Re: [Rosegarden-user] Controlling Rosegarden by a MIDI-Keyboard

2020-09-27 Thread Ted Felix

On 8/11/20 5:15 AM, Werner Flügel wrote:
Now I want to control Rosegarden with the 
start/Stop/Rewind-Buttons on the Keyboard. Is this possible?


  It is now.  For the most part.  I've implement 
play/stop/record/previous track/next track.  See [feature-requests:#423]:


https://sourceforge.net/p/rosegarden/feature-requests/423/

  If you can get the latest svn, build, and test, I would really 
appreciate it.  Thanks.


  Fast-Forward and Rewind are on the todo for this week.

Ted.


___
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] [Solved] CPU load rises when Rosegarden is idle

2020-08-07 Thread Ted Felix
> I thought this is the new auto composition number one hit producing 
tool (ACNOHPT) :-)


  I was wondering when this would cause trouble.  I didn't even notice 
it.  We now avoid connecting to thru if we've found no other connections.


  Should be fixed in [r15906].  Please test latest svn.  Thanks.

Ted.

On 8/7/20 9:44 AM, MST wrote:

Ted,
I see. That makes sense because another observation I made were the 
never ending midi notes I recorded with the virtual keyboard.
I thought this is the new auto composition number one hit producing tool 
(ACNOHPT) :-)

Thank you and greetings,
Michael
*Gesendet:* Freitag, 07. August 2020 um 14:35 Uhr
*Von:* "Ted Felix" 
*An:* rosegarden-user@lists.sourceforge.net
*Betreff:* Re: [Rosegarden-user] [Solved] CPU load rises when Rosegarden 
is idle

Turns out it's because out is connected to MIDI Thru and in is also
connected to MIDI Thru. This ends up creating a feedback loop.

I will make the connection algorithm avoid the local MIDI Thru ports.

Ted.

On 8/6/20 8:19 PM, Ted Felix wrote:
 > On 8/6/20 5:16 PM, MST wrote:
 >> My "unfavorable" settings in the recording parameter menu (all midi
 >> input devices, all channels, midi through auto routing) are causing a
 >> cpu load of 104% according to "top" - didn't know that is possible :-)
 >> Choosing only one midi input device, one fixed midi channel and midi
 >> auto routing to off solves this problem.
 >
 >   This is interesting.  It appears as if it is the Thru Routing setting
 > that needs to be set to Off.  Then the CPU usage goes down and the Thru
 > Routing can be changed back to Auto and the CPU usage will not go 
back up.

 >
 >   I need to take a closer look at this one.  Thanks for catching it and
 > letting me know.
 >
 > 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


Re: [Rosegarden-user] [Solved] CPU load rises when Rosegarden is idle

2020-08-07 Thread Ted Felix
  Turns out it's because out is connected to MIDI Thru and in is also 
connected to MIDI Thru.  This ends up creating a feedback loop.


  I will make the connection algorithm avoid the local MIDI Thru ports.

Ted.

On 8/6/20 8:19 PM, Ted Felix wrote:

On 8/6/20 5:16 PM, MST wrote:
My "unfavorable" settings in the recording parameter menu (all midi 
input devices, all channels, midi through auto routing) are causing a 
cpu load of 104% according to "top" - didn't know that is possible :-)
Choosing only one midi input device, one fixed midi channel and midi 
auto routing to off solves this problem.


   This is interesting.  It appears as if it is the Thru Routing setting 
that needs to be set to Off.  Then the CPU usage goes down and the Thru 
Routing can be changed back to Auto and the CPU usage will not go back up.


   I need to take a closer look at this one.  Thanks for catching it and 
letting me know.


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


Re: [Rosegarden-user] [Solved] CPU load rises when Rosegarden is idle

2020-08-06 Thread Ted Felix

On 8/6/20 5:16 PM, MST wrote:

My "unfavorable" settings in the recording parameter menu (all midi input devices, all 
channels, midi through auto routing) are causing a cpu load of 104% according to "top" - 
didn't know that is possible :-)
Choosing only one midi input device, one fixed midi channel and midi auto 
routing to off solves this problem.


  This is interesting.  It appears as if it is the Thru Routing setting 
that needs to be set to Off.  Then the CPU usage goes down and the Thru 
Routing can be changed back to Auto and the CPU usage will not go back up.


  I need to take a closer look at this one.  Thanks for catching it and 
letting me know.


Ted.


___
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] Repeatable crash :(

2020-08-04 Thread Ted Felix
  We've made some very significant changes to the way we discover and 
connect to synths.  You're probably running into a problem there.


  Run your usual test that is failing, then in the preferences, click 
on the Details... button and copy/paste the text to an email for me. 
That should help me figure out what we need to change to accommodate 
your particular setup.


  In the meantime, to avoid the connection annoyances, you might want 
to consider going back to [r15892].


Thanks.
Ted.

On 8/4/20 3:33 PM, Will Godfrey wrote:

Bit slow getting back to this. It seems fine now thanks.

I'm actually running build 15904. I don't know if I'm missing something but
Rosegarden no longer seems to connect to Yoshimi, if Yoshimi is already
running.
If Rosegarden is started first, then the connection is made.

On Sat, 1 Aug 2020 13:35:50 -0400 Ted Felix  wrote:


   Ok, should be fixed in [r15899].  I'm not too happy as I'm not 100%
sure what is going on to cause this problem.  It appears that a
mysterious refresh is invalidating the item pointer.  Anyway, should be
solid enough.

Ted.

On 8/1/20 9:33 AM, Ted Felix wrote:

On 8/1/20 2:41 AM, Will Godfrey wrote:

However one thing that's different is that when getting the crash I
opened the
segment in the first bar and left the entry of the initial CC at zero.
then
when changing time just incrementing it with the arrows so it's steps
of 60.


    Got it.  It is crashing for me now.  Specifically the issue is most
noticeable in EventView::slotPopupEventEditor() where the Segment
pointer becomes NULL before the creation of the EventEditCommand which
then tries to use the null Segment pointer.

    I will continue digging.

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


Re: [Rosegarden-user] CPU load rises when Rosegarden is idle

2020-08-03 Thread Ted Felix

On 8/3/20 3:22 PM, MST wrote:
Since the latest commitments (code version 1589x - 15904) the cpu load 
rises strongly (in my case 45-60% and more) when Rosegarden is idle and 
in QJackCtl a lot of xruns are generated. After loading a song into 
Rosegarden and starting to play this song the CPU load falls again 
(maybe 5%) and no more xruns are generated.


  Try using the "top" command (at the command line) to make sure 
Rosegarden is to blame.  It will show you the processes that are using 
the most CPU.


Ted.


___
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] Track colour options have reset -- I am only seeing "Default"

2020-08-03 Thread Ted Felix

On 8/3/20 7:59 AM, Hugo van Galen wrote:
When wanting to choose a different track colour, the dropdown only shows 
the "Default" option. Did I screw up my build somehow?


  The colors are stored in the .rg file.  The base set is in your 
autoload.rg file.  You will find that here:


~/.local/share/rosegarden/autoload/autoload.rg

  If you create a new document and the colors are gone, then your 
autoload.rg is missing the colors.


  If this is happening in a specific document, then the colors are 
missing from that document.


  It's the  section in the .rg file and you can copy it 
from a known good .rg file into the broken one and colors should be 
restored.


I did try a `make clean all` but the problem persists. It does seem that 
`build/src/qrc_data.cpp` includes a reference to 
`colourmap-fragment.xml` so I'm a bit confused as to why I don't see the 
options.


  The colors are in autoload.rg and nowhere else.

According to the documentation, I should also have an "Add New Colour" 
option in this dropdown, but that isn't there either.


  That was never implemented AFAICT.  I removed what little code was 
there related to this.  If somebody wants to re-implement, be my guest. 
I think we have enough colors as it is.  And worst case, you can always 
hack together your own set in your autoload.rg.


Ted.


___
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] New ALSA port matching algorithm

2020-08-01 Thread Ted Felix

On 8/1/20 3:16 AM, Malhaire Christophe wrote:

It works perfectly (see enclosed file), well done.


  Glad to hear it.  Hopefully everyone else is happy with it as well. 
Haven't heard any complaints yet.


I think it's better this way. Thus, if for one reason or another a 
device is not correctly detected (because it is switched off, or because 
a usb cable is defective ...) the corresponding track will not send data 
anywhere ("no connection" is preferable to "wrong connection").


  Agreed.


-
"40:0 Streichfett MIDI 1 (duplex)"
"40:0 Streichfett MIDI 1 (duplex)"
   Final score: 75
-
"36:0 XV-5050 MIDI 1 (duplex)"
"32:0 XV-5050 MIDI 1 (duplex)"
   Final score: 75

I wonder why the "Final score" indicator has the same value (75) in both 
cases?


I would have thought that the score would be higher in the first case 
because the objective/source strings are equal, right?While in the 
second case we have 36 different from 32.


  The client numbers (36 and 32) are ignored.  This was one of the 
reasons the old algorithm would get so confused.  It would give very 
high priority to the client number.  That's really not a good idea as 
the client numbers are pretty random.


  The client numbers are still used to make sure the device class 
(internal/hardware/software) matches, and if they don't, the comparison 
isn't even done.  That's one reason why you might see some with no score 
at all.


  The code's pretty easy to read now.  See 
AlsaDriver::setPlausibleConnection():


https://sourceforge.net/p/rosegarden/code/15895/tree/trunk/rosegarden/src/sound/AlsaDriver.cpp#l1274

  Comments appreciated.

  Thanks for helping with this.  Let me know if you run into anything else.

Ted.


___
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] Repeatable crash :(

2020-08-01 Thread Ted Felix
  Ok, should be fixed in [r15899].  I'm not too happy as I'm not 100% 
sure what is going on to cause this problem.  It appears that a 
mysterious refresh is invalidating the item pointer.  Anyway, should be 
solid enough.


Ted.

On 8/1/20 9:33 AM, Ted Felix wrote:

On 8/1/20 2:41 AM, Will Godfrey wrote:
However one thing that's different is that when getting the crash I 
opened the
segment in the first bar and left the entry of the initial CC at zero. 
then
when changing time just incrementing it with the arrows so it's steps 
of 60.


   Got it.  It is crashing for me now.  Specifically the issue is most 
noticeable in EventView::slotPopupEventEditor() where the Segment 
pointer becomes NULL before the creation of the EventEditCommand which 
then tries to use the null Segment pointer.


   I will continue digging.

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


Re: [Rosegarden-user] Repeatable crash :(

2020-08-01 Thread Ted Felix

On 8/1/20 2:41 AM, Will Godfrey wrote:

However one thing that's different is that when getting the crash I opened the
segment in the first bar and left the entry of the initial CC at zero. then
when changing time just incrementing it with the arrows so it's steps of 60.


  Got it.  It is crashing for me now.  Specifically the issue is most 
noticeable in EventView::slotPopupEventEditor() where the Segment 
pointer becomes NULL before the creation of the EventEditCommand which 
then tries to use the null Segment pointer.


  I will continue digging.

Ted.


___
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] Build oddity

2020-07-31 Thread Ted Felix

  Just pushed the fix as [r15896].

Ted.

On 7/31/20 1:33 PM, Will Godfrey wrote:

Thanks for getting back to me so quickly.
NULL gave the same result, but item.reset() worked fine.
Sorry I took so long letting you know but my 'office' is right under the roof
and it became way too hot to even walk in the room today :(

Will.

On Fri, 31 Jul 2020 08:35:24 -0400
Ted Felix  wrote:


   The line in question is this:

   item = nullptr;

   It's not the compiler, it's the version of Qt you are using.  Prior
to 5.8, QSharedPointer didn't handle nullptr.

   Removing that line will simply leak memory.

   It can be fixed by changing it to NULL, or doing it the shared_ptr way:

   item.reset();

   I will fix this for the future.

   See also [r15476].

Ted.

On 7/31/20 7:01 AM, Will Godfrey wrote:

I have two machines, the older one with gcc 6.3.0 and the newer one running
gcc 8.3.0.

