freedink-dfarc-3.14-8.1.mga8
freedv-1.4-4.1.mga8
freefilesync-11.3-2.1.mga8
fskbsetting-0.3.5-6.1.mga8
gnudl-1.0.0-1.3.mga8
gnuplot-5.4.1-3.1.mga8
gnuplot-mode-5.4.1-3.1.mga8
gnuplot-nox-5.4.1-3.1.mga8
gnuplot-qt-5.4.1-3.1.mga8
guayadeque-0.4.6-7.git20201222.2.1.mga8
java-plplot-5.15.0-18.1.mga8
lib(64)codeblocks0-20.03-4.1.mga8
lib(64)codeblocks-devel-20.03-4.1.mga8
lib(64)csiro0-5.15.0-18.1.mga8
lib(64)openbabel5-2.4.1-14.1.mga8
lib(64)openbabel-devel-2.4.1-14.1.mga8
lib(64)plplot17-5.15.0-18.1.mga8
lib(64)plplotada4-5.15.0-18.1.mga8
lib(64)plplotcxx15-5.15.0-18.1.mga8
lib(64)plplot-devel-5.15.0-18.1.mga8
lib(64)plplotfortran0-5.15.0-18.1.mga8
lib(64)plplotqt2-5.15.0-18.1.mga8
lib(64)plplottcltk14-5.15.0-18.1.mga8
lib(64)plplottcltk_main1-5.15.0-18.1.mga8
lib(64)plplotwxwidgets1-5.15.0-18.1.mga8
lib(64)qsastime0-5.15.0-18.1.mga8
lib(64)tclmatrix10-5.15.0-18.1.mga8
lib(64)wx_baseu3.1_5-3.1.5-1.mga8
lib(64)wx_baseu_net3.1_5-3.1.5-1.mga8
lib(64)wx_baseu_xml3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_adv3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_aui3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_core3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_gl3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_html3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_media3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_propgrid3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_qa3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_ribbon3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_richtext3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_stc3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_webview3.1_5-3.1.5-1.mga8
lib(64)wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8
lib(64)wx_gtk3u_wxsqlite3_3.1-devel-4.6.4-1.1.mga8
lib(64)wx_gtk3u_xrc3.1_5-3.1.5-1.mga8
lib(64)wxgtku3.1-devel-3.1.5-1.mga8
lib(64)wxsvg3-1.5.22-3.1.mga8
lib(64)wxsvg-devel-1.5.22-3.1.mga8
lua-plplot-5.15.0-18.1.mga8
mediainfo-20.09-2.1.mga8
mediainfo-gui-common-20.09-2.1.mga8
mediainfo-gui-qt-20.09-2.1.mga8
mediainfo-gui-wx-20.09-2.1.mga8
ocaml-plplot-5.15.0-18.1.mga8
openbabel-2.4.1-14.1.mga8
openmsx-0.15.0-9.1.mga8
openmsx-catapult-0.15.0-9.1.mga8
openyahtzee-1.9.3-10.1.mga8
perl-openbabel-2.4.1-14.1.mga8
plplot-5.15.0-18.1.mga8
python3-openbabel-2.4.1-14.1.mga8
python3-plplot-5.15.0-18.1.mga8
python3-plplot-qt5-5.15.0-18.1.mga8
radiotray-ng-0.2.7-12.1.mga8
ruby-openbabel-2.4.1-14.1.mga8
scorched3d-44-14.1.mga8
slade-3.2.0-0.b2.3.1.mga8
spek-0.8.3-15.1.mga8
tintii-2.10.0-12.1.mga8
trustedqsl-2.5.7-2.1.mga8
urbanlightscape-1.4.0-12.1.mga8
vbam-2.1.4-1.1.mga8
veracrypt-1.24u7-7.1.mga8
woeusb-3.3.1-4.1.mga8
wxformbuilder-3.9.0-1.git20201102.2.1.mga8
wxgtk3.1-3.1.5-1.mga8
wxhexeditor-0.24-9.1.mga8
wxmaxima-20.12.1-1.1.mga8
xchm-1.30-5.1.mga8

(In reply to Jani Välimaa from comment #20)
> (In reply to Philippe Didier from comment #16)
> > 4) BUT
> > - updated dvdstyler crashes (core dumped) with wxgtk 3.1.5 final when it
> > worked perfectly with the pre release version of wxgtk
> dvdstyler needs also updated wxsvg. dvdstyler starts with updated wxsvg, but
> emits a msg:
> dvdstyler: Symbol `_ZTV9wxSVGCtrl' has different size in shared object,
> consider re-linking.
> 
> I'll push another rebuild of dvdstyler with rebuilt wxsvg.

Hi Jani
Congratulations 
dvdstyler-3.1.2-3.2 rebuilt upon wxsvg is OK now

Does opencpn continue to work with the updated wxgtk packages installed?

If not, all of these updates will have to be moved to backports rather then
updates, so users can choose which packages break on their systems.

I.E. Have opencpn work or have kicad work, but not both.

Hi Dave and Jani

There is no problem for opencpn and its plugins : they remain built with gtk3.0 (the spec file is preventing to update this BuildRequire) 
And since gtk3.0 and gtk3.1.5 can stay side by side for Mageia8 opencpn is not affected

Beside this I can confirm everything works now for
audacity,
dvdstyler, 
freefilesync
guayadeque, 

tested on a real install of mageia8 up-to-date on  x86-64 hardware

PS I can't test the other programs : I never installed and don't use nor know them

Installed whole list and tried some
audacity OK
opencpn seems OK as far as I can see.
kicad segmentation fault
flamerobin opens in a far too small window,but it can be resized
openyahtzee opens in a far too small window,but it cannot be resized: useless
Radiotray-NG works OK
wxHexEditor gives after welcome window: ./src/common/socketiohandler.cpp(50): assert ""socket->m_fd != -1"" failed in Install_Callback(): shouldn't be called on invalid socket
Traceer/breekpunt-instructie (memorydump made)

CC: (none) => herman.viaene

(In reply to Herman Viaene from comment #27)
> kicad segmentation fault

Did you  also install python-wxpython4-4.1.1-1.1.mga8, bug 29291?

@Jani, for wxHexEditor, I see the same fault as Herman, #27

@Herman: correct kicad version: 5.1.12-1 ?

CC: (none) => jani.valimaa

(In reply to Morgan Leijström from comment #29)
> @Jani, for wxHexEditor, I see the same fault as Herman, #27
> 

Same seems to be happening with pkg from core/release also. Hitting "Continue" multiple times makes app window to open at some point.

Thanks Jani.
Opened bug 29869 for wxHexEditor, depending on this bug.

CC: jani.valimaa => (none)

@Morgan Comment 30:this new version seems OK, I will update bug 28532 accordingly.

I here hopefully identify what applications have been tested OK with the new
wxgtk-3.1.5-1, to aid people testing other applications.
Jani's monumental lists comment 21 & comment 22 are the references.

In addition, it seems to me that the following updated packages are part of the deal:
 kicad-5.1.12-1 OK https://bugs.mageia.org/show_bug.cgi?id=28352#c62
 python3-wxpython4-4.1.1-1.1 OK bug 29291
 wxsvg comment 20, comment 23 (need more detail)

Tested OK
---------
 audacity-2.4.2-9.1 comment 16 & comment 27
 boinc-manager-7.16.14-2.1 comment 18
 dvdstyler-3.1.2-3.2 comment 23
 freefilesync comment 16, comment 18
 opencpn comment 8, comment 27 (but this had no need to be updated. Is it still valid here? comment 25)
 python3-wxpython4-4.1.1-1.1 OK comment 19 &
  https://bugs.mageia.org/show_bug.cgi?id=29291#c6 &
  https://bugs.mageia.org/show_bug.cgi?id=28717#c31
 guayadeque comment 16
 radiotray-ng-0.2.7-12.1 comment 27
 urbanlightscape comment 18

OK?
---
 flamerobin-0.9.3.1-13.1 comment 27

No good
-------
 openyahtzee-1.9.3-10.1 comment 27
 wxhexeditor-0.24-9.1 bug 29869

I will try the last three to see whether the fault is in the current issued versions.

CC: (none) => lewyssmith

Previous flamerobin-0.9.3.1-13.mga8.x86_64
opens with a smallish window, but nothing unusual. OK.

Previous openyahtzee-1.9.3-10.mga8.x86_64
looks fine to me, will attach screenshot. True, the window cannot be resized, but it doesn not need to be on this system.

Previous wxhexeditor-0.24-9.mga8.x86_64
From Tools menu, license window briefly pops up & disappears. From terminal:
 $ wxHexEditor
 Segmentation fault (core dumped)
so the current fault was already there; bug 29869.

Created attachment 13083 [details]
Correct window for pre-update openyahtzee-1.9.3-10

Re Herman's comment 27 for the updated application's window being too small, hence useless. It was not before. To re-check after this update. It may be system dependant, here 1366x768.

Used the list at http://madb.mageia.org/tools/listRpmsForQaBug/bugnum/29848/application/0 with qarepo to get the 64-bit rpms. Had to change to wxhexeditor-0.24-9.2 before qarepo would go ahead.

Updated what was already installed first, and gnuplot and guayadeque seem to be working. Have to try the 32-bit version of guayadeque before that gets a full OK. (bug 29742)

0ad would not install without the corresponding 0ad-data. It's there, just missed getting on the list. Went back after it with qarepo, installed it, and ran it. Seems OK.

That appears to be the limit of what I have used before. The rest are new to me, and I can't tell if problems are from this update, or from user(me) error.

The slade gui crashes without loading. It gives a bunch of "crash information," some of which seems to be related to wxWindow and wxControl. There is more, but I don't understand any of it. Too much to list here - this one may need its own bug.

Starting the wxformbuilder gui from the menu throws error message after error message, all related to wx in some fashion. I never did get it to run; after about 10 of these messages I gave up.

Installed the wx gui for mediainfo without the others. It seems to be OK.

wxhexeditor-0.24-9.3.mga8 (this new version) bug 29869 is now perfectly OK

"Have to try the 32-bit version of guayadeque before that gets a full OK. (bug 29742)"

The 32-bit version of guayadeque still crashes, just as it did before, in a Plasma install on 64-bit hardware using the server kernel.

For those confused by the reference to 0ad in Comment 36, it was mentioned in Comment 0, but was left off lists in later comments.

(In reply to Thomas Andrews from comment #36)
> Used the list at
> http://madb.mageia.org/tools/listRpmsForQaBug/bugnum/29848/application/0
> with qarepo to get the 64-bit rpms. Had to change to wxhexeditor-0.24-9.2
> before qarepo would go ahead.
> 
Discovered this list was incomplete, so my 64-bit tests all have to be done again. This time I'm using the list in Comment 22, adding in packages for 0ad, and using a wild card to get the current wxhexeditor. (Hoping this list is now complete.)

> Updated what was already installed first, and gnuplot and guayadeque seem to
> be working. Have to try the 32-bit version of guayadeque before that gets a
> full OK. (bug 29742)

These two are still OK in 64-bit. Mentioned in Comment 38, 32-bit version of guayadeque still crashes.
> 
> 0ad would not install without the corresponding 0ad-data. It's there, just
> missed getting on the list. Went back after it with qarepo, installed it,
> and ran it. Seems OK.

0ad still OK.
> 
> That appears to be the limit of what I have used before. The rest are new to
> me, and I can't tell if problems are from this update, or from user(me)
> error.
> 
> The slade gui crashes without loading. It gives a bunch of "crash
> information," some of which seems to be related to wxWindow and wxControl.
> There is more, but I don't understand any of it. Too much to list here -
> this one may need its own bug.

With all of the packages updated, the slade gui now runs to a "first time SLADE Setup" screen, and with a full-screen main window.
> 
> Starting the wxformbuilder gui from the menu throws error message after
> error message, all related to wx in some fashion. I never did get it to run;
> after about 10 of these messages I gave up.

wxformbuilder now comes up with a proper-looking gui. I didn't go any farther than that, but I'd say it's probably OK.
> 
> Installed the wx gui for mediainfo without the others. It seems to be OK.

Mediainfo is still OK.

CC: (none) => andrewsfarm

(In reply to Thomas Andrews from comment #38)
> "Have to try the 32-bit version of guayadeque before that gets a full OK.
> (bug 29742)"
> 
> The 32-bit version of guayadeque still crashes, just as it did before, in a
> Plasma install on 64-bit hardware using the server kernel.
> 
The 32bits crash is not linked to wxgtk3.1.5 (if you install the same final release
as the one used to create the rpm)

1) The problem of guayadeque 32 bits is caused by the Mageia Build Flags used to build rpms : if I install the rpm on real 32bits Hardware guayadeque crashes

2) When I don't use the rpm but I simply build guayadeque from source with this build script : I got no crash 
#!/bin/bash

# avoid full rebuild if script is runs from symlinked location
cd $(readlink -f $(dirname $0))

./buildt
rm CMakeCache.txt
cmake ${1} . \
 -DCMAKE_BUILD_TYPE='Debug' \
 -DCMAKE_INSTALL_PREFIX=/usr \
 -DCMAKE_CXX_STANDARD="11" \
 && make ${2}

3) When I build it from source adding the mageia Build Flags it crashes :
#!/bin/bash

# avoid full rebuild if script is runs from symlinked location
cd $(readlink -f $(dirname $0))

./buildt
rm CMakeCache.txt
cmake ${1} . \
 -DCMAKE_BUILD_TYPE='Debug' \
 -DCMAKE_INSTALL_PREFIX=/usr \
 -DCMAKE_CXX_STANDARD="11" \
 && make ${2}
 CFLAGS="${CFLAGS:--O2 -g -pipe -Wformat -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -fstack-protector --param=ssp-buffer-size=4 -fasynchronous-unwind-tables}" ; export CFLAGS ; 
  CXXFLAGS="${CXXFLAGS:--O2 -g -pipe -Wformat -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -fstack-protector --param=ssp-buffer-size=4 -fasynchronous-unwind-tables}" ; export CXXFLAGS ; 
  FFLAGS="${FFLAGS:--O2 -g -pipe -Wformat -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -fstack-protector --param=ssp-buffer-size=4 -fasynchronous-unwind-tables }" ; export FFLAGS ; 
  FCFLAGS="${FCFLAGS:--O2 -g -pipe -Wformat -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -fstack-protector --param=ssp-buffer-size=4 -fasynchronous-unwind-tables }" ; export FCFLAGS ; 
  LDFLAGS="${LDFLAGS:- -Wl,--as-needed -Wl,--no-undefined -Wl,-z,relro -Wl,-O1 -Wl,--build-id -Wl,--enable-new-dtags}" ; export LDFLAGS



The problem already existed for the rpm from the core repo : look at https://bugs.mageia.org/show_bug.cgi?id=29742
There's no solution for this since we can't prevent the use of the build time flags when building a rpm for Mageia

(In reply to Philippe Didier from comment #40)

> The problem already existed for the rpm from the core repo : look at
> https://bugs.mageia.org/show_bug.cgi?id=29742
> There's no solution for this since we can't prevent the use of the build
> time flags when building a rpm for Mageia

I have a comment on this, but it doesn't belong here. I will take it to the other bug.

But, if there is one of these packages that doesn't work on one arch but does on the other, could there be others? 

If we are going to support 32-bit systems, then we should be testing on them when there might be a specific 32-bit problem, whatever it is. That is part of QA's job.

So, I have real 32-bit hardware, and given enough time I can test several of these packages. But the hardware is slow, and I have a life beyond Mageia, so "enough time" might be longer than we are willing to wait. Should I begin, or is it probably unnecessary? Or, if some are more likely to exhibit 32-bit problems than others, which ones should be tested?

Packages have same problem now as before the wxgtk change, I think we should ship them now with this update as is, in order not to delay it.

The big fixes users are waiting for and are ready, are PlayOnLinux and KiCAD.

We should just make sure this update does not break something that is not already broken.

(and remember to ship all rebuilt packages that depend on wxgtk3.1 simultaneously...)

Then open separate bugs per package that still does not work.

Hi Morgan

We may say that this update doesn't block 
https://bugs.mageia.org/show_bug.cgi?id=29742  !

the Mageia8 i586rpm of Guayadeque from Core Repo (built with the prerelease of wxgtk)
already crashed on some 32 bits hardwares when it was OK for 64bits

This is not at all linked to wxgtk but only a consequence of the rpm build flags used after Mageia7 

You may remove this blocking link since guayadeque works perfectly on 64bits, as it did before, after having been rebuilt in Updates-Testing upon wgtk 3.1.5 final release 
if you update all the wxgtk stuff in the same time


https://bugs.mageia.org/show_bug.cgi?id=29742 is stricly a "rpm build flags" bug

So, what is left to test so we can send this on its way? (I'm a little lost...)

With new wxgtk3.1-3.1.5-1.mga8
and all lib64wx_* pkgs currently on my system that were eligible to be updated to 3.1.5-1
and new python3-wxpython4-4.1.1-1.mga8
all 64-bit.

aegisub OK
-------
Installed previous aegisub-3.2.2-14.git20191006.9.1.mga8.x86_64, it worked.
Updated it to new  aegisub-3.2.2-14.git20191006.9.2.mga8, it worked.

ariamaestosa ???
------------
Installed prev version ariamaestosa-1.4.13-11.mga8 did not work, failed repeatedly with:
"./src/common/socketiohandler.cpp(50): assert ""socket->m_fd != -1"" failed in Install_Callback(): shouldn't be called on invalid socket"
Updated to new version ariamaestosa-1.4.13-11.1.mga8, failed likewise.
This may be due to inconsistent pre/post update packages. To re-visit.

asc OK
---
Installed prev version asc-2.6.1.0-20.mga8, it worked (but beware: it runs fullscreen, Ctl/Alt/arrow stays in the game, so you cannot swap desktops thus).
Updated to new version asc-2.6.1.0-20.1.mga8, works likewise.

openyahtzee ???
-----------
Already had it installed, comment 34: openyahtzee-1.9.3-10.mga8, it worked.
Updated to new openyahtzee-1.9.3-10.1.mga8. Herman's comment 27 about a too-small-and-cannot-resize window I confirm; will attach screenshot.

Created attachment 13084 [details]
Showing too-small yahtzee window after update

With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.

With new wxgtk3.1-3.1.5-1.mga8
and all lib64wx_* pkgs currently on my system that were eligible to be updated to 3.1.5-1; and new python3-wxpython4-4.1.1-1.mga8. Aall 64-bit

cwstudio OK
--------
Before update: cwstudio-0.9.6-8.mga8 does not work with:
"./src/common/framecmn.cpp(407): assert ""m_frameStatusBar != __null"" failed in SetStatusText(): no statusbar to set text for"
This may be due to mixed versions of wxgtk....
After update: cwstudio-0.9.6-8.1.mga8, worked!

diff-pdf ???
--------
Before update: diff-pdf-0.4.1-4.mga8, could not make work, lots of errors suggesting re-linking.
Update to: diff-pdf-0.4.1-4.1.mga8. An improvement, all the errors went. So the update is OK from that point of view. EXCEPT:
- I could not find a menu entry for it; and under Cinnamon, the menu search field did not show it. (Same before update).
- Running from terminal, with a correct command line, it simply returned to the prompt without doing anything!

flamerobin OK
----------
Already installed: flamerobin-0.9.3.1-13.mga8 failed, dumped; not to worry. Mixed wxgtk stuff.
Update to: flamerobin-0.9.3.1-13.1.mga8 weems to work. Despite Herman's comment 27 about the small window size, this looked quite adequate for a starters; and as he said, can be re-sized.

freedv OK
------
Installed current: freedv-1.4-4.mga8; did not work mixed wxgtk stuff.
Updated to:        freedv-1.4-4.1.mga8, looked to work fine.

freedink-dfarc OK
--------------
Installed current: freedink-dfarc-3.14-8.mga8
Having installed instead just 'freedink', I got:
 freedink-data-1.08.20190120-2.mga8
 freedink-109.6-2.mga8
 freedink-dfarc-3.14-8.mga8
& 3 menu entries/
- DFarc (crashed, GTK problems, crashed, acceptable here).
- Freedink, fullscreen game, seemed alive & kicking.
- FreedinkEdit, fullscreen something, displayed & 'lived'.
Updated to new: freedink-dfarc-3.14-8.1.mga8, it displayed & worked fine.

codeblocks OK ?
----------
Installed the current versions:
 lib64codeblocks0-20.03-4.mga8
 codeblocks-20.03-4.mga8, which crashed; no surprise.
Updated to the new:
 lib64codeblocks0-20.03-4.1.mga8
 codeblocks-20.03-4.1.mga8
which does launch OK. However, it starts with an error:
"./src/gtk/choice.cpp(284): assert ""Assert failure"" failed in GetString(): invalid index"
dismissing which 'continue' then starts the full GUI which nominally works. (The first time, it first displays a list of found compilers).
Can anyone check this in a totally UN-updated system?

Regarding diff-pdf, tested it with a couple of unrelated pdf files I have.
diff-pdf --output-diff=diff.pdf 210grandsale.pdf Slinky_flyer.pdf

It generated diff.pdf showing the differences when viewed with xpdf, so
it's working now. It also sets the return code to indicate the files are
different.

Note there is no menu entry or man page. Just
/usr/share/doc/diff-pdf/README.md describing how to use the command line
tool.

diff-pdf works here too.
It is a command line utility, but including option --view it opens a window.
Compared two large pdf manuals, with worked nicely.

BTW, do not confuse it with diffpdf (whithout a hyphen) which is a GUI tool. Which runs faster and is better at being a GUI tool anyways. Regarding features I have not compared them.

(In reply to Lewis Smith from comment #46)
> Created attachment 13084 [details]
> Showing too-small yahtzee window after update
> 
> With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.
Finally figured out what's the problem.

The issue with openyathzee is caused by a regression in wxgtk 3.1.5, but it's fixed in upstream GIT with the following commits:
https://github.com/wxWidgets/wxWidgets/commit/136574b1e0ca3a7165d379261952dfb3fb2f5ca6
https://github.com/wxWidgets/wxWidgets/commit/912f4b76ac42a79ff772cc9ea5cf8cb5c3f09960

The problem is that if I apply those patches, I need to rebuild all pkgs requiring lib(64)wx_gtk3u_core3.1_5 again. :\

Openyahtzee launched with fixed wxgtk crashes with undefined symbols:
$ openyahtzee 
openyahtzee: symbol lookup error: openyahtzee: undefined symbol: _ZN19wxTopLevelWindowGTK29WXSetInitialFittingClientSizeEi, version WXU_3.1

Re-checking new diff-pdf-0.4.1-4.1.mga8   OK
                --------
It seems to need either the --output-diff parameter (Dave) to produce an output file; or the --view parameter (Morgan); to actually do something.

(In reply to Jani Välimaa from comment #50)
> (In reply to Lewis Smith from comment #46)
> > Created attachment 13084 [details]
> > Showing too-small yahtzee window after update
> > With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.
> Finally figured out what's the problem.
> The issue with openyathzee is caused by a regression in wxgtk 3.1.5, but
> it's fixed in upstream GIT with the following commits:

> The problem is that if I apply those patches, I need to rebuild all pkgs
> requiring lib(64)wx_gtk3u_core3.1_5 again. :\
Out of the question in the circumstances.
What do we do? Hope nobody uses it (very likely)? If they do, they will raise a bug, and we will have to go through the whole wxgtk3.1-3.1.5-x business again. For a game?
 $ urpmq --whatrequires lib64wx_gtk3u_core3.1_5 | uniq
produces a huge list - 76.

fskbsetting OK
-----------
Installed prev fskbsetting-0.3.5-6.mga8, it crashed, no surprise.
Updated it to  fskbsetting-0.3.5-6.1.mga8, it opens OK. Initial window slightly small, but resizable.

codelite OK ?
--------
Note this pulled in also:
  lib64wx_gtk3u_wxsqlite3_3.1_0  4.6.4        1.mga8        x86_64
which is not in our list to update.
Current codelite-14.0-4.mga8, crashed, expected.
Updated:
- codelite-14.0-4.1.mga8.x86_64
- lib64wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8.x86_64
Selecting codelit to update did *not* automatically select the lib64wx_gtk3u_wxsqlite3_3.1_0 update; I think it should have. I looked for it & ticked it manually. Where did this update come from?
BTAIM, the updated application worked.

(In reply to Jani Välimaa from comment #50)
> (In reply to Lewis Smith from comment #46)
> > Created attachment 13084 [details]
> > Showing too-small yahtzee window after update
> > 
> > With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.
> Finally figured out what's the problem.
> 

> 
> The problem is that if I apply those patches, I need to rebuild all pkgs
> requiring lib(64)wx_gtk3u_core3.1_5 again. :\

what about going the other way... would it work to rebuild openyahtzee against wxgtk3.0 ?

(In reply to Lewis Smith from comment #51)
> Re-checking new diff-pdf-0.4.1-4.1.mga8   OK
>                 --------
> It seems to need either the --output-diff parameter (Dave) to produce an
> output file; or the --view parameter (Morgan); to actually do something.

With neither option, it just sets a return code, which can be useful in
scripts.

(In reply to Jani Välimaa from comment #50)
> (In reply to Lewis Smith from comment #46)
> > Created attachment 13084 [details]
> > Showing too-small yahtzee window after update
> > 
> > With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.
> Finally figured out what's the problem.
> 
> The issue with openyathzee is caused by a regression in wxgtk 3.1.5, but
> it's fixed in upstream GIT with the following commits:
> https://github.com/wxWidgets/wxWidgets/commit/
> 136574b1e0ca3a7165d379261952dfb3fb2f5ca6
> https://github.com/wxWidgets/wxWidgets/commit/
> 912f4b76ac42a79ff772cc9ea5cf8cb5c3f09960
> 
> The problem is that if I apply those patches, I need to rebuild all pkgs
> requiring lib(64)wx_gtk3u_core3.1_5 again. :\
> 
> Openyahtzee launched with fixed wxgtk crashes with undefined symbols:
> $ openyahtzee 
> openyahtzee: symbol lookup error: openyahtzee: undefined symbol:
> _ZN19wxTopLevelWindowGTK29WXSetInitialFittingClientSizeEi, version WXU_3.1

FYI. Installed all pkgs requiring lib(64)wx_gtk3u_core3.1_5 and grepped libraries and binaries for WXSetInitialFittingClientSize usage and the result is that 44 pkgs are using the symbol.

Created attachment 13086 [details]
Package list

Package list for comment 50.

(In reply to Jani Välimaa from comment #55)
> Created attachment 13086 [details]
> Package list
> 
> Package list for comment 50.
For comment 54 actually.

@Jani
(In reply to Thomas Backlund from comment #52)
> (In reply to Jani Välimaa from comment #50)
> > (In reply to Lewis Smith from comment #46)
> > > Created attachment 13084 [details]
> > > Showing too-small yahtzee window after update
> > > With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.
> >
> > Finally figured out what's the problem.
> > The problem is that if I apply those patches, I need to rebuild all pkgs
> > requiring lib(64)wx_gtk3u_core3.1_5 again. :\
> 
> what about going the other way... would it work to rebuild openyahtzee
> against wxgtk3.0 ?
I had already thought of this, it looks ideal for the moment.
It is exactly what Philippe did for opencpn comment 8.

(In reply to Lewis Smith from comment #57)
> @Jani
> (In reply to Thomas Backlund from comment #52)
> > (In reply to Jani Välimaa from comment #50)
> > > (In reply to Lewis Smith from comment #46)
> > > > Created attachment 13084 [details]
> > > > Showing too-small yahtzee window after update
> > > > With openyahtzee-1.9.3-10.1.mga8; the window cannot be resized.
> > >
> > > Finally figured out what's the problem.
> > > The problem is that if I apply those patches, I need to rebuild all pkgs
> > > requiring lib(64)wx_gtk3u_core3.1_5 again. :\
> > 
> > what about going the other way... would it work to rebuild openyahtzee
> > against wxgtk3.0 ?
> I had already thought of this, it looks ideal for the moment.
> It is exactly what Philippe did for opencpn comment 8.
It's doable, but please also note that we have 36 out of 44 pkgs mentioned in attachment 13086 [details] using wxSizer::SetSizeHints() directly or indirectly via wxWindow::SetSizerAndFit(). Hopefully not all are affected.

gnudl OK ???
-----
OLD: gnudl-1.0.0-1.2.mga8, pulls in a lot (36).
Enigmatic. No menu entry for gnudl, nor the only binary gdl. It started from the command prompt,
 $ gdl
  GDL - GNU Data Language, Version 1.0.0
- For basic information type HELP,/INFO
- Default library routine search path used (GDL_PATH/IDL_PATH env. var. not set): /usr/share/gnudatalanguage/lib
- Using WxWidgets as graphics library (windows and widgets).
- No startup file read (GDL_STARTUP/IDL_STARTUP env. var. not set). 
- Please report bugs, feature or help requests and patches at: https://github.com/gnudatalanguage/gdl
OpenCL: Failed to get number of platforms with clGetPlatformIDs(): -1001
OpenCL: Failed to get number of platforms with clGetPlatformIDs(): -1001
GDL> 
'exit' to leave. No GUI.
Can someone else try this?

NEW: gnudl-1.0.0-1.3.mga8 behaved identically. Do not know how to push it to a GUI.

gnuplot OK (comments 36,39)
-------

openbabel OK ???
---------
OLD: openbabel-2.4.1-14.mga8, the only other listed rpms pulled in was:
 lib64openbabel5-2.4.1-14.mga8
Menu entry did nothing, command is not 'babel' but:
 $ obabel
No input file or format spec or possibly a misplaced option.
Most options must come after the input files. (-i -o -O -m can be anywhwere.)
Open Babel 2.4.1 -- Jan  1 2021 -- 18:07:43
Usage:
obabel [-i<input-type>] <infilename> [-o<output-type>] -O<outfilename> [Options]
Try  -H option for more information.
 so do not know how to drive it.

NEW:
- lib64openbabel5-2.4.1-14.1.mga8.x86_64
- openbabel-2.4.1-14.1.mga8.x86_64
Again, ticking just the application did *not* automatically add the lib, I ticked it manually.
From menu, Lo! a GUI. So it basically works, but a problem: the window is deeper than the desktop. You can Alt/drag it higher to see the bottom, and shrink it vertically; but when you have, there is *no scroll bar*.
Can somebody else play with this pre the gtk updates to see whether this behaviour is new?

mediainfo OK ?
---------
BEFORE update:
 mediainfo-gui-common-20.09-2.mga8
 mediainfo-20.09-2.mga8
 mediainfo-gui-qt-20.09-2.mga8
 mediainfo-gui-wx-20.09-2.mga8
-gui:
"./src/gtk/window.cpp(2443): assert "m_wxwindow" failed in AddChildGTK(): Cannot add a child to a window without a client area"  Sort of expected.
-qt    GUI worked, starts by wanting a file.
AFTER update:
 mediainfo-gui-wx-20.09-2.1.mga8
 mediainfo-gui-common-20.09-2.1.mga8
 mediainfo-20.09-2.1.mga8
 mediainfo-gui-qt-20.09-2.1.mga8
-gui   Looks good, but LH icons have no bubble help. Different GUI from...
-qt    Works like previously, bubble help works.

radiotray-ng XXX
------------
Curr version: radiotray-ng-0.2.7-12.mga8 fails, expected due to mixed lib versions; from terminal:
"(radiotray-ng:31529): Gdk-CRITICAL **: 21:22:46.771: gdk_window_thaw_toplevel_updates: assertion 'window-update_and_descendants_freeze_count > 0' failed"
 But the program continues to run without GUI.

New version:  radiotray-ng-0.2.7-12.1.mga8
The GUI 'Radiotray-ng bookmark editor' works.
Launching the base application does *not* pop a GUI. Should it? However, it is left running invisibly. From terminal produces the same error as pre-update. Not good.

With updated GTK etc bits. X64.

plplot XXX
------
Old: plplot-5.15.0-18.mga8, which pulled in 14 things in all (see comment 21).
Of the 4 binaries:
/usr/bin/plserver
/usr/bin/pltcl
/usr/bin/pltek
/usr/bin/wxPLViewer
I could find none of the names in any menu, and from a terminal, they all - no panick - failed.

New:
- lib64csiro0-5.15.0-18.1.mga8.x86_64
- lib64plplot17-5.15.0-18.1.mga8.x86_64
- lib64plplotcxx15-5.15.0-18.1.mga8.x86_64
- lib64plplottcltk14-5.15.0-18.1.mga8.x86_64
- lib64plplottcltk_main1-5.15.0-18.1.mga8.x86_64
- lib64plplotwxwidgets1-5.15.0-18.1.mga8.x86_64
- lib64qsastime0-5.15.0-18.1.mga8.x86_64
- lib64tclmatrix10-5.15.0-18.1.mga8.x86_64
- plplot-5.15.0-18.1.mga8.x86_64

 $ pltcl
did nothing, returned at once.
 $ pltek
Usage: pltek filename 
At the prompt, the following replies are recognized:
   h,?	  Give this help message.
    q	  Quit program.
   <n>	  Go to the specified page number.
   -<n>   Go back <n> pages.
   +<n>   Go forward <n> pages.
 <Return> Go to the next page.
 $ wxPLViewer
xPlViewerApp::OnInit: error when creating wxPlFrame instance.  The message was Error initializing the shared memory and/or mutex needed for the application. The application will close
Can somebody else try this application *before* all the updates?

spek OK
----
Prev: spek-0.8.3-15.mga8 failed "./src/gtk/window.cpp(2443): assert "m_wxwindow" failed in AddChildGTK(): Cannot add a child to a window without a client area", no surprise.
New:  spek-0.8.3-15.1.mga8, worked. (Dated 2013!)

scorched3d OK
----------
Old: scorched3d-44-14.mga8, not unexpectedly (mixed libs) crashed.
New: scorched3d-44-14.1.mga8, works fine.

slade OK
-----
Prev: slade-3.2.0-0.b2.3.mga8, not trying any more!
Upd:  slade-3.2.0-0.b2.3.1.mga8 looks to work fine.

tintii XXX
------
Old: tintii-2.10.0-12.mga8
Upd: tintii-2.10.0-12.1.mga8, did not work:
"./src/common/wincmn.cpp(2490): assert ""!m_containingSizer"" failed in SetContainingSizer(): Adding a window already in a sizer, detach it first!"
twice, dismissing these 'continue' did lead to the GUI.
Opening an image file led to:
"./src/common/sizer.cpp(2109): assert "!(flags & (wxALIGN_RIGHT | wxALIGN_CENTRE_HORIZONTAL))" failed in DoInsert(): Horizontal alignment flags are ignored with wxEXPAND"
Continuing eventually showed a tiny 'thumbnail', then it crashes.
Can someone see whether this worked before? Or even now.

trustedqsl OK
----------
issued:  trustedqsl-2.5.7-2.mga8, crashed, normal
updated: trustedqsl-2.5.7-2.1.mga8 looks to work fine.

urbanlightscape OK ?
---------------
issued:  urbanlightscape-1.4.0-12.mga8, not tried.
updated: urbanlightscape-1.4.0-12.1.mga8 pops its GUI, then:
"./src/common/sizer.cpp(2109): assert "!(flags & (wxALIGN_RIGHT | wxALIGN_CENTRE_HORIZONTAL))" failed in DoInsert(): Horizontal alignment flags are ignored with wxEXPAND"
clicking 'continue' ends up with a working application.

re earlier comment 59
> mediainfo OK ?, OK seconded in comment 59.
re the previous comment 60, I should have looked at my own earlier comment 33:
> radiotray-ng-0.2.7-12.1 comment 27 tested OK
> urbanlightscape comment 18 tested OK

vbam OK
----
Prev: vbam-2.1.4-1.mga8, not tested (mixed gtk libs)
Upd:  vbam-2.1.4-1.1.mga8, seems to work, enigmatic how. Responsive GUI.
On urpme'ing this, its library lib64sfml2.5 (installed with it) was *not* pulled also, nor noted as an orphan. I removed it explicitly.

veracrypt OK
---------
Curr: veracrypt-1.24u7-7.mga8, did not try.
Upd:  veracrypt-1.24u7-7.1.mga8 looks to work, responsive GUI.

openmsx OK ?
-------
Original: openmsx-0.15.0-9.mga8, just installed it.
Updated:  openmsx-0.15.0-9.1.mga8, no menu entry, from terminal:
 $ openmsx
 info: SRAM file cbios-msx2+.cmos not found, assuming blank SRAM content.
pops an intial CBIOS window, which quickly goes to another saying:
" No cartridge found. This version of C-BIOS can only start cartridges. Please re-start your MSX (emulator) with a cartridge inserted".
Was this so with the pre-update version of everything?
No evident way to end it other than killing either the window or command.

woeusb OK ?
------
Create a Windows USB from a real Windows DVD or image
Current: woeusb-3.3.1-4.mga8, not tried;
Update:  woeusb-3.3.1-4.1.mga8
"./src/common/sizer.cpp(2109): assert "!(flags & (wxALIGN_RIGHT | wxALIGN_CENTRE_HORIZONTAL))" failed in DoInsert(): Horizontal alignment flags are ignored with wxEXPAND"
'Continue' leads to a working GUI.

wxformbuilder OK (also in comment 39)
-------------
Original: wxformbuilder-3.9.0-1.git20201102.2.mga8, pulled in also:
          lib64wx_gtk3u_media3.1_5. Did not try it.
Updated:  wxformbuilder-3.9.0-1.git20201102.2.1.mga8
          lib64wx_gtk3u_media3.1_5-3.1.5-1.mga8
Displays a large & comprehensive GUI - with no help! Looks good.

wxhexeditor X ???
-----------
I already had installed wxhexeditor-0.24-9.mga8, known not to work properly (bug 29869). Updated it to wxhexeditor-0.24-9.3.mga8
ex https://bugs.mageia.org/show_bug.cgi?id=29869#c10
and as far as I know, all the necessary wxgtk3.1-3.1.5-1 updates in place:
 lib64wx_baseu3.1_5-3.1.5-1.mga8
 lib64wx_baseu_net3.1_5-3.1.5-1.mga8
 lib64wx_baseu_xml3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_aui3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_core3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_gl3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_html3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_media3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_propgrid3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_qa3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_ribbon3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_richtext3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_stc3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_xrc3.1_5-3.1.5-1.mga8
 python3-opengl-3.1.5-1.mga8
 wxgtk3.1-3.1.5-1.mga8
It segfaulted immediately. This is not in agreement with:
 https://bugs.mageia.org/show_bug.cgi?id=29869#c11

Can somebody cross-check? Is there a wxgtk3.1 pkg missing? All my 3.1.5 versions of these are 3.1.5-1, i.e. none not updated.

wxhexeditor on x86_64 is working ok here.
$ rpm -qa|grep -e wxhex -e wx_gtk|sort
lib64wx_gtk3u_aui3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_core3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_html3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_propgrid3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_qa3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_richtext3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_webview3.1_5-3.1.5-1.mga8
lib64wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8
lib64wx_gtk3u_xrc3.1_5-3.1.5-1.mga8
wxhexeditor-0.24-9.3.mga8

Double check /etc/urpmi/skip.list if the package list installed doesn't match
the above.

Confirming the woeusb assertion failure and that selecting continue does get
to the gui. The backtrace ...
ASSERT INFO:
./src/common/sizer.cpp(2109): assert "!(flags & (wxALIGN_RIGHT | wxALIGN_CENTRE_HORIZONTAL))" failed in DoInsert(): Horizontal alignment flags are ignored with wxEXPAND

BACKTRACE:
[1] wxBoxSizer::DoInsert(unsigned long, wxSizerItem*)
[2] std::__cxx11::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >::~basic_stringbuf()
[3] std::__cxx11::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >::~basic_stringbuf()
[4] wxEntry(int&, wchar_t**)
[5] __libc_start_main

(In reply to Dave Hodgins from comment #62)
... [the list]
> the package list installed doesn't match the above.
My list shows many duplicates, _4-3.1.4-1 / _5-3.1.5-1, but note that this difference is also in the package name itself, before -version.
 lib64wx_gtk3u_aui3.1_4-3.1.4-1.mga8
 lib64wx_gtk3u_aui3.1_5-3.1.5-1.mga8
and similarly for:
 lib64wx_gtk3u_core3.1
 lib64wx_gtk3u_gl3.1
 lib64wx_gtk3u_html3.1
 lib64wx_gtk3u_qa3.1
 lib64wx_gtk3u_xrc3.1
I wonder whether I should delete the '_4-3.1.4-1' variants.
I also have duplicates of the form '3.0_0-3.0.5.1-1', but think this is OK.

OTOH I do not have webview3 or wxsqlite3, which I will add directly from Testing.

Continuing testing of updated applications with new gtk bits. Added 'new':
 lib64wx_gtk3u_webview3.1_5-3.1.5-1.mga8
 lib64wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8

wxmaxima OK
--------
CURR: wxmaxima-20.12.1-1.mga8
NEW:  wxmaxima-20.12.1-1.1.mga8. In the menus:
wxmaxima looks great.
xmaxima pops 2 wondows, a Browser & a Console. Looks fine.

lib64wx_gtk3u_wxsqlite3_3.1_0 ?
-----------------------------
lib64wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8
Installed (above), but how to know whether it works? Tests to date have not seemed to ask for it. Re-run failures?

lib64wxsvg3 OK ? [See also dvdstyler]
-----------
Installed from core/release:  lib64wxsvg3-1.5.22-3.mga8
Updated from updates_testing: lib64wxsvg3-1.5.22-3.1.mga8
Same question as for previous lib. I seem to recall Jani saying something needed it; yes, 'dvdstyler' cited OK, see comment 33. The testing of that did *not* cite the exact version of wxsvg used. Please do.
See comment 20, comment 23.

xchm OK
----
Old: xchm-1.30-5.mga8
New: xchm-1.30-5.1.mga8
starts up OK with a "Hello, and welcome to xCHM, the UNIX CHM viewer" screen, with an 'About' button with an 'X'. If you click that, the program ends immediately. Should it? Apparently yes, this looks another way of ending it.
Following the "If you'd like to know more about CHM, go to Pabs' CHM Specification page" link works a treat, and the button changes its title according to the context. Self-contained documentation!

Trying to summarise applications which do not (?) work post-updates.
Can others please, where possible, check the behaviour of applications PRE the big updates, where noted, to see whether doubtful or unclear behaviour post-update is new or not.

gnudl
----- needs more testing, comment 59

guayadeque
---------- 32-bit, existing problem. 64-bit OK.
comment 40, comment 43

openbabel
--------- Needs more testing, comment 59.

openmsx
------- Needs pre-update cross-checking, comment 61.

openyahtzee-1.9.3-10.1
----------- (window too small, no resize)
comment 50, comment 52, comment 54, comment 58
Does this warrant a separate bug? I do not think we should re-build & re-test again all the applications cited by Jani comment 56.

plplot (pltcl, pltek, wxPLViewer)
------ Unsure, needs pre-update cross-check, comment 60

radiotray-ng
------------ Needs confirmation testing.
Herman found it OK comment 27, I was not sure, comment 59.

tintii
------Needs more pre & post update testing, comment 60.

wxhexeditor
----------- (bug 29869) still crashes with me, but not for others. Deem OK.

(In reply to Lewis Smith from comment #66)
> Trying to summarise applications which do not (?) work post-updates.
> Can others please, where possible, check the behaviour of applications PRE
> the big updates, where noted, to see whether doubtful or unclear behaviour
> post-update is new or not.
> 
> 
> tintii
> ------Needs more pre & post update testing, comment 60.
> 

Please open a new bug and I'll switch tintii to build against wxgtk3.0.

Tintii is very old and probably dead upstream and I don't want to spend time to fix the code for wxgtk3.1.

(In reply to Lewis Smith from comment #66)
> Trying to summarise applications which do not (?) work post-updates.
> Can others please, where possible, check the behaviour of applications PRE
> the big updates, where noted, to see whether doubtful or unclear behaviour
> post-update is new or not.
> 
> radiotray-ng
> ------------ Needs confirmation testing.
> Herman found it OK comment 27, I was not sure, comment 59.
> 
Radiotray-ng doesn't open a GUI when launched. It's added to system tray with a gray icon looking like a radio tower.

Only part using wxgtk is the bookmark editor.

$ ldd /usr/bin/radiotray-ng | grep libwx

$ ldd /usr/bin/rtng-bookmark-editor | grep libwx
	libwx_gtk3u_core-3.1.so.5 => /lib64/libwx_gtk3u_core-3.1.so.5 (0x00007f8e81752000)
	libwx_baseu-3.1.so.5 => /lib64/libwx_baseu-3.1.so.5 (0x00007f8e814cd000)

(In reply to Dave Hodgins from comment #63)
> Confirming the woeusb assertion failure and that selecting continue does get
> to the gui. The backtrace ...
> ASSERT INFO:
> ./src/common/sizer.cpp(2109): assert "!(flags & (wxALIGN_RIGHT |
> wxALIGN_CENTRE_HORIZONTAL))" failed in DoInsert(): Horizontal alignment
> flags are ignored with wxEXPAND
> 
> BACKTRACE:
> [1] wxBoxSizer::DoInsert(unsigned long, wxSizerItem*)
> [2] std::__cxx11::basic_stringbuf<char, std::char_traits<char>,
> std::allocator<char> >::~basic_stringbuf()
> [3] std::__cxx11::basic_stringbuf<char, std::char_traits<char>,
> std::allocator<char> >::~basic_stringbuf()
> [4] wxEntry(int&, wchar_t**)
> [5] __libc_start_main

woeusb issue should be fixed with RPM/SRPM woeusb-3.3.1-4.2.mga8.

Thanks Jani for your several comments; and the huge amount of work you have done on this saga.

(In reply to Jani Välimaa from comment #68)
> (In reply to Lewis Smith from comment #66)
> > radiotray-ng
> > ------------ Needs confirmation testing.
> > Herman found it OK comment 27, I was not sure, comment 59.
> Radiotray-ng doesn't open a GUI when launched. It's added to system tray
> with a gray icon looking like a radio tower.
> Only part using wxgtk is the bookmark editor.
This explains what I saw; I did note that the GUI editor bit worked, so:
 radiotray-ng OK
 ------------

(In reply to Jani Välimaa from comment #67)
> > tintii
> > ------Needs more pre & post update testing, comment 60.
> Please open a new bug and I'll switch tintii to build against wxgtk3.0.
> Tintii is very old and probably dead upstream and I don't want to spend time
> to fix the code for wxgtk3.1.
Will open bug tomorrow. A quick look at the site revealed no dates. If it really is old & not being developed, should we not drop it from Cauldron?

(In reply to Jani Välimaa from comment #69)
> woeusb issue should be fixed with RPM/SRPM woeusb-3.3.1-4.2.mga8.
Will try this tomorrow.

I did not understand why this cannot be done also for just openyahtzee (comment 58). I will open a bug for this also.

(In reply to Jani Välimaa from comment #67)
> Please open a new bug and I'll switch tintii to build against wxgtk3.0.
Done, Bug 29906 for 'tintii.'.

> I did not understand why this cannot be done also for just openyahtzee
> (comment 58). I will open a bug for this also.
This (in previous comment) should have followed 'tintii' (build against wxgtk3.0), not 'woeusb'.
Opened separate Bug 29907 for 'openyahtzee'.

woeusb                       OK
------ (comments 61, 63, 69)
Issued:  woeusb-3.3.1-4.mga8
New upd: woeusb-3.3.1-4.2.mga8
from menu & command '$ woeusbgui' pops its little window without the previous initial error message. Thanks Jani.

openyahtzee OK
-----------
Bug 29907 closed fixed.

real 32bit hardware LXDE updated to latest.
added all required packages as per comment 22

attempted to run the following applications:
  
aegisub - Aborted (core dumped)

asc - ok
audacity - ok
babel  : info text in terminal
boincmgr - ok
codeblocks - ok
codelite - crash window
diff-pdf - no window
dvdstyler - ok
erl -ok (terminal)  (erlang)
flamerobin -ok

freedink
Segmentation fault (core dumped)

freedink-dfac - ok

freedv - ok
gnuplot - opens a wxt terminal

guayadeque Aborted (core dumped)

mediainfo-gui - ok
openyahtzee - (forgot to test)
radiotray - ok
openmsx-catapult - ok
scorched3d3 - ok

slade - Illegal instruction (core dumped)

spek - ok
tintii - complaint windows but application opens
urbanlightscape - ok

vbam - Illegal instruction (core dumped)

veracrypt - Illegal instruction (core dumped

woeusb - ok
wxformbuilder - ok

wxPLViewer ([plplot) not ok

wxmaxima - ok
xchm - ok
-----------------------------------------------------------------------

To satisfy dependencies, the following packages are going to be installed:
  Package                        Version      Release       Arch    
(medium "Core Updates Testing (distrib5)")
  libwx_baseu3.1_5               3.1.5        1.mga8        i586    
  libwx_baseu_net3.1_5           3.1.5        1.mga8        i586    
  libwx_baseu_xml3.1_5           3.1.5        1.mga8        i586    
  libwx_gtk3u_adv3.1_5           3.1.5        1.mga8        i586    
  libwx_gtk3u_aui3.1_5           3.1.5        1.mga8        i586    
  libwx_gtk3u_core3.1_5          3.1.5        1.mga8        i586    
  libwx_gtk3u_gl3.1_5            3.1.5        1.mga8        i586    
  libwx_gtk3u_html3.1_5          3.1.5        1.mga8        i586    
  libwx_gtk3u_media3.1_5         3.1.5        1.mga8        i586    
  libwx_gtk3u_propgrid3.1_5      3.1.5        1.mga8        i586    
  libwx_gtk3u_qa3.1_5            3.1.5        1.mga8        i586    
  libwx_gtk3u_ribbon3.1_5        3.1.5        1.mga8        i586    
  libwx_gtk3u_richtext3.1_5      3.1.5        1.mga8        i586    
  libwx_gtk3u_stc3.1_5           3.1.5        1.mga8        i586    
  libwx_gtk3u_webview3.1_5       3.1.5        1.mga8        i586    
  libwx_gtk3u_wxsqlite3_3.1_0    4.6.4        1.1.mga8      i586    
  libwx_gtk3u_xrc3.1_5           3.1.5        1.mga8        i586    
  libwxgtku3.1-devel             3.1.5        1.mga8        i586
  wxgtk3.1                       3.1.5        1.mga8        i586
  
aegisub - Aborted (core dumped)
asc - ok
audacity - segmentation fault (core dumped)
babel  : info text in terminal
boincmgr - not ok
codeblocks - not ok
codelite - Aborted (core dumped)
diff-pdf - no window / info text in terminal
dvdstyler Aborted (core dumped)
erl -ok (terminal)  (erlang)
flamerobin -Aborted (core dumped)
freedink -Aborted (core dumped)
freedink-dfac - Aborted (core dumped)
freedv - Aborted (core dumped)
gnudl - command not found
gnuplot - opens a wxt terminal
guayadeque - Aborted (core dumped)
mediainfo-gui - segmentation fault (core dumped)
openmsx-catapult - segmentation fault (core dumped)
openyahtzee - (forgot to test)
radiotray - ok
scorched3d3 - segmentation fault (core dumped)
slade - Illegal instruction (core dumped)
spek - segmentation fault (core dumped)
tintii - segmentation fault (core dumped)
urbanlightscape - segmentation fault (core dumped)
vbam - Illegal instruction (core dumped)
veracrypt - segmentation fault (core dumped)
woeusb - ok
wxformbuilder - segmentation fault (core dumped)
wxPLViewer ([plplot) - segmentation fault (core dumped)
wxmaxima - Aborted (core dumped)
xchm - segmentation fault (core dumped)

update system:
  
aegisub - Aborted (core dumped)
asc - ok
audacity - ok
babel  : info text in terminal
boincmgr - ok
codeblocks - ok
codelite - crash handler window
diff-pdf - no window / info text in terminal
dvdstyler - ok
erl -ok (terminal)  (erlang)
flamerobin -ok
freedink -segmentation fault (core dumped)
freedink-dfac - ok
freedv - ok
gnuplot - opens a wxt terminal
guayadeque - crash report / Aborted (core dumped)
mediainfo-gui - ok
openmsx-catapult - ok
openyahtzee - ok
radiotray-ng - ok
scorched3d3 - ok
slade - Illegal instruction (core dumped)
spek - ok
tintii - ok
urbanlightscape - ok
vbam - Illegal instruction (core dumped)
veracrypt - Illegal instruction (core dumped)
woeusb - ok
wxformbuilder - ok

wxPLViewer 

(wxPLViewer:13111): Gtk-WARNING **: 20:35:13.917: Theme parsing error: gtk.css:3:33: Failed to import: Error opening file /home/home/.config/gtk-3.0/window_decorations.css: No such file or directory
wxPlViewerApp::OnInit: error when creating wxPlFrame instance.  The message was Error initializing the shared memory and/or mutex needed for the application. The application will close

wxmaxima - ok
xchm - ok

CC: (none) => westel

Thank you for testing a lot on 32 bit!

§ You may need updated python3-wxpython4-4.1.1-1.1 for some apps
(Bug 29291 - python3-wxpython4... - to be released at the same time)
Does that help against some crashes?

§ diff-pdf is a command line tool, for graphic output see comment 48 & 49

(In reply to Lewis Smith from comment #65)
> Continuing testing of updated applications with new gtk bits. Added 'new':
>  lib64wx_gtk3u_webview3.1_5-3.1.5-1.mga8
>  lib64wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8
>
> 
> lib64wx_gtk3u_wxsqlite3_3.1_0 ?
> -----------------------------
> lib64wx_gtk3u_wxsqlite3_3.1_0-4.6.4-1.1.mga8
> Installed (above), but how to know whether it works? Tests to date have not
> seemed to ask for it. Re-run failures?
> 
>
This library is used by guayadeque
Since guayadeque is OK (only on 64bits ...) then lib64wx_gtk3u_wxsqlite3_3.1_0 is OK

Trying at last erlang 64-bit.
erlang OK
------
PRE:  erlang-23.2.1-3.mga8, installing just this pulled in 39 pkgs.
POST: erlang-23.2.1-3.1.mga8, all 39 got updated.
But how on earth do you fire it up? Nothing 'erlang*' in the menus, nor from the command line that I can find (/usr/bin/erlang). Ah, from Ben:
 $ erl
which pops an erlang console - not easy to get out of either! As for a GUI ?

Pinging Len in case he knows better. (Remove yourself from CCs if not).

CC: (none) => tarazed25

(In reply to Morgan Leijström from comment #74)
> Thank you for testing a lot on 32 bit!
> 
> § You may need updated python3-wxpython4-4.1.1-1.1 for some apps
> (Bug 29291 - python3-wxpython4... - to be released at the same time)
> Does that help against some crashes?

no

> 
> § diff-pdf is a command line tool, for graphic output see comment 48 & 49

thanks, 
diff-pdf --view
but no window generated

add:

ariamaestosa - crash text boxes
cwstudio ok from application launcher

(In reply to ben mcmonagle from comment #77)
> diff-pdf --view
> but no window generated

It needs the two pdf files to be compared. It's the output of the diff-pdf that
is shown in the window, with no window created if there are no files to compare.

@lewis in reply to comment 76:
Nope.  erlang is one of the things on my todo list (sometime/never).  tutorialspoint is a good starting point.  As far as development environments are concerned it mentions plugins for Eclipse but does not enlarge on graphics programming.  If the programmer is familiar with wxgtk it might become fairly obvious how to call graphics functions.  There are hints on the web that wxwidgets has been ported to erlang and it is possible that the three wx_gtk3 packages in dependencies cover that.

See what you mean about exiting.  ^G^Ca

(In reply to ben mcmonagle from comment #78)
> add:
> 
> ariamaestosa - crash text boxes

Please create a separate bug report about ariamaestosa and I'll fix it. It's already fixed in Cauldron.

tintii OK
------
tintii-2.10.0-12.3.mga8, just adding it to this list from its own Bug 29906 now fixed.

(In reply to Jani Välimaa from comment #81)
> (In reply to ben mcmonagle from comment #78)
> > add:
> > 
> > ariamaestosa - crash text boxes
> 
> Please create a separate bug report about ariamaestosa and I'll fix it. It's
> already fixed in Cauldron.


see bug 29932

Best I could do ;)

CC: (none) => westel

One by one the doubtful applications have been sorted. I attempt here to list those remaining to fix - or at least check out:

 ariamaestosa : in the workshop, bug 29932. I will test the update.
ANYBODY please clarify these uncertainties:
 gnudl : unsure, comment 66.
 openbabel : unsure, comment 59, comment 73/update system
 plplot (wxPLViewer) : unsure, comment 60, comment 73, same 'error'.
-----------------------------------------------------
Ben had a few 32-bit dumps for programs OK'd for x64, comment 63 (update); (for guayadeque - known 32-bit problem, nothing we can do).
ANYONE WITH REAL 32-bit hardware, please test these updates:
 aegisub : OK comment 45
 codelite : OK comment 51
 freedink : OK comment 47
 slade : OK comment 60
 vbam : OK comment 61
 veracrypt : OK comment 61
If the 32-bit crashes are confirmed, these may be condemned to a permanent 'not for 32-bit' status.

> Ben had a few 32-bit dumps for programs OK'd for x64, comment 63 (update)
above, should read "comment 73/update system".

(In reply to Lewis Smith from comment #84)
> One by one the doubtful applications have been sorted. I attempt here to
> list those remaining to fix - or at least check out:
> 
>  plplot (wxPLViewer) : unsure, comment 60, comment 73, same 'error'.

According to upstream "The wxPLViewer application is only used internally by the wxwidgets device and is not a general utility that is directly invoked by a user."

https://sourceforge.net/p/plplot/plplot/ci/4e9679b2266e190f66969f21392cdcf9766a58c9

ariamaestosa OK
------------
New ariamaestosa-1.4.13-11.2.mga8 tested OK:
 https://bugs.mageia.org/show_bug.cgi?id=29932#c3

plplot (wxPLViewer) OK
-------------------
New plplot-5.15.0-18.1.mga8, deemed OK according to prev comment.

additional; 32bit test via 64bit hardware install.

using automated test launchGUIs, it was determined there were 47 applications added via comment 21 entries, including some via dependencies.

test resulted in 3 failures:
FreeFileSync~%F
gnuplot
ocaml

of these, only  FreeFileSync was a true failure for me.

$ FreeFileSync 

(FreeFileSync:11014): Gtk-WARNING **: 22:46:48.802: Theme parsing error: gtk.css:2:33: Failed to import: Error opening file /home/home/.config/gtk-3.0/window_decorations.css: No such file or directory

(FreeFileSync:11014): GLib-GObject-WARNING **: 22:46:49.032: invalid cast from 'GtkImage' to 'wxGtkImage'
Segmentation fault (core dumped)


erlang was not found to test, as it does not have a .desktop file in /usr/share/applications.

11 minutes to test the basic operation of 47 applications, plus a little extra for testing the 3 "fails" and erlang.

big disclaimer:
does not include my slow typing to install the pre-update packages with slow download time, and then upgrade the system via the same slow download speed.

Created attachment 13106 [details]
launchGUIs test file

Installed the whole stack less those items already pushed, about 130 packages.
The updates ran smoothly.
Tried out interactive erlang using some of the introductory examples from
https://www.erlang.org/doc/man/erlang.html

$ cat test.txt
atom_to_list('Erlang').
atom_to_list('你好').
Bin = <<1,2,3,4,5,6,7,8,9,10>>.
binary_part(Bin,{byte_size(Bin), -5}).
Zed = <<1,2,3>>.
binary_part(Zed,{0,2}).

Invoked erl shell and pasted test.txt onto the command line:
$ erl
Erlang/OTP 23 [erts-11.1.5] [source] [64-bit] [smp:8:8] [ds:8:8:10] [async-threads:1] [hipe]

Eshell V11.1.5  (abort with ^G)
1> atom_to_list('Erlang').
"Erlang"
2> atom_to_list('你好').
[20320,22909]
3> Bin = <<1,2,3,4,5,6,7,8,9,10>>.
<<1,2,3,4,5,6,7,8,9,10>>
4> binary_part(Bin,{byte_size(Bin), -5}).
<<6,7,8,9,10>>
5> Zed = <<1,2,3>>.
<<1,2,3>>
6> binary_part(Zed,{0,2}).
<<1,2>>
7> ^Ca

Found sudoku.erl in the examples folder and copied sudoko.hrl and wx.erl to the working directory and compiled a BEAM file, hoping to be able to demonstrate wx graphics.
$ erlc -I ../../sudoku.hrl,.././wx.hrl sudoku.erl
$ ll sudoku*
-rw-r--r-- 1 lcl lcl 1352 Jan 25 16:33 sudoku.beam
-rw-r--r-- 1 lcl lcl 1540 Jan 25 12:26 sudoku.erl
-rw-r--r-- 1 lcl lcl 1360 Jan 25 12:32 sudoku.hrl

However, I have no idea how to use the BEAM file, which appears to be an erlang module.  There are indications that such files need to be imported into an emulator of some kind.

Tried to look at it in eclipse but was told that there is no editor to deal with that format but it did show the embedded strings in the bytecode, referring to "sudoku" and "fun game".
That is as far as this goes without a few more weeks of erlanging.

(In reply to ben mcmonagle from comment #88)
> additional; 32bit test via 64bit hardware install.
Is this real 32-bit h/w, needed because of its often restricted graphics instructions.
> test resulted in 3 failures:
> FreeFileSync~%F
> gnuplot
> ocaml
> of these, only  FreeFileSync was a true failure for me.
>  $ FreeFileSync 
> (FreeFileSync:11014): GLib-GObject-WARNING **: 22:46:49.032: invalid cast
> from 'GtkImage' to 'wxGtkImage'
> Segmentation fault (core dumped)
Does this mean to add it to the 32-bit only failures in comment 84 ?

What were the (acceptable) problems with gnuplot (certified OK for x64) & ocaml ? You are the first to have tried ocaml, I will for x64.

Re aegisub, codelite, freedink, slade, vbam, veracrypt previously noted as post-update 32-bit failures comments 73, 84, do you confirm those failures, or are you saying they now test OK for x32? [No time yet to peruse your test results].
-----
@Len: thanks for battling with Erlang, we accept it is OK for the update.

(In reply to Lewis Smith from comment #91)
> (In reply to ben mcmonagle from comment #88)
> > additional; 32bit test via 64bit hardware install.
> Is this real 32-bit h/w, needed because of its often restricted graphics
> instructions.

no, it was, in this instance 32bit install on 64bit hardware.
I was  working out the method of using launchGUIs to run just the added applications, and took advantage of the trial to  run the tests.

> > test resulted in 3 failures:
> > FreeFileSync~%F
> > gnuplot
> > ocaml
> > of these, only  FreeFileSync was a true failure for me.
> >  $ FreeFileSync 
> > (FreeFileSync:11014): GLib-GObject-WARNING **: 22:46:49.032: invalid cast
> > from 'GtkImage' to 'wxGtkImage'
> > Segmentation fault (core dumped)
> Does this mean to add it to the 32-bit only failures in comment 84 ?
> 

as this was different hardware, I will, now that  I have the guiApps file, will  later run it on the 32bit hardware.
will advise.

my tests were: invoke in terminal as above and note response.

I had not  previously tested FFS, and a "ctrl+f" search of this page finds no test confirmation, 32 or 64 bit.  
as  you are aware  launchGUIs lets you test all applications, it does  not forget to invoke an application, and we have 40+ to test in this update.

> What were the (acceptable) problems with gnuplot (certified OK for x64) &
> ocaml ? You are the first to have tried ocaml, I will for x64.

acceptable as ocaml opens a terminal window when run from the application menu.
acceptable as gnuplot gives  a positive response when run in terminal, I still cannot find it in the application menu! (all DE install)

> 
> Re aegisub, codelite, freedink, slade, vbam, veracrypt previously noted as
> post-update 32-bit failures comments 73, 84, do you confirm those failures,
> or are you saying they now test OK for x32? [No time yet to peruse your test
> results].

because launchGUIs found them and  gives the correct invocation to run, they worked (on the 64bit hardware).


e.g: vbam correct invocation is  *visualboyadvance-m*, knowing this, then able to locate in application launcher subheadings  the vba-m  launcher. 
vbam in a terminal  fails, *vba* from the menu  text box run command gives a *vba-m* icon to choose, but no action. I certainly was not looking for *VisualBoyAdvance-m* in the menu launcher. 

many of these applications I was not even aware of existed, let alone know how to invoke successfully from a terminal.

Thanks for the explanations. And all your work on this.

@TJ : I think you have a real 32-bit box, are you able to try just the few 32-bit doubtful (64-bit OK) programs noted in comment 84 ? POST update. Should be quick to do.
TIA

Working on it. Currently using qarepo to download the list in Comment 22, plus 0ad, which was left off. Even with Foolishness connected directly to my router with a wired connection, it's taking a while. Will check the programs in Comment 84, plus 0ad, right after Foolishness quits complaining about his arthritis.

@Ben: Gnuplot is in the "Graphics" menu on my Plasma systems, will try it from the Xfce menu on Foolishness.

Foolishness has spoken:

0ad: 0ad-data is in the 32-bit testing repos, but the game itself is not. It seems it is not a new situation here, I missed it when testing bug 29091 because I only checked 64-bits. There is a 32-bit 0ad-0.0.23b, but not a 32-bit 0ad-0.0.25 from bug 29091, or a 0ad-0.0.25b here. This should go on, I think, and a new bug filed to get the 32-bit package built.

The rest of these were run from the Xfce menus:

aegisub - Looks OK
codelite - crashed
freedink-dfarc - gui runs. Trying to play the game throws an error, but I believe that's from a non-experienced user. Didn't look like a crash from wxgtk to me.
slade - looks OK
vbam - opens a window that's completely black. Gameboy emulator that IIRC from another update needs a downloaded file to work. Not crashing - I think it's OK.
veracrypt - Gui looks OK, but I didn't try to do anything with it.
gnuplot - OK

Anything else?

(In reply to Lewis Smith from comment #91)

> Does this mean to add it to the 32-bit only failures in comment 84 ?
> 
> What were the (acceptable) problems with gnuplot (certified OK for x64) &
> ocaml ? You are the first to have tried ocaml, I will for x64.
> 
> Re aegisub, codelite, freedink, slade, vbam, veracrypt previously noted as
> post-update 32-bit failures comments 73, 84, do you confirm those failures,
> or are you saying they now test OK for x32? [No time yet to peruse your test
> results].
> -----


back onto the real 32bit hardware,  cpu is: AMD Athlon(tm) XP 2400+ :                      
see bug 29932 for hardware.

edited contents of guiFails.text:

aegisub~%f       correct failure - will not launch from application menu
codelite~%f      correct failure - will not launch from application menu / crash report windows
freedink         correct failure - will not launch from application menu
freedinkedit     correct failure - will not launch from application menu
FreeFileSync~%F  correct failure - will not launch from application menu
slade            correct failure - will not launch from application menu
veracrypt        correct failure - will not launch from application menu
visualboyadvance-m~%f (vbam) correct failure - will not launch from application menu


guayadeque  unlisted failure - will not launch from application menu - an assertation failed! text windows (many!)

false failures:
gnuplot   launches a virtual terminal window from application menu
ocaml     launches a virtual terminal window from application menu

seems these are CPU/ hardware specific failures, as TJs 32bit hardware they seem ok

(In reply to ben mcmonagle from comment #96)
> back onto the real 32bit hardware,  cpu is: AMD Athlon(tm) XP 2400+ :       
> 
> see bug 29932 for hardware.
> 
> edited contents of guiFails.text:
> 
> aegisub~%f       correct failure - will not launch from application menu
> codelite~%f      correct failure - will not launch from application menu /
> crash report windows
> freedink         correct failure - will not launch from application menu
> freedinkedit     correct failure - will not launch from application menu
> FreeFileSync~%F  correct failure - will not launch from application menu
> slade            correct failure - will not launch from application menu
> veracrypt        correct failure - will not launch from application menu
> visualboyadvance-m~%f (vbam) correct failure - will not launch from
> application menu
> 
> 
> guayadeque  unlisted failure - will not launch from application menu - an
> assertation failed! text windows (many!)
> 
> false failures:
> gnuplot   launches a virtual terminal window from application menu
> ocaml     launches a virtual terminal window from application menu
> 
> seems these are CPU/ hardware specific failures, as TJs 32bit hardware they
> seem ok

Indeed that's possible. I see your processor doesn't support sse2, where my P4 does, and that has been known to knock out a lot of applications. One thing, it's probably not the graphics capability, as Foolishness has a RV200 gpu, which is a little older and probably less capable than your RV280. Another difference, but probably not a factor, is that you are using LXDE, where I'm using Xfce.

guayadeque is an old problem, dating back before this update. It fails on all 32-bit installs that I've tried on real hardware, even on 64-bit hardware. Curiously, it doesn't seem to fail in a 32-bit VirtualBox guest. Those who are studying it are blaming "sse2 creep" for the problem. I disagree, but not being a developer in any sense of the word I may not know what I'm talking about, so disagreeing is about all I can do.

What happened here?  Looks like the wxgtk update got pushed with no announcement and without including the packages that were rebuilt against it.

I see kicad bug 28352 got marked fixed by robot and packages are in updates on the mirror I use.

And yes wxgtk packages are vanished from testing, and in updates I see lib64wx_gtk3u_core3.1_5-3.1.5-1.mga8.x86_64.rpm
 - but with file date 2021-07-25 !?
https://ftp.acc.umu.se/mirror/mageia/distrib/8/x86_64/media/core/updates/

In current situation i believe it is best to push all depending packages as is, and later fix any problems left.

That's going to take a while to get the advisories ready and added to svn.

Another solution would be to rebuild the old version of wxgtk an kicad with
higher release numbers, and push those two only, then rebuild the new
versions.

(In reply to David Walser from comment #98)
> What happened here?  Looks like the wxgtk update got pushed with no
> announcement and without including the packages that were rebuilt against it.
Is this true? If so, it is serious. How did it happen? I thought nothing could go out without 'validated' and 'advisory'.

I was going to review where we stand with this main bug and its various offshoots, and ask how to organise it. The advisory will be long, basically comment 21 but with the few subsequent application updates incorporated. And what about the at-the-same-time updates like kicad, that games pkg dropped from M8 but here re-introduced.

So long as the post-c21 updated applications are included, this was/is probably good to go on condition that the related bits go at the same time.

As for the few (& variable) 32-bit problems, we should just admit them. I wonder about enlarging the guayadeque bug to include them, generalise it.

Least total work is to just chew on and get it all out...

Most packages are just rebuild against new wxgtk, i don't think it is much work with advisory.  Possibly note we know about 32-bit problems in the packages it concern, and that it is nothing new for this update (right?)

Suggested work order: first finish easy and popular needs:


__Already having advisory - just push?

Bug 29291 - needed by kicad which is already pushed

Bug 29869 - wxHexEditor


__Have suggested advisory - upload, push?

Bug 29808 - aom 


__Separate bug of popular package:
Bug 28717 PlayOnLinux (need advisory)

guayadeque is bug 29742

See Also: (none) => https://bugs.mageia.org/show_bug.cgi?id=29742

(In reply to Lewis Smith from comment #102)
> (In reply to David Walser from comment #98)
> > What happened here?  Looks like the wxgtk update got pushed with no
> > announcement and without including the packages that were rebuilt against it.
> Is this true? If so, it is serious. How did it happen? I thought nothing
> could go out without 'validated' and 'advisory'.
> 
Bug 28352 (Kicad) was validated and had an advisory, but somehow got pushed with unresolved "depends on" bugs. That, apparently, drew along at least some of the new wxgtk packages with it.

Once this now-emergency is taken care of, we need to get to the bottom of how that bug (28352) got pushed AS SOON AS POSSIBLE.

It is also strange that it happened suddenly after long idle time.

All wxgtk packages got moved.

In my comment 99 I questioned the wxgtk packages file date but they are correct.
https://bugs.mageia.org/show_bug.cgi?id=28352#c33
I forgot how long this have been hanging.

Regarding the depending (set either way) bugs
Forgot to list tintii, openyahtzee, ariamaestosa.  Refining my list:

(By "OK", I mean they seem to be able to set Validated, or already is)



__OK and advisory uploaded - just push?

Bug 29291 - python3-wxpython4, needed by kicad which is already pushed

Bug 29869 - wxHexEditor



__OK and have suggested advisory - upload, push?

Bug 29808 - aom 



__OK but need advisory written and uploaded

Bug 28717 - PlayOnLinux (need advisory) (also need 29291 listed above)

Bug 29906 - tintii

Bug 29907 - openyahtzee

Bug 29932 - ariamaestosa 

Bug 29742 - guayadeque "crashes on i586 but not on x86_64; packaging problem with Mageia's build flag macro"  Yes; The 64 bit works and the 32 bit did not before the update either, right?  IMO, Now advisory with note on that, push, new bug to fix 32 bit build problem later.

Blocks: (none) => 29742
See Also: https://bugs.mageia.org/show_bug.cgi?id=29742 => (none)

Comment 6 has approx 44 packages listed. I haven't looked yet to see how many
of those are fixed, need to be fixed, or are ok as is.

My understanding is that the script that pushes updates looks for:
- At least one ok whiteboard entry
- Validated keyword
- Advisory keyword and advisory committed to svn
- All bugs that the update depends also being ready to push

Bug 28352 should not have been processed by that script without this bug
being validated with the advisory committed to svn and advisory keyword
added.

Unless I misunderstand the way the script works, my best guess is that a
sysadmin manually pushed the update for bug 28352 and wxgtk, not using
the normal script.

Looks like the easiest way to go for now is to finish getting the rest of the
packages moved from updates testing to updates.

Normally, I'd add an advisory to svn listing the wxgtk and all of the other
srpms that need to be pushed and they would all get moved from updates testing
to updates at the same time. All of the rpm packages generated from a given srpm are automatically included in the move.

If any of the srpms listed in the advisory are not present in the updates
testing repo, none of the listed srpms get moved (no action taken). 

As some srpms/rpms have already been moved, those srpms now have to be excluded
from the advisory.

So I need to put together a list of which srpms still need to be moved for
an advisory that I'll add to svn for this bug report, even though wxgtk itself
will have to be excluded.

I'll work on that now.

(In reply to Dave Hodgins from comment #108)
> Comment 6 has approx 44 packages listed. I haven't looked yet to see how many
> of those are fixed, need to be fixed, or are ok as is.
> 

Dave, The list from Comment 6 is incomplete. 

0ad is not on that list, but it has been mentioned elsewhere. Also "urpmq --whatrequires-recursive wxgtk3.1" lists 0ad as needing it. A rebuilt 0ad IS in testing, and I have tried it out. See comment  95. I believe it should go out with the rest of this, and a new bug filed to add in the 32-bit version that was left out of the last update.

I don't know what else may have been left off that list.

(In reply to Morgan Leijström from comment #107)
> Regarding the depending (set either way) bugs
> (By "OK", I mean they seem to be able to set Validated, or already is)
> 
> __OK and advisory uploaded - just push?
> Bug 29291 - python3-wxpython4, needed by kicad which is already pushed
> Bug 29869 - wxHexEditor
to include with *this* bug update & advisory.
> 
> __OK and have suggested advisory - upload, push?
> Bug 29808 - aom
started as a security bug, so will have itw own advisory;
but can only be pushed with this bug (depends on it) - and should therefore *not* include it in its updates.
> 
> __OK but need advisory written and uploaded
> Bug 28717 - PlayOnLinux (need advisory) (also need 29291 listed above)
This one is complicated, because it re-introduces a dropped package, so cannot be pushed as an update to something that does not exist. OTOH it needs to be in core to be upgradeable from Mageia 7. Needs expert advice.

> Bug 29906 - tintii
> Bug 29907 - openyahtzee
> Bug 29932 - ariamaestosa
These 3 all closed fixed, but only in the context of this bug update, on which they depend. They should be included in this bug's advisory, do not need their own.

> Bug 29742 - guayadeque "crashes on i586 but not on x86_64; packaging problem
This is arkward. The pkg is part of *this* bug, so should be pushed (updated) with it - included in this advisory. Its own bug (which depends on this one) can be:
- left open for itself alone
- closed wontfix
- expanded to include other applications known not to work on 32-bit.

(In reply to Dave Hodgins from comment #108)
> Comment 6 has approx 44 packages listed. I haven't looked yet to see
> how manyof those are fixed, need to be fixed, or are ok as is.
This bug notes in detail that all have now been tested OK for 64-bit, most for 32-bit except a few now known not to work. I think this should be noted in the advisory: tough.
Repeat that the advisory will (I think) be largely comment 21 [+ 0ad] with a few application versions updated since. What think others?

(In reply to Lewis Smith from comment #111)
> > Bug 28717 - PlayOnLinux (need advisory) (also need 29291 listed above)
> This one is complicated, because it re-introduces a dropped package, so
> cannot be pushed as an update to something that does not exist. OTOH it
> needs to be in core to be upgradeable from Mageia 7. Needs expert advice.

In Bug 28717, Dave Hodgins, #29:
> It should be in updates since it was in m7, and upgrading only uses release
> and updates repositories, with no option to include packages from backports.

Created attachment 13108 [details]
Advisory prepared for pushing remaining packages

Please review the attached advisory (already committed to svn, but can still be
changed).

aom and python3-wxpython have their own bug reports.

kicad already pushed. Also the actual wxgtk srpm that would normally go in this
advisory was also already pushed as part of the kicad update.

Thomas didn't like my description. :-) and changed it to ...
  Mageia 8 was released with a wxgtk 3.1.5 prerelease. This update provides
  the final 3.1.5 release. In this update the wxWidgets has been built without
  disable-glcanvasegl in order to support KiCad hardware acceleration, since
  Kicad does not support the EGL canvas backend which is otherwise the default
  in wxWidgets 3.1.5.

  Packages depending on wxgtk3.1 have been rebuilt so they work with the the
  new version.

Adding the tags to allow the updates to now get pushed.

Whiteboard: (none) => MGA8-64-OK
Keywords: (none) => advisory, validated_update
CC: (none) => sysadmin-bugs

An update for this issue has been pushed to the Mageia Updates repository.

https://advisories.mageia.org/MGAA-2022-0017.html

Status: NEW => RESOLVED
Resolution: (none) => FIXED