Re: [Rosegarden-user] Fedora and LV2
Confirmed - [76e11482] builds with lilv and without gtk2
and lv2 plugins are operative.
Chuck.
On 12/06/2024 20:11, Ted Felix wrote:
Just pushed Philip's fix as [76e11482]. Please test latest git. You
should now be able to build with lilv and without gtk2.
Philip is looking further into the gtk2 issue. I'll let you know if
there are any developments. We are tracking this on Feature Request
462, but at 7 pages it's pretty huge to navigate.
https://sourceforge.net/p/rosegarden/feature-requests/462/
Ted.
On 6/12/24 11:41 AM, Ted Felix wrote:
Ok, I'll see if Philip can fix this. Hopefully it's just a
forgotten #ifdef someplace.
Ted.
On 6/12/24 11:34 AM, chuck elliot wrote:
'tried a new build with lilv-devel installed but not gtk2-devel.
build failed at the link stage:
/usr/bin/ld: ../librosegardenprivate.so: undefined reference to
`Rosegarden::LV2Gtk::getInstance()'
collect2: error: ld returned 1 exit status
make[2]: *** [src/CMakeFiles/rosegarden.dir/build.make:127:
rosegarden] Error 1
make[1]: *** [CMakeFiles/Makefile2:264:
src/CMakeFiles/rosegarden.dir/all] Error 2
make: *** [Makefile:146: all] Error 2
so it looks like there is a dependency...
Regards,
Chuck.
On 12/06/2024 14:44, Ted Felix wrote:
Thanks for tracking this down. Yes, the LV2 stuff can use gtk2,
but it doesn't need it. You should be able to install lilv-devel
without gtk2-devel and that might work. Let me know how it goes if
you try it.
Note: A fresh clone shouldn't be necessary. Just delete your
build directory, create a new one, switch into it and run cmake again.
Ted.
On 6/11/24 4:40 AM, chuck elliot wrote:
...but a fresh git clone of RG is now compiling without
gtk2-devel and lilv-devel...
cmake has disabled LILV support
...now it runs again.
Chuck.
On 09/06/2024 11:10 pm, Chris Fancsy wrote:
DuckDuckGo was unable to verify sender identity
Hi Chuck,
I have the same problem with the compiled 24.06 version on Fedora
40.
I compiled the 23.12 version and it runs fine.
The problem is from the LV2 plugin support which is using GTK2. I
don't think there's a way to disable LV2 plugin support at compile
time.
If I find a solution I will post it here.
Chris
On Sun, Jun 9, 2024 at 5:05 AM chuck elliot
wrote:
Thanks for the link Ted but I'm afraid I didn't glean anything
useful
there...
RG 21.12 runs ok on this system (lying around in /usr/local/bin)
Rosegarden version: 21.12 ("Cloud Nine")
Build key: ac38e2622f (21.06-24-gb9da2db9d)
Built against Qt version: 5.15.2
Current version info is:
Rosegarden version: 24.12 ("Incredible")
Build key: 442818511e (24.06-1-gf91b7e227)
Built against Qt version: 5.15.13
How would I be "requesting the gtk widgets"? I am not aware of
having,
or how I would have, done so...
I assume that my RG preferences would be applied the same to
both of the
above versions?
TIA
Chuck.
On 07/06/2024 7:14 pm, Ted Felix wrote:
Just some limited searching seems to indicate that the "gtk"
widget
style was dropped around Qt 5.7 or so. Maybe you're running a
newer
version of Qt and you are requesting the gtk widgets?
https://bbs.archlinux.org/viewtopic.php?id=214147&p=2
See post #36.
All the trails are from 2016/2017, though. Not sure they are
relevant today.
Ted.
On 6/7/24 6:24 AM, chuck elliot wrote:
Rosegarden won't run at the moment.
The problem seems to be something to do with "GtkWidget" (see
below).
24.06 built fine from git source.
I have recently upgraded to fedora 39 and this is probably the
first
time I have tried to
run RG since. Same result on a similar/identical system.
gtk3-3.24.41-1.fc39.x86_64
gtk4-4.12.5-1.fc39.x86_64
RG output:
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
add private field to invalid (non-instantiatable) type
''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
register existing type 'GtkBuildable'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_interface_add_prerequisite: assertion 'G_TYPE_IS_INTERFACE
(interface_type)' failed
(rosegarden:725222): GLib-CRITICAL **: 11:12:47.844:
g_once_init_leave: assertion 'result != 0' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_widget_get_style_context: assertion 'GTK_IS_WIDGET
(widget)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_style_context_lookup_color: assertion 'GTK_IS_STYLE_CONTEXT
(context)' failed
(rosegarden:7
Re: [Rosegarden-user] Fedora and LV2
That was quick!
'will test and get back to you.
Thanks,
Chuck.
On 12/06/2024 20:11, Ted Felix wrote:
Just pushed Philip's fix as [76e11482]. Please test latest git. You
should now be able to build with lilv and without gtk2.
Philip is looking further into the gtk2 issue. I'll let you know if
there are any developments. We are tracking this on Feature Request
462, but at 7 pages it's pretty huge to navigate.
https://sourceforge.net/p/rosegarden/feature-requests/462/
Ted.
On 6/12/24 11:41 AM, Ted Felix wrote:
Ok, I'll see if Philip can fix this. Hopefully it's just a
forgotten #ifdef someplace.
Ted.
On 6/12/24 11:34 AM, chuck elliot wrote:
'tried a new build with lilv-devel installed but not gtk2-devel.
build failed at the link stage:
/usr/bin/ld: ../librosegardenprivate.so: undefined reference to
`Rosegarden::LV2Gtk::getInstance()'
collect2: error: ld returned 1 exit status
make[2]: *** [src/CMakeFiles/rosegarden.dir/build.make:127:
rosegarden] Error 1
make[1]: *** [CMakeFiles/Makefile2:264:
src/CMakeFiles/rosegarden.dir/all] Error 2
make: *** [Makefile:146: all] Error 2
so it looks like there is a dependency...
Regards,
Chuck.
On 12/06/2024 14:44, Ted Felix wrote:
Thanks for tracking this down. Yes, the LV2 stuff can use gtk2,
but it doesn't need it. You should be able to install lilv-devel
without gtk2-devel and that might work. Let me know how it goes if
you try it.
Note: A fresh clone shouldn't be necessary. Just delete your
build directory, create a new one, switch into it and run cmake again.
Ted.
On 6/11/24 4:40 AM, chuck elliot wrote:
...but a fresh git clone of RG is now compiling without
gtk2-devel and lilv-devel...
cmake has disabled LILV support
...now it runs again.
Chuck.
On 09/06/2024 11:10 pm, Chris Fancsy wrote:
DuckDuckGo was unable to verify sender identity
Hi Chuck,
I have the same problem with the compiled 24.06 version on Fedora
40.
I compiled the 23.12 version and it runs fine.
The problem is from the LV2 plugin support which is using GTK2. I
don't think there's a way to disable LV2 plugin support at compile
time.
If I find a solution I will post it here.
Chris
On Sun, Jun 9, 2024 at 5:05 AM chuck elliot
wrote:
Thanks for the link Ted but I'm afraid I didn't glean anything
useful
there...
RG 21.12 runs ok on this system (lying around in /usr/local/bin)
Rosegarden version: 21.12 ("Cloud Nine")
Build key: ac38e2622f (21.06-24-gb9da2db9d)
Built against Qt version: 5.15.2
Current version info is:
Rosegarden version: 24.12 ("Incredible")
Build key: 442818511e (24.06-1-gf91b7e227)
Built against Qt version: 5.15.13
How would I be "requesting the gtk widgets"? I am not aware of
having,
or how I would have, done so...
I assume that my RG preferences would be applied the same to
both of the
above versions?
TIA
Chuck.
On 07/06/2024 7:14 pm, Ted Felix wrote:
Just some limited searching seems to indicate that the "gtk"
widget
style was dropped around Qt 5.7 or so. Maybe you're running a
newer
version of Qt and you are requesting the gtk widgets?
https://bbs.archlinux.org/viewtopic.php?id=214147&p=2
See post #36.
All the trails are from 2016/2017, though. Not sure they are
relevant today.
Ted.
On 6/7/24 6:24 AM, chuck elliot wrote:
Rosegarden won't run at the moment.
The problem seems to be something to do with "GtkWidget" (see
below).
24.06 built fine from git source.
I have recently upgraded to fedora 39 and this is probably the
first
time I have tried to
run RG since. Same result on a similar/identical system.
gtk3-3.24.41-1.fc39.x86_64
gtk4-4.12.5-1.fc39.x86_64
RG output:
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
add private field to invalid (non-instantiatable) type
''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
cannot
register existing type 'GtkBuildable'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_interface_add_prerequisite: assertion 'G_TYPE_IS_INTERFACE
(interface_type)' failed
(rosegarden:725222): GLib-CRITICAL **: 11:12:47.844:
g_once_init_leave: assertion 'result != 0' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_widget_get_style_context: assertion 'GTK_IS_WIDGET
(widget)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_style_context_lookup_color: assertion 'GTK_IS_STYLE_CONTEXT
(context)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:
Re: [Rosegarden-user] Fedora and LV2
Just pushed Philip's fix as [76e11482]. Please test latest git. You
should now be able to build with lilv and without gtk2.
Philip is looking further into the gtk2 issue. I'll let you know if
there are any developments. We are tracking this on Feature Request
462, but at 7 pages it's pretty huge to navigate.
https://sourceforge.net/p/rosegarden/feature-requests/462/
Ted.
On 6/12/24 11:41 AM, Ted Felix wrote:
Ok, I'll see if Philip can fix this. Hopefully it's just a forgotten
#ifdef someplace.
Ted.
On 6/12/24 11:34 AM, chuck elliot wrote:
'tried a new build with lilv-devel installed but not gtk2-devel.
build failed at the link stage:
/usr/bin/ld: ../librosegardenprivate.so: undefined reference to
`Rosegarden::LV2Gtk::getInstance()'
collect2: error: ld returned 1 exit status
make[2]: *** [src/CMakeFiles/rosegarden.dir/build.make:127:
rosegarden] Error 1
make[1]: *** [CMakeFiles/Makefile2:264:
src/CMakeFiles/rosegarden.dir/all] Error 2
make: *** [Makefile:146: all] Error 2
so it looks like there is a dependency...
Regards,
Chuck.
On 12/06/2024 14:44, Ted Felix wrote:
Thanks for tracking this down. Yes, the LV2 stuff can use gtk2, but
it doesn't need it. You should be able to install lilv-devel without
gtk2-devel and that might work. Let me know how it goes if you try it.
Note: A fresh clone shouldn't be necessary. Just delete your build
directory, create a new one, switch into it and run cmake again.
Ted.
On 6/11/24 4:40 AM, chuck elliot wrote:
...but a fresh git clone of RG is now compiling without
gtk2-devel and lilv-devel...
cmake has disabled LILV support
...now it runs again.
Chuck.
On 09/06/2024 11:10 pm, Chris Fancsy wrote:
DuckDuckGo was unable to verify sender identity
Hi Chuck,
I have the same problem with the compiled 24.06 version on Fedora 40.
I compiled the 23.12 version and it runs fine.
The problem is from the LV2 plugin support which is using GTK2. I
don't think there's a way to disable LV2 plugin support at compile
time.
If I find a solution I will post it here.
Chris
On Sun, Jun 9, 2024 at 5:05 AM chuck elliot
wrote:
Thanks for the link Ted but I'm afraid I didn't glean anything useful
there...
RG 21.12 runs ok on this system (lying around in /usr/local/bin)
Rosegarden version: 21.12 ("Cloud Nine")
Build key: ac38e2622f (21.06-24-gb9da2db9d)
Built against Qt version: 5.15.2
Current version info is:
Rosegarden version: 24.12 ("Incredible")
Build key: 442818511e (24.06-1-gf91b7e227)
Built against Qt version: 5.15.13
How would I be "requesting the gtk widgets"? I am not aware of
having,
or how I would have, done so...
I assume that my RG preferences would be applied the same to both
of the
above versions?
TIA
Chuck.
On 07/06/2024 7:14 pm, Ted Felix wrote:
Just some limited searching seems to indicate that the "gtk" widget
style was dropped around Qt 5.7 or so. Maybe you're running a newer
version of Qt and you are requesting the gtk widgets?
https://bbs.archlinux.org/viewtopic.php?id=214147&p=2
See post #36.
All the trails are from 2016/2017, though. Not sure they are
relevant today.
Ted.
On 6/7/24 6:24 AM, chuck elliot wrote:
Rosegarden won't run at the moment.
The problem seems to be something to do with "GtkWidget" (see
below).
24.06 built fine from git source.
I have recently upgraded to fedora 39 and this is probably the
first
time I have tried to
run RG since. Same result on a similar/identical system.
gtk3-3.24.41-1.fc39.x86_64
gtk4-4.12.5-1.fc39.x86_64
RG output:
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add private field to invalid (non-instantiatable) type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkBuildable'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_interface_add_prerequisite: assertion 'G_TYPE_IS_INTERFACE
(interface_type)' failed
(rosegarden:725222): GLib-CRITICAL **: 11:12:47.844:
g_once_init_leave: assertion 'result != 0' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_widget_get_style_context: assertion 'GTK_IS_WIDGET (widget)'
failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_style_context_lookup_color: assertion 'GTK_IS_STYLE_CONTEXT
(context)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.868:
gtk_style_context_get_background_color: assertion
'GTK_IS_STYLE_CONTEXT (context)' failed
qt.qpa.gtk: Property
Re: [Rosegarden-user] MIDI Instrument creation script
I'd like to jump in on this.
The original MIDI spec made no requirements about when Bank MSB or LSB should
be applied... because it doesn't matter!
The only thing that's relevant is that these should stable and be between 0 and
127 when a program change comes along. It's up to the destination to decide
what to do if the values are invalid - all the synths I've used simply ignore
values that point to a non-existent bank and leave things as previously set.
All the hardware synths I've used don't need you to send bank information at
all. If you don't, they simply assume you want to use the same as the last time
it was powered up.
It also doesn't matter if you repeatedly send the same value (apart from
wasting MIDI bandwidth).
Not naming names but there was one sequencer I knew of that insisted on sending
a bank pair for every bank change, and if you didn't set both it assumed zero
for the one you didn't set.
On Wed, 12 Jun 2024 11:35:23 -0400
Ted Felix wrote:
> Looking at this now. I think I see what you are getting at. We
>absolutely support multiple LSB's per MSB and multiple MSB's per LSB.
>The problem is that the UI makes it annoying to achieve this. (In fact
>based on the pattern of MSB/LSB values you could end up in a very
>frustrating maze.)
>
> Procedure to reproduce:
>
>- Create a new device.
>- Create a bank with MSB:LSB 10:0.
>- Create another bank.
>- Adjust the bank MSB to 10. You will find that you cannot because the
>LSB is 0 and it won't let you select 10:0 again. So the MSB will jump
>from 9 to 11.
>- Adjust the LSB to 1, then you can adjust the MSB to 10.
>
> This proactive error prevention seems irritating. Maybe we should
>redo this as a pop-up dialog for setting MSB and LSB at the same time
>and checking them for dupes only on dismissal?
>
> Is this what you are seeing? Are there any other issues you've
>encountered with this dialog?
>
>Ted.
>
>On 5/31/24 6:49 PM, Olivier S. wrote:
>> According to my tests, both Patch banks and Rhythm Sets work,
>> nevertheless with a limitation that I think is a Rosegarden bug, unless
>> you can explain where I'm wrong: it's impossible within the same device
>> to manage both Patches and Rhythm Sets for the same MSB. Rosegarden only
>> accepts one bank with the same LSB, even if you check the percussion
>> box, which I think is precisely designed to manage this conflict. At
>> this stage, my only solution is to create two devices for my JV1080, one
>> for patches on MIDI channels 1-9/11-16 and another specific to
>> percussion on channel 10 to manage rhythm sets. Frustrating.
>
>
>___
>Rosegarden-user mailing list
>Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
>https://lists.sourceforge.net/lists/listinfo/rosegarden-user
--
Will J Godfrey {apparently now an 'elderly'}
https://willgodfrey.bandcamp.com/
http://yoshimi.github.io
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user
Re: [Rosegarden-user] Fedora and LV2
Ok, I'll see if Philip can fix this. Hopefully it's just a forgotten
#ifdef someplace.
Ted.
On 6/12/24 11:34 AM, chuck elliot wrote:
'tried a new build with lilv-devel installed but not gtk2-devel.
build failed at the link stage:
/usr/bin/ld: ../librosegardenprivate.so: undefined reference to
`Rosegarden::LV2Gtk::getInstance()'
collect2: error: ld returned 1 exit status
make[2]: *** [src/CMakeFiles/rosegarden.dir/build.make:127: rosegarden]
Error 1
make[1]: *** [CMakeFiles/Makefile2:264:
src/CMakeFiles/rosegarden.dir/all] Error 2
make: *** [Makefile:146: all] Error 2
so it looks like there is a dependency...
Regards,
Chuck.
On 12/06/2024 14:44, Ted Felix wrote:
Thanks for tracking this down. Yes, the LV2 stuff can use gtk2, but
it doesn't need it. You should be able to install lilv-devel without
gtk2-devel and that might work. Let me know how it goes if you try it.
Note: A fresh clone shouldn't be necessary. Just delete your build
directory, create a new one, switch into it and run cmake again.
Ted.
On 6/11/24 4:40 AM, chuck elliot wrote:
...but a fresh git clone of RG is now compiling without
gtk2-devel and lilv-devel...
cmake has disabled LILV support
...now it runs again.
Chuck.
On 09/06/2024 11:10 pm, Chris Fancsy wrote:
DuckDuckGo was unable to verify sender identity
Hi Chuck,
I have the same problem with the compiled 24.06 version on Fedora 40.
I compiled the 23.12 version and it runs fine.
The problem is from the LV2 plugin support which is using GTK2. I
don't think there's a way to disable LV2 plugin support at compile
time.
If I find a solution I will post it here.
Chris
On Sun, Jun 9, 2024 at 5:05 AM chuck elliot wrote:
Thanks for the link Ted but I'm afraid I didn't glean anything useful
there...
RG 21.12 runs ok on this system (lying around in /usr/local/bin)
Rosegarden version: 21.12 ("Cloud Nine")
Build key: ac38e2622f (21.06-24-gb9da2db9d)
Built against Qt version: 5.15.2
Current version info is:
Rosegarden version: 24.12 ("Incredible")
Build key: 442818511e (24.06-1-gf91b7e227)
Built against Qt version: 5.15.13
How would I be "requesting the gtk widgets"? I am not aware of having,
or how I would have, done so...
I assume that my RG preferences would be applied the same to both
of the
above versions?
TIA
Chuck.
On 07/06/2024 7:14 pm, Ted Felix wrote:
Just some limited searching seems to indicate that the "gtk" widget
style was dropped around Qt 5.7 or so. Maybe you're running a newer
version of Qt and you are requesting the gtk widgets?
https://bbs.archlinux.org/viewtopic.php?id=214147&p=2
See post #36.
All the trails are from 2016/2017, though. Not sure they are
relevant today.
Ted.
On 6/7/24 6:24 AM, chuck elliot wrote:
Rosegarden won't run at the moment.
The problem seems to be something to do with "GtkWidget" (see
below).
24.06 built fine from git source.
I have recently upgraded to fedora 39 and this is probably the first
time I have tried to
run RG since. Same result on a similar/identical system.
gtk3-3.24.41-1.fc39.x86_64
gtk4-4.12.5-1.fc39.x86_64
RG output:
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add private field to invalid (non-instantiatable) type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkBuildable'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_interface_add_prerequisite: assertion 'G_TYPE_IS_INTERFACE
(interface_type)' failed
(rosegarden:725222): GLib-CRITICAL **: 11:12:47.844:
g_once_init_leave: assertion 'result != 0' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_widget_get_style_context: assertion 'GTK_IS_WIDGET (widget)'
failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_style_context_lookup_color: assertion 'GTK_IS_STYLE_CONTEXT
(context)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.868:
gtk_style_context_get_background_color: assertion
'GTK_IS_STYLE_CONTEXT (context)' failed
qt.qpa.gtk: Property name "theme_unfocused_bg_color" not found.
Falling back to QGtk3Interface::QGtkColorDefault::Background
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
add private field to invalid (non-instanti
Re: [Rosegarden-user] MIDI Instrument creation script
Looking at this now. I think I see what you are getting at. We absolutely support multiple LSB's per MSB and multiple MSB's per LSB. The problem is that the UI makes it annoying to achieve this. (In fact based on the pattern of MSB/LSB values you could end up in a very frustrating maze.) Procedure to reproduce: - Create a new device. - Create a bank with MSB:LSB 10:0. - Create another bank. - Adjust the bank MSB to 10. You will find that you cannot because the LSB is 0 and it won't let you select 10:0 again. So the MSB will jump from 9 to 11. - Adjust the LSB to 1, then you can adjust the MSB to 10. This proactive error prevention seems irritating. Maybe we should redo this as a pop-up dialog for setting MSB and LSB at the same time and checking them for dupes only on dismissal? Is this what you are seeing? Are there any other issues you've encountered with this dialog? Ted. On 5/31/24 6:49 PM, Olivier S. wrote: According to my tests, both Patch banks and Rhythm Sets work, nevertheless with a limitation that I think is a Rosegarden bug, unless you can explain where I'm wrong: it's impossible within the same device to manage both Patches and Rhythm Sets for the same MSB. Rosegarden only accepts one bank with the same LSB, even if you check the percussion box, which I think is precisely designed to manage this conflict. At this stage, my only solution is to create two devices for my JV1080, one for patches on MIDI channels 1-9/11-16 and another specific to percussion on channel 10 to manage rhythm sets. Frustrating. ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user
Re: [Rosegarden-user] Fedora and LV2
'tried a new build with lilv-devel installed but not gtk2-devel.
build failed at the link stage:
/usr/bin/ld: ../librosegardenprivate.so: undefined reference to
`Rosegarden::LV2Gtk::getInstance()'
collect2: error: ld returned 1 exit status
make[2]: *** [src/CMakeFiles/rosegarden.dir/build.make:127: rosegarden]
Error 1
make[1]: *** [CMakeFiles/Makefile2:264:
src/CMakeFiles/rosegarden.dir/all] Error 2
make: *** [Makefile:146: all] Error 2
so it looks like there is a dependency...
Regards,
Chuck.
On 12/06/2024 14:44, Ted Felix wrote:
Thanks for tracking this down. Yes, the LV2 stuff can use gtk2, but
it doesn't need it. You should be able to install lilv-devel without
gtk2-devel and that might work. Let me know how it goes if you try it.
Note: A fresh clone shouldn't be necessary. Just delete your build
directory, create a new one, switch into it and run cmake again.
Ted.
On 6/11/24 4:40 AM, chuck elliot wrote:
...but a fresh git clone of RG is now compiling without
gtk2-devel and lilv-devel...
cmake has disabled LILV support
...now it runs again.
Chuck.
On 09/06/2024 11:10 pm, Chris Fancsy wrote:
DuckDuckGo was unable to verify sender identity
Hi Chuck,
I have the same problem with the compiled 24.06 version on Fedora 40.
I compiled the 23.12 version and it runs fine.
The problem is from the LV2 plugin support which is using GTK2. I
don't think there's a way to disable LV2 plugin support at compile
time.
If I find a solution I will post it here.
Chris
On Sun, Jun 9, 2024 at 5:05 AM chuck elliot wrote:
Thanks for the link Ted but I'm afraid I didn't glean anything useful
there...
RG 21.12 runs ok on this system (lying around in /usr/local/bin)
Rosegarden version: 21.12 ("Cloud Nine")
Build key: ac38e2622f (21.06-24-gb9da2db9d)
Built against Qt version: 5.15.2
Current version info is:
Rosegarden version: 24.12 ("Incredible")
Build key: 442818511e (24.06-1-gf91b7e227)
Built against Qt version: 5.15.13
How would I be "requesting the gtk widgets"? I am not aware of having,
or how I would have, done so...
I assume that my RG preferences would be applied the same to both
of the
above versions?
TIA
Chuck.
On 07/06/2024 7:14 pm, Ted Felix wrote:
Just some limited searching seems to indicate that the "gtk" widget
style was dropped around Qt 5.7 or so. Maybe you're running a newer
version of Qt and you are requesting the gtk widgets?
https://bbs.archlinux.org/viewtopic.php?id=214147&p=2
See post #36.
All the trails are from 2016/2017, though. Not sure they are
relevant today.
Ted.
On 6/7/24 6:24 AM, chuck elliot wrote:
Rosegarden won't run at the moment.
The problem seems to be something to do with "GtkWidget" (see
below).
24.06 built fine from git source.
I have recently upgraded to fedora 39 and this is probably the first
time I have tried to
run RG since. Same result on a similar/identical system.
gtk3-3.24.41-1.fc39.x86_64
gtk4-4.12.5-1.fc39.x86_64
RG output:
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add private field to invalid (non-instantiatable) type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkBuildable'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_interface_add_prerequisite: assertion 'G_TYPE_IS_INTERFACE
(interface_type)' failed
(rosegarden:725222): GLib-CRITICAL **: 11:12:47.844:
g_once_init_leave: assertion 'result != 0' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_widget_get_style_context: assertion 'GTK_IS_WIDGET (widget)'
failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_style_context_lookup_color: assertion 'GTK_IS_STYLE_CONTEXT
(context)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.868:
gtk_style_context_get_background_color: assertion
'GTK_IS_STYLE_CONTEXT (context)' failed
qt.qpa.gtk: Property name "theme_unfocused_bg_color" not found.
Falling back to QGtk3Interface::QGtkColorDefault::Background
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
add private field to invalid (non-instantiatable) type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATAB
Re: [Rosegarden-user] Add key change [WAS] Re: Testing Appreciated for 24.06
Here's my procedure: - New Composition - Draw a four bar segment on track 1 at bar 1. - Switch to arrow (select and edit) tool. - Ctrl+Drag to Track 2, 3, 4, etc... until I have a segment on all 16 tracks. - Select all and launch notation. - First staff is selected, position pointer is at bar 1. - Repeat until all segments have key changes. - Alt+G+K, pick Bb, Ok. - Key change appears. - Hover over the segment selector wheel and mouse wheel down to the next segment. No problems at all. Does the above process work for you? Ted. On 6/5/24 1:55 AM, Lorenzo Sutton wrote: On 04/06/2024 22:04, Ted Felix wrote: I have a feeling those messages in the output are not related to the issue. I tried adding some key changes to a new document, and it worked fine. How often does the problem happen? I think I found a possible culprit / criteria to trigger this: - Create an empty segment in track 1 - Now *duplicate* that segment dragging it to track 2 - OR create another segment on another track - All these segments are at measure 1 After this trying to add a key change on those segments seems to fail. moving them around seems to 'unlock' this but randomly, as sometimes it doesn't. TIn all cases where adding a key change fails the relevant message on the terminal seems to be: Internal error in NotationTypes, Pitch::transpose() Hope this helps Lorenzo Ted. On 6/4/24 11:19 AM, Lorenzo Sutton wrote: Hi Ted, Very quick answer... Actually this was on empty segments. The idea is like by default you are in C major and already know you want to change key kind of scenario, so no notes involved. Two segments on two different tracks was the only common denominator. Sorry if this isn't too helpful. Lorenzo On Tue, 4 Jun 2024 at 17:04, Ted Felix wrote: I edited two segments and made a key change at every bar, but no luck reproducing. Probably because transposition might be involved and I had no notes. The "internal error" message, if it is relevant, seems to indicate that the transposition resulted in notes that were outside (below) the MIDI note range. Are you working with some really low notes? I think you might be able to insert the key change without transposition then transpose separately? That's probably really inconvenient, but it might help determine whether it is the key change or the transposition that is the problem. I'll keep trying to reproduce the message with some low notes and some wild key changes. As for the other messages, those may have been caused by some baseclass rearranging I did in response to warnings in the translation stuff. I sure hope not. I'll take a closer look. Ted. On 6/3/24 5:56 AM, Lorenzo Sutton wrote: Sorry for using the mailing list for this but it's on a bit of a deadline project. When trying to add a key change in notation editor nothing happens and in the terminal I get this: [NotationTool] NotationTool::findActionInParentView: Can't find ActionFileClient in parent widget hierarchy [NotationTool] NotationTool::invokeInParentView: No action " "no_accidental" " found in parent view Internal error in NotationTypes, Pitch::transpose() It seems to happen randomly... I managed to "trigger" this by editing two segments in the same notation editor window On 08/05/2024 18:01, Ted Felix wrote: Rosegarden 24.06 will be released on June 5. Please test the latest version in git as much as possible. See the release notes for areas that need testing: https://www.rosegardenmusic.com/wiki/dev:24.06 Let me know if you have any questions about the release notes and I will elaborate. If you've not built Rosegarden before, give it a shot and let us know if you run into trouble. We'll walk you through it. I've written up instructions on the wiki that should help you through the build process: https://www.rosegardenmusic.com/wiki/dev:building_rosegarden_from_source To help out with 24.06 you can either download a source snapshot, or use git to stay up to date with the latest. Thanks. Ted. ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user ___ R
Re: [Rosegarden-user] Fedora and LV2 (was: Can't launch Rosegarden)
Thanks for tracking this down. Yes, the LV2 stuff can use gtk2, but
it doesn't need it. You should be able to install lilv-devel without
gtk2-devel and that might work. Let me know how it goes if you try it.
Note: A fresh clone shouldn't be necessary. Just delete your build
directory, create a new one, switch into it and run cmake again.
Ted.
On 6/11/24 4:40 AM, chuck elliot wrote:
...but a fresh git clone of RG is now compiling without
gtk2-devel and lilv-devel...
cmake has disabled LILV support
...now it runs again.
Chuck.
On 09/06/2024 11:10 pm, Chris Fancsy wrote:
DuckDuckGo was unable to verify sender identity
Hi Chuck,
I have the same problem with the compiled 24.06 version on Fedora 40.
I compiled the 23.12 version and it runs fine.
The problem is from the LV2 plugin support which is using GTK2. I
don't think there's a way to disable LV2 plugin support at compile
time.
If I find a solution I will post it here.
Chris
On Sun, Jun 9, 2024 at 5:05 AM chuck elliot wrote:
Thanks for the link Ted but I'm afraid I didn't glean anything useful
there...
RG 21.12 runs ok on this system (lying around in /usr/local/bin)
Rosegarden version: 21.12 ("Cloud Nine")
Build key: ac38e2622f (21.06-24-gb9da2db9d)
Built against Qt version: 5.15.2
Current version info is:
Rosegarden version: 24.12 ("Incredible")
Build key: 442818511e (24.06-1-gf91b7e227)
Built against Qt version: 5.15.13
How would I be "requesting the gtk widgets"? I am not aware of having,
or how I would have, done so...
I assume that my RG preferences would be applied the same to both of the
above versions?
TIA
Chuck.
On 07/06/2024 7:14 pm, Ted Felix wrote:
Just some limited searching seems to indicate that the "gtk" widget
style was dropped around Qt 5.7 or so. Maybe you're running a newer
version of Qt and you are requesting the gtk widgets?
https://bbs.archlinux.org/viewtopic.php?id=214147&p=2
See post #36.
All the trails are from 2016/2017, though. Not sure they are
relevant today.
Ted.
On 6/7/24 6:24 AM, chuck elliot wrote:
Rosegarden won't run at the moment.
The problem seems to be something to do with "GtkWidget" (see below).
24.06 built fine from git source.
I have recently upgraded to fedora 39 and this is probably the first
time I have tried to
run RG since. Same result on a similar/identical system.
gtk3-3.24.41-1.fc39.x86_64
gtk4-4.12.5-1.fc39.x86_64
RG output:
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
add private field to invalid (non-instantiatable) type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844: cannot
register existing type 'GtkBuildable'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_interface_add_prerequisite: assertion 'G_TYPE_IS_INTERFACE
(interface_type)' failed
(rosegarden:725222): GLib-CRITICAL **: 11:12:47.844:
g_once_init_leave: assertion 'result != 0' failed
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.844:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_widget_get_style_context: assertion 'GTK_IS_WIDGET (widget)'
failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.859:
gtk_style_context_lookup_color: assertion 'GTK_IS_STYLE_CONTEXT
(context)' failed
(rosegarden:725222): Gtk-CRITICAL **: 11:12:47.868:
gtk_style_context_get_background_color: assertion
'GTK_IS_STYLE_CONTEXT (context)' failed
qt.qpa.gtk: Property name "theme_unfocused_bg_color" not found.
Falling back to QGtk3Interface::QGtkColorDefault::Background
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
register existing type 'GtkWidget'
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
add class private field to invalid type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868: cannot
add private field to invalid (non-instantiatable) type ''
(rosegarden:725222): GLib-GObject-CRITICAL **: 11:12:47.868:
g_type_add_interface_static: assertion 'G_TYPE_IS_INSTANTIATABLE
(instance_type)' failed
-DCMAKE_BUILD_TYPE=Debug-DCMAKE_BUILD_TYPE=Debug-DCMAKE_BUILD_TYPE=Debugqt.qpa.events.reader:
[heap] 0
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to
unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user
___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to
unsubscribe
https://lists.sourceforge.net/lists/listinfo/roseg