On the newer one rosegarden build 15895 compiles without issues, but on the
older one I get a mismatch type failure at:
rosegarden-15895/src/gui/rulers/PropertyControlRuler.cpp line 265

If I comment this line out the build completes and the image *seems* to run
OK, but I've no idea of the implications.
   



___
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


Re: [Rosegarden-user] Repeatable crash :(

2020-07-31 Thread Ted Felix
  I can't reproduce this.  Perhaps my procedure is incorrect?  Here's 
what I'm doing:


- Launch rg.
- Left-click to drop a segment.
- Press "E" to launch the event editor.
- Press "I" to insert an event.
- Event Type: Controller
- Absolute Time: 7680
- Controller Number: 7
- Controller Value: 100
- Ok
- Begin Loop
  - Right-click the event and select "Open in event editor".
  - Add a little to Time, Number, and Value.
  - Ok
- End Loop, repeat several times

  Am I doing something wrong?

Ted.

On 7/31/20 4:57 PM, Will Godfrey wrote:

I seem to have identified this while using Rosegarden to send test CCs to
Yoshimi.

Open a new Rosegarden document and display the matrix editor.

Create an empty track segment (I just right click on a track).

Using the event editor create a new controller and fill in any values.

Keep changing these including time, without saving, and after 5 or 6 changes
Rosegarden exits!

If you save the file after each edit, it never crashes.




___
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] Build oddity

2020-07-31 Thread Ted Felix

On 7/31/20 8:35 AM, Ted Felix wrote:

   Removing that line will simply leak memory.


  Actually, it will not leak memory.  It appears as if it will always 
select the last item.  Whatever it might do, it's easily fixed with reset().


Ted.


___
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] Build oddity

2020-07-31 Thread Ted Felix

  The line in question is this:

  item = nullptr;

  It's not the compiler, it's the version of Qt you are using.  Prior 
to 5.8, QSharedPointer didn't handle nullptr.


  Removing that line will simply leak memory.

  It can be fixed by changing it to NULL, or doing it the shared_ptr way:

  item.reset();

  I will fix this for the future.

  See also [r15476].

Ted.

On 7/31/20 7:01 AM, Will Godfrey wrote:

I have two machines, the older one with gcc 6.3.0 and the newer one running
gcc 8.3.0.

On the newer one rosegarden build 15895 compiles without issues, but on the
older one I get a mismatch type failure at:
rosegarden-15895/src/gui/rulers/PropertyControlRuler.cpp line 265

If I comment this line out the build completes and the image *seems* to run
OK, but I've no idea of the implications.




___
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] extended quantization

2020-07-30 Thread Ted Felix

On 7/30/20 12:29 PM, chuck via Rosegarden-user wrote:

I am using v19.12 package in Fedora 32 until I get a chance to compile
the latest version.


  You'll probably want to do that and try again.  Might have been fixed 
in the meantime.



The fixed triplet quantizations seem to be working fine but when I
tried to get a quintuplet by using an 'arbitrary grid unit' setting
of 96, RG crashed.


  So that's 5 in the time of a quaver?  I just played with it a little 
bit, but I'm not seeing any crashes.  It's working fine.  Might also be 
a problem with the specific notes you are quantizing.  If you can send a 
small composition with just the notes in question, that might help me to 
reproduce this.



Apparently an automatic bug report was sent.


  Yeah, but they won't let me see those.  Something about sensitive 
private information.  藍



'happy to pursue this if required.


  Try building a newer version.  If you can reproduce the crash there, 
send me an example to quantize.  Worst-case, we can try gdb on your 
machine and see if that provides any clues.


  This is feature request #477.  Let's move the conversation there if 
that's ok with you.


https://sourceforge.net/p/rosegarden/feature-requests/477/

Ted.


___
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] Slow grace note (acciaccatura)

2020-07-30 Thread Ted Felix

On 7/27/20 1:08 PM, Davide Castellano wrote:

I checked on the matrix view and the note length is half than 1/64 box.
When I hear the grace note, it's like 1/8 duration.

Is there a way to solve it?


  Currently, Rosegarden only supports short grace notes (1/4 the 
duration of the main note).  And grace notes are a special case that 
appear very small in the Matrix editor, but the actual play time is longer.


  We would need to add code and some sort of UI to support longer grace 
notes.


  From SegmentPerformanceHelper::getGraceNoteTimeAndDuration():

// For our present purposes, we will assume that grace notes start
// at the same time as their host note was intended to, and
// "steal" a proportion of the duration of their host note.  This
// causes the host note to start later, and be shorter, by that
// same proportion.

[...]

// To begin with we will set the proportion to 1/4, but we will
// probably want it to be (a) something different [because I don't
// really know what I'm doing], (b) adaptive [e.g. shorter host
// note or more grace notes = longer proportion], (c)
// configurable, or (d) all of the above.

Ted.


___
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] New ALSA port matching algorithm

2020-07-29 Thread Ted Felix

On 7/29/20 7:16 PM, Ted Felix wrote:

   Let me try adding the <= 20 check and see how that works for you.


  Just committed this as [r15895].  Please test the latest svn.

Ted.


___
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] New ALSA port matching algorithm

2020-07-29 Thread Ted Felix

On 7/29/20 4:16 PM, Malhaire Christophe wrote:
Maybe if a piece of hardware is turned off or not connected, maybe the 
program shouldn't try to "force" connect it to a port that doesn't match?


  I don't think there's a lot we can really do if things aren't turned 
on.  That's going to lead to confusion, and I don't think there's a way 
around it.


  However, yeah, we could give up based on some threshold with the name 
matching.  E.g. a score of 20 or less will not match.  Then it is fuzzy 
up to a point.  Or we could even just get rid of the fuzziness.  It's 
not really very useful, honestly.  Some soft synths will put their 
process ID in the name, but you can always override it to get a more 
stable setup.  That's what I do with Fluidsynth.


  Let me try adding the <= 20 check and see how that works for you.

Ted.


___
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] New ALSA port matching algorithm

2020-07-29 Thread Ted Felix

  Just pushed [r15894] to svn which should provide the requested behavior.

Ted.

On 7/29/20 7:27 AM, Ted Felix wrote:

On 7/29/20 6:10 AM, Malhaire Christophe wrote:
If the Waldorf Streichfett is turned off, it is connected to "14:0 
Midi Through Port-0 (duplex)" instead of "no port"


   The previous version would completely refuse to connect to a 
mismatched client class.  In this case a hardware client (Streichfett) 
connected to an internal port (MIDI through).


   This behavior can easily be added.

When both XV-5050s are turned on, one is connected to “32: 0 XV-5050 
MIDI 1 (duplex)” and the second is connected to “40: 0 XV-5050 MIDI 1 
(duplex)”. Good. But when the second is off, it is connected to "32: 0 
XV-5050 MIDI 1 (duplex)" like the other.


   The previous version would refuse to connect to an already connected 
playback port.  I think I meant to include that, but forgot.  I will put 
that in.


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


Re: [Rosegarden-user] New ALSA port matching algorithm

2020-07-29 Thread Ted Felix

On 7/29/20 6:10 AM, Malhaire Christophe wrote:
If the Waldorf Streichfett is turned off, it is connected to "14:0 Midi 
Through Port-0 (duplex)" instead of "no port"


  The previous version would completely refuse to connect to a 
mismatched client class.  In this case a hardware client (Streichfett) 
connected to an internal port (MIDI through).


  This behavior can easily be added.

When both XV-5050s are turned on, one is connected to “32: 0 XV-5050 
MIDI 1 (duplex)” and the second is connected to “40: 0 XV-5050 MIDI 1 
(duplex)”. Good. But when the second is off, it is connected to "32: 0 
XV-5050 MIDI 1 (duplex)" like the other.


  The previous version would refuse to connect to an already connected 
playback port.  I think I meant to include that, but forgot.  I will put 
that in.


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] New ALSA port matching algorithm

2020-07-27 Thread Ted Felix
  I've implemented a new/simpler ALSA port matching algorithm 
(AlsaDriver::setPlausibleConnection()).  Would appreciate some testing 
of the latest svn.  [r15893] is the commit in question.


  The new algorithm does a fuzzy string compare on the name, then 
checks the client class, port, and availability to compute a match 
score.  The port with the highest score wins and the connection is made.


  There is also a small improvement in handling of blank port names. 
We now check playback and record ports separately.  This means in some 
situations where we might not have connected a record port to a 
keyboard, we will now do that.


  So, when you load a composition, we should do a slightly better job 
of restoring connections.  At the very least, this should be as good as 
it was.


  Anyway, beat it up and let me know if you run into any trouble.  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


Re: [Rosegarden-user] Add Control Ruler not working as expected in Matrix Editor and a few other questions

2020-07-22 Thread Ted Felix

On 7/22/20 5:56 AM, Hugo van Galen wrote:

Should I [...] add a comment to bug #1543


  I've copied over the fix and added a comment to bug #1543.  Please 
test the latest svn and respond in the bug report with the results.  Thanks.


https://sourceforge.net/p/rosegarden/bugs/1543/

Ted.


___
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] Add Control Ruler not working as expected in Matrix Editor and a few other questions

2020-07-22 Thread Ted Felix

On 7/22/20 5:56 AM, Hugo van Galen wrote:

I looked into NotationWidget.cpp to see the following comment and code:

     // FIX #1543: If name happens to come to us with an & included 
somewhere,
     // strip the & so the string will match the one we are comparing 
later on.

     //
     name.replace(QRegExp("&"), "");

I can confirm that adding this code to the MatrixWidget.cpp's 
addControlRuler() fixes the issue.


  Thanks for tracking this down.  I will get it in there right now.

Ted.


___
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] Questions about audio playback and MIDI connections

2020-07-04 Thread Ted Felix

On 6/30/20 1:11 PM, Malhaire Christophe wrote:
Thanks a lot for your answer, here is the content of the Details window. 


  Ok, I've added comments to the code and improved the logging in the 
details window.  With this we should be able to start making some 
improvements to the matching algorithm.  Need you to do the following if 
you can:


1. I've re-opened bug #1572 which is this exact problem.  Can we move 
the discussion there to avoid cluttering the user mailing list?  You'll 
need to create a sourceforge account if you don't have one.


  https://sourceforge.net/p/rosegarden/bugs/1572/

2. Get the latest code from svn, build, and re-run your test.  Send me 
the contents of the Details window directly, or post them to bug #1572.


  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


Re: [Rosegarden-user] Questions about audio playback and MIDI connections

2020-06-30 Thread Ted Felix

On 6/30/20 1:11 PM, Malhaire Christophe wrote:
AlsaDriver::setPlausibleConnection: connection like "24:0 Unitor8 MIDI 1 
(duplex)" requested for device 2
AlsaDriver::setPlausibleConnection: fuzzy match 24:0 XV-5050 MIDI 1 
(duplex) available with fitness 1


  Interesting.  So it looks like we need to more heavily weight the 
name portion and downplay the device/port numbers.  Especially, it 
seems, the device numbers (24 in this case) which can maybe be 
completely ignored since they are completely unpredictable.


  This looks like it might be really easy to fix.  I will have a closer 
look.  Hopefully you can hang around for a couple of weeks and help me 
with the testing.  I should be back to working on rg this weekend.


Ted.


___
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] Questions about audio playback and MIDI connections

2020-06-30 Thread Ted Felix

On 6/30/20 5:20 AM, Malhaire Christophe wrote:
1) While using Rosegarden, I cannot play a video on Youtube, or play a 
piece of music using Parole Media Player.


  Off the top of my head, it sounds like those programs likely use 
ALSA, but you have JACK running.  When JACK is running, it takes over 
ALSA and anything trying to use ALSA will be "broken".


  There may be a way to get ALSA and JACK apps to cooperate.  I seem to 
remember an article on the JACK website.  This is a common complaint.


2) Each time the PC and Rosegarden are restarted, I have to check and 
change my MIDI connections because the MIDI ports of my external USB 
sound modules are systematically inconsistent (MIDI Playback devices 
related to MIDI Outputs).


  This is interesting as we've had this complaint before.  I really 
would like to work on fixing this, if possible, but the other user 
reported that the problem went away.


  To get a closer look at what rg is doing to make those connections, 
load a composition, then go to Edit > Preferences... and press the 
"Details..." button next to "Sequencer Status".  This should show you 
the connections it is trying to make.  If you can post what's in there, 
that might help with figuring out what we might be able to do to fix this.


I believe Rosegarden has an internal function 
for "smartly" attaching a MIDI device, according to its name, but it 
doesn't work for me. Surely I must miss something but what?


  It's all automated, so not really anything you can do.  We need to 
have a closer look and see if anything can be done at all.



Is there anything to add to the .rgd file?


  .rgd files are only relevant when you create a new device.  It is the 
.rg file that is driving things from that point on.  It's the exact same 
info, though.


But since the port number can change at each boot, or depending on the 
order in which the devices are turned on, maybe this is not a good method?


  I suspect that is what is really happening.  There's a limit to what 
we can obtain from the drivers.  Therefore there is a limit as to how 
good of a job we can do matching things up.


3) And by the way, where can we find a complete manual about the xml 
format of .rgd files?


  You'll have to read the code.  We've barely got time to write code 
let alone documentation.  Feel free to create some, though.  We'd be 
happy to post it to the wiki.


Ted.


___
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] Add Control Ruler not working as expected in Matrix Editor and a few other questions

2020-06-30 Thread Ted Felix

On 6/29/20 4:41 AM, Hugo van Galen wrote:
I have been using Rosegarden for about a decade now and have been able 
to figure it out using search engines. Lately I ran into an issue that I 
don't remember having before: I noticed that, when in the Matrix Editor, 
the option "View / Rulers / Add Control Ruler / Volume Controller" 
doesn't do anything; yet when in the "Notation Editor" this works fine.


  I can't reproduce this.  But the code does look like it needs some 
review.  The issue will likely become apparent upon cleaning up and 
syncing the relevant routines.


It seems that any other controller except Velocity and Pitch bend only 
work in the Notation Editor.


  Those are handled separately, so it isn't surprising they are not 
affected.


Is there an underlying logical reason for 
this to behave this way -- or is this a bug?


  Sounds like you've found a legitimate bug.  The key routines in 
question are:


MatrixWidget::addControlRuler()
NotationWidget::slotAddControlRuler()

  There are differences that need to be resolved between those two and 
that might clear up the issue.  I don't have time at the moment, but 
should be back to Rosegarden this weekend.  I will have a look then.


- I also wonder whether it is already possible to "remember" the zoom 
options for any given segment as I always find myself tweaking 
these, the default zoom doesn't work for me here. Or am I overlooking 
something obvious in the settings? (Ideally, it would be great if all 
"view" options could be memorised, so when opening the same segment 
later, all zoom options, and controller rulers could be restored.)


  This is very likely an oversight.  I recommend opening a feature 
request on sourceforge:


https://sourceforge.net/p/rosegarden/feature-requests/


- I wonder how the Figuration functionality actually works


  Not my area of expertise.  Somebody here can probably fill in the 
blanks, though.  Then we really should update the documentation based on 
what was found.


Ted.


___
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] Building error....?

2020-06-15 Thread Ted Felix

  Just pushed the fix as r15845.  Latest svn should build fine.

Ted.

On 6/16/20 12:17 AM, Ted Felix wrote:

   Please apply patch #72 to fix this:

https://sourceforge.net/p/rosegarden/patches/72/

Ted.

On 6/15/20 12:02 PM, Mario Moles wrote:
error: aggregate ‘QPainterPath path’ has incomplete type and cannot be 
defined

   138 | QPainterPath path;
   |  ^~~~



___
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] Building error....?

2020-06-15 Thread Ted Felix

  Please apply patch #72 to fix this:

https://sourceforge.net/p/rosegarden/patches/72/

Ted.

On 6/15/20 12:02 PM, Mario Moles wrote:

-- The following features have been enabled:

  * ALSA, Alsa library (Advanced Linux Sound Architecture), used for 
MIDI support

  * SNDFILE, Better support for WAV files
  * JACK, Library for accessing the JACK server (http://jackaudio.org).
  * LIRCCLIENT, The LIRC client library, for remote control support

-- The following REQUIRED packages have been found:

  * Qt5Core
  * Qt5Gui
  * Qt5Widgets
  * Qt5Xml
  * Qt5Network
  * Qt5PrintSupport
  * Qt5Test
  * Qt5 (required version >= 5.1.0)
  * ZLIB
  * PkgConfig
  * X11

-- Not building unit tests, using a static library for rosegarden
-- Configuring done
-- Generating done
-- Build files have been written to: /home/mario/Documenti/rosegarden/build
[  1%] Generating svn_header
-- Found Subversion: /usr/bin/svn (found version "1.14.0")
-- SVN revision 15844
-- Build key 53d3c38410
[  1%] Built target svnheader
[  1%] Automatic MOC for target rosegardenprivate
[  1%] Built target rosegardenprivate_autogen
[  1%] Building CXX object 
src/CMakeFiles/rosegardenprivate.dir/gui/widgets/Panner.cpp.o
/home/mario/Documenti/rosegarden/src/gui/widgets/Panner.cpp: In member 
function ‘virtual void Rosegarden::Panner::paintEvent(QPaintEvent*)’:
/home/mario/Documenti/rosegarden/src/gui/widgets/Panner.cpp:138:18: 
error: aggregate ‘QPainterPath path’ has incomplete type and cannot be 
defined

   138 | QPainterPath path;
   |  ^~~~
make[2]: *** [src/CMakeFiles/rosegardenprivate.dir/build.make:2417: 
src/CMakeFiles/rosegardenprivate.dir/gui/widgets/Panner.cpp.o] Errore 1
make[1]: *** [CMakeFiles/Makefile2:197: 
src/CMakeFiles/rosegardenprivate.dir/all] Errore 2

make: *** [Makefile:161: all] Errore 2
[  1%] Generating svn_header
-- Found Subversion: /usr/bin/svn (found version "1.14.0")
-- SVN revision 15844
-- Build key 53d3c38410
[  1%] Built target svnheader
[  1%] Automatic MOC for target rosegardenprivate
[  1%] Built target rosegardenprivate_autogen
[  1%] Building CXX object 
src/CMakeFiles/rosegardenprivate.dir/gui/widgets/Panner.cpp.o
/home/mario/Documenti/rosegarden/src/gui/widgets/Panner.cpp: In member 
function ‘virtual void Rosegarden::Panner::paintEvent(QPaintEvent*)’:
/home/mario/Documenti/rosegarden/src/gui/widgets/Panner.cpp:138:18: 
error: aggregate ‘QPainterPath path’ has incomplete type and cannot be 
defined

   138 | QPainterPath path;
   |  ^~~~
make[2]: *** [src/CMakeFiles/rosegardenprivate.dir/build.make:2417: 
src/CMakeFiles/rosegardenprivate.dir/gui/widgets/Panner.cpp.o] Errore 1
make[1]: *** [CMakeFiles/Makefile2:197: 
src/CMakeFiles/rosegardenprivate.dir/all] Errore 2

make: *** [Makefile:161: all] Errore 2



___
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


Re: [Rosegarden-user] Shorten default MIDI event duration for drum notes inserted from the piano roll.

2020-06-14 Thread Ted Felix

On 6/14/20 9:02 AM, Ted Felix wrote:
   I guess for those that don't use notation, we could shorten the notes 
to make working with the editors a little easier.  Maybe we should offer 
the ability to switch back and forth.


  Or even better, implement a percussion mode for the velocity ruler. 
That way it will show the notes as short for percussion like the matrix 
editor does.


Ted.


___
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] Shorten default MIDI event duration for drum notes inserted from the piano roll.

2020-06-14 Thread Ted Felix

On 6/13/20 10:40 PM, Ted Felix wrote:
   Before I push this, does anyone have any ideas why there might be 
such a complicated algorithm for computing the end of a percussion note? 


  I think I've figured it out.  By making the notes last up until the 
next note or the bar, the notation comes out looking correct.  If we 
were to make all percussion notes 64th notes, the notation becomes an 
unreadable pile of rests.


  I guess for those that don't use notation, we could shorten the notes 
to make working with the editors a little easier.  Maybe we should offer 
the ability to switch back and forth.


Ted.


___
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] Shorten default MIDI event duration for drum notes inserted from the piano roll.

2020-06-13 Thread Ted Felix

On 6/12/20 3:29 PM, Arnaldo Pirrone wrote:
It looks like rosegarden sets the duration of the notes manually 
inserted from the piano roll into the drum track to 3840. This causes 
the velocity ruler to appear clogged up like you can see in the attached 
picture.


  Actually, the code to do this is more complicated than just 3840.  It 
adjusts the size to end at the end of the bar, or at the beginning of 
the next note(s) of the same pitch.  I have no idea why.  My best guess 
is that it is trying to make all percussion notes contiguous to a 
certain extent.  No drum synth that I've seen cares about note-offs, though.


  I modified the relevant part of the code to go with 64th notes and it 
seems to work fine.  Even open/closed hat work exactly as expected.


  Before I push this, does anyone have any ideas why there might be 
such a complicated algorithm for computing the end of a percussion note? 
 The relevant code is here if you want to have a look:


  MatrixPercussionInsertionCommand::getEndTime()

  Warning: There are no comments.

Ted.


___
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] Shorten default MIDI event duration for drum notes inserted from the piano roll.

2020-06-12 Thread Ted Felix

On 6/12/20 3:29 PM, Arnaldo Pirrone wrote:
It looks like rosegarden sets the duration of the notes manually 
inserted from the piano roll into the drum track to 3840. This causes 
the velocity ruler to appear clogged up like you can see in the attached 
picture.


  That is rather annoying and should be easy to fix.  I'll add it to my 
todo.


Feature request: can we have a way to selectively mute notes into the 
piano roll/sheet without actually deleting them? (LPX have a similar 
function)


  Would this be a checkbox to the left to mute a specific note? 
Probably doable.  Can you add a feature request to the tracker:


https://sourceforge.net/p/rosegarden/feature-requests/


Any way to fix a drum track with wrong durations?


  While you can select all the notes in a Segment, I don't see an easy 
way to set the duration for all of them.  Sounds like an idea for a 
feature.  Or at least an improvement in the way editing is handled when 
multiple events are selected.


Have you ever considered moving the development of the rosegarden 
project to GitLab/GitHub?


  Feel free to fork from here:

https://github.com/tedfelix/rosegarden

Ted.


___
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] Controlling play/stop/rewind from external DAW (Ardour)

2020-06-06 Thread Ted Felix

On 6/6/20 12:17 PM, stefano franchi wrote:
2. In Rosegarden's Midi Sync preferences, selected "accept Start, Stop, 
and Continue", and set to "slave" as MIDI MMC and MIDI TCM


  Both set to slave seems reasonable.  I've not done any of this 
before, though.


2. in Jack2 (actually, Cadence's Catia GUI for jack2) connect the midi 
track input port to Rosegarden's "out 1 - General MIDI Device" port


  So Ardour is recording rg's output?  Should be ok if that's what you 
want to do.


3.in  Jack2, connected Ardour's midi track out to 
Rosegarden's "External Controller" in port


  This might be the problem.  I'm pretty sure rg doesn't respond to MMC 
and MTC via the "External Controller" port.  From digging through the 
code (see testForMMCSysex()), it looks like rg responds to MMC and MTC 
via the "record in" port.  Try connecting to that.


  Hopefully someone else with more experience with this can chime in 
and confirm.


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] ROSEGARDEN 20.06, codename "Zepherine Drouhin" RELEASED

2020-06-03 Thread Ted Felix

The Rosegarden team is proud to announce the release of version 20.06
of Rosegarden, a MIDI sequencer that features a rich understanding of
music notation along with basic support for digital audio.

  http://www.rosegardenmusic.com/

Included in this release...

= Bug Fixes =

  * Fixed multiple memory leaks (r15684, r15685, etc...)
  * Fixed 100% CPU usage when drawing lines in the controller rulers,
bug #1571 (r15705)
  * Fixed multiple invalid reads (r15707, r15708)
  * Increased the height of the loop ruler, feature #481 (r15718)
  * Fixed CCs not going out when playback begins in the middle of
a segment, bug #1560 (r15735)
  * Fixed volume and pan CCs always going out even if the controllers
were removed from the Device, feature #482 (r15750)
  * Stop sending CCs on ch10 when metronome muted, feature #482
(r15751)
  * Fixed a problem with showing pitchbend ruler (r15771)
  * Made default for expression CC 127 instead of 100 (r15773)
  * Fixed a crash in Audio File Manager, bug #1576 (r15786)
  * Fixed a crash when importing .ly as X11/RG21, bug #1579 (r15806)
  * Fixed a crash on chords in pitch tracker, bug #1582 (r15835)

= New Features =

  * Track height can now be adjusted in the preferences, feature #479
(r15710-r15712, r15737)
  * New Segment > Transpose by Semitones... available in the Segment
Canvas, feature #480 (r15715)
  * Right-click can now be used in the loop ruler to set the loop,
feature #481 (r15717)
  * The current segment label is now displayed in the Matrix editor
when multiple segments have been selected for editing, feature
#470 (r15738-r15740)
  * Segment "For Notation" feature, bug #1578 (r15805)
  * Source distribution now includes these release notes in a
CHANGELOG file.  The release notes will continue to be available
on the wiki as well:
https://www.rosegardenmusic.com/wiki/start#release_notes

= Significant Code Cleanups =

  * PitchBend and Controller Sequence Dialog cleanup (r15662-r15684,
r15696-r15701, etc...)
  * MappedEventBuffer cleanup (r15719-r15730)
  * Colour removal and ColourMap cleanup (r15795-r15803)

= Additional Contributors =

  * Michael Stockinger (www.stockinger.org) - German Translation
  * Juan Carlos Pineda Arredondo - Roland-D5.rgd and ZynAddSubFX.rgd
  * Philip Leishman (lman) - Segment "For Notation" feature

Ted.


___
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] Cannot connect to jack

2020-05-24 Thread Ted Felix

On 5/24/20 12:46 PM, stefano franchi wrote:
some progress: the errors above were due to pulseaudio being active and 
messing with jack. Once I cut pulseaudio off the loop with a direct 
ALSA--> jackd connection (done with the cadence jack manager), most of 
the errors disappeared.


  Sounds like you are making good progress.  I recommend going through 
my Linux MIDI Guide.  It's designed to help you understand and 
troubleshoot JACK and ALSA and MIDI.  Should save you a lot of time in 
the long run.


  http://www.tedfelix.com/linux/linux-midi.html

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] Rosegarden App Icon

2020-05-22 Thread Ted Felix
  Just posted some new ideas for the rosegarden app icon on the wiki. 
Head over to the logo page and have a look at the pdf at the top.  Then 
let me know what you think.  We can discuss here on the list or send me 
an email directly if you prefer a silent ballot.


  https://www.rosegardenmusic.com/wiki/rosegarden_logo

Ted.


___
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] Guitar to USB setup

2020-05-22 Thread Ted Felix

On 5/22/20 11:13 AM, rhkra...@gmail.com wrote:

I was going to ask earlier, but what does a Guitar to USB cable do


  It's just a USB audio interface with an instrument level input stage. 
 No MIDI.


Ted.


___
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] Guitar to USB setup

2020-05-19 Thread Ted Felix

On 5/19/20 1:57 PM, Mike Witt wrote:
I see a lot of "Guitar to USB" cables advertised. They don't appear to 
have any "active" components. Does this really work?


  Probably.  The active components are usually hidden in the USB end. 
It's like one of those microscopic flash drives that are so easy to lose.


Anybody using one 
of these with Rosegarden? On Linux, hopefully?


  I'm *not* using one, but some suggestions from researching these 
things in the past:


1. The cheap ones probably aren't worth it.  I wouldn't spend less than $50.

2. They probably all work just fine with Linux.  But to be safe, pick a 
model and check the reviews on amazon to see if anyone's tested that 
model with Linux.  A google search might turn up something as well. 
E.g. a google search on 'behringer ucg102 linux' turns up a link on 
linuxmusicians.com that talks about that one and all of the others 
working fine in Linux.


3. Consider getting one someplace you can return it easily (e.g. 
brick-and-mortar) in case you can't find reviews indicating it will work 
with Linux.  Might not be easy in the time of coronavirus.


Ted.


___
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] Rosegarden still to 'aggressive' in sending Bank and Program changes?

2020-05-18 Thread Ted Felix

On 5/17/20 10:59 PM, r...@hydrophones.com wrote:

I'd like to do that but the Sourceforge platform will not let me register
or login -- it is seriously unfriendly, at least to this Linux user.


  That's strange.  It's always been really easy to work with in the 
past.  Maybe they are doing maintenance or something?  It does act funny 
every now and then.


Ted.


___
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] Rosegarden still to 'aggressive' in sending Bank and Program changes?

2020-05-17 Thread Ted Felix

On 5/17/20 7:46 AM, r...@hydrophones.com wrote:

I agree.


  Lorenzo will be opening a feature request for this.  Be sure to vote 
for it or chime in in the comments.


Ted.


___
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] Rosegarden still to 'aggressive' in sending Bank and Program changes?

2020-05-17 Thread Ted Felix

On 5/17/20 5:43 AM, Lorenzo Sutton wrote:
Rosegarden sends a Bank selection (CC 0 + CC 32) and Program change 
every single time playback starts from the beginning (i.e. measure 1, 
and this includes if you are looping with a measure which includes 
measure 1).


  I was waiting for someone to complain about this.  Now that CCs are 
sorted, this needs to be looked at next.


  Please open a feature request and copy your message into it.

  Should be able to get this done for 20.12.

Ted.


___
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] Testing Appreciated for 20.06

2020-05-17 Thread Ted Felix

On 5/17/20 4:27 AM, Lorenzo Sutton wrote:
I think for 'users' this might also be useful (the link you provide does 
ask for a password when trying the first svn command ;)


  Yeah, I probably need to switch to recommending the other one and 
keep it maintained as well.



Also, anything in particular we should try out, look for?


  The controller rulers have been problematic.  They could use some 
testing.


  In the editors: Controllers -> Insert Pitch Bend Sequence and Insert 
Controller Sequence.


  Other than that, there's been a lot of changes throughout that would 
be best tested by normal use.


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] HiDPI and the Toolbar Icons

2020-05-16 Thread Ted Felix

  A couple questions for those with HiDPI displays...

  Are the current rg toolbar icons too small on a HiDPI display?  At 
22px, I'm assuming they must be very tiny.  Unless Qt scales them up for 
us.  Does it?  If so, do they look ok?


  Can you send me (directly) a screenshot of an app (any app) with 
toolbar buttons that are appropriately sized on a HiDPI display?


  Looking into maybe improving the toolbar icons if a specific need can 
be identified.


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] Testing Appreciated for 20.06

2020-05-11 Thread Ted Felix
  20.06 will be released on June 3.  Right now, what you see in svn is 
the release candidate.  Please test the latest svn as much as you can. 
If you've not built from svn, give it a shot and let us know if you run 
into trouble.  We'll walk you through it.


  The steps I follow for building from svn are here:

https://www.rosegardenmusic.com/wiki/dev:eclipse

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


Re: [Rosegarden-user] Rosegarden Crashes when I stop a recording.

2020-04-27 Thread Ted Felix

On 4/27/20 1:51 AM, liebre...@grossmann-venter.com wrote:

Rosegarden Crashes repeatedly when I stop a recording.
You also lose the entire recording even if autosave was on.
It just eats everything.

Is this a known bug with 17.12 ?


  There is no known bug with that description.  You can search the bug 
reports here:


https://sourceforge.net/p/rosegarden/bugs/

  In general, I try to address crashes within a day or two.  The only 
current outstanding crash involves opening a file of one type as another 
(#1579).


I cannot upgrade to the most recent rosegarden as there is seemingly a 
bug in the cmake instructions.


  The square brackets, "[]", in the instructions mean that the value or 
option are *optional*.  Do not type in the square brackets.  This is a 
common convention when describing commands.  You'll see it in man pages. 
 As an example, if I look at the man page for the "ls" command:


$ man ls

...I see this:

SYNOPSIS
   ls [OPTION]... [FILE]...

  The [OPTION] means that options are optional.  It does not mean to 
type this:


$ ls [OPTION]

  It means I can run ls with or without an option.  Like this:

$ ls
$ ls -l

  I also noticed that you are running a shell as root.  That is 
universally considered dangerous as it allows you to make whatever 
modifications you want to the system.  Including ones that could damage 
your system and cause Rosegarden to crash.  I would recommend running as 
a limited user instead.


  Anyway, let us know if you need any more help.  We should be able to 
get you through the build and run process.  And we always need more 
people doing testing.


Ted.


___
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] MIDI Monitor Recommendation

2020-04-25 Thread Ted Felix

On 4/25/20 3:14 AM, Will Godfrey wrote:

It's available from sourceforge, alternatively Will QmidiRoute do what you want?


  I'll try building kmidimon and see what happens.  This is for my 
Linux MIDI tutorial, though, so I need something that's easily installed.


  Ubuntu still carries QMidiRoute and it looks like it does what I 
need.  I'll give it a shot.  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


Re: [Rosegarden-user] MIDI Monitor Recommendation

2020-04-25 Thread Ted Felix

On 4/25/20 12:50 AM, david wrote:

Hmm. Add Debian Buster or KXStudio repositories and get it from there?
I have it here in Debian Buster.


  That's interesting.  I originally assumed it was dropped due to 
dependencies that Ubuntu no longer supports.  Sounds like that's not the 
case for Debian.  I'll have to try building it and see if anything is 
missing.


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] MIDI Monitor Recommendation

2020-04-24 Thread Ted Felix
  So, it turns out Ubuntu 20.04 has dropped kmidimon from its packages. 
 Does anyone have any suggestions for other similar MIDI monitoring 
tools that display MIDI messages as they are sent via ALSA?


Ted.


___
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] RGD files for Roland D5 and ZynAddSubFX synthesizers

2020-04-14 Thread Ted Felix

On 4/14/20 7:09 AM, Juan Carlos Pineda Arredondo wrote:

Here they are .rgd files available for anyone interested:


  Thanks.

  I've created a patch in the tracker and added these.  We will add 
them to the next delivery of rg (20.06).


https://sourceforge.net/p/rosegarden/patches/71/

Ted.


___
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] Pitchbend bug

2020-04-10 Thread Ted Felix

On 4/10/20 4:17 PM, Lorenzo Sutton wrote:
That makes sense, I wasn't aware it had always been sent and for some 
reason though CC 123 would take care of this, but thinking about it this 
doesn't make sense (I wonder if CC 121 would, but again an explicit 
pitchbend is probably better with many synths).


  I suspect that pitchbend isn't reset by the usual resets.  Reason I 
know this is that it wasn't while I was testing.  Also rg's panic button 
doesn't zero pitchbend either.  Of course, each synth probably does 
whatever it wants.  So, yeah, probably safe in most cases.


Ted.


___
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] Pitchbend bug

2020-04-10 Thread Ted Felix

On 4/10/20 4:07 AM, Lorenzo Sutton wrote:
Unfortunately it seems now the bug seems to still be there, but sending 
out a pitchbend of 8192 - this is in r15766


  That is working as designed.  8192 is the correct number (in the 
0-16383 system).  That means no pitchbend at all.  So, from a fresh 
power-on of your synths, if you stop and play in the middle of a 
segment, it shouldn't suddenly end up pitched down.


  If you don't want pitchbend going out ever, remove pitchbend from the 
device's controllers.  If it's still going out after this, let me know 
and I will add it to my todo to track down.  I have another pitchbend 
ruler issue to track down, so I can take a look at both of them while 
I'm in there.


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] Pitchbend bug

2020-04-08 Thread Ted Felix
  I just pushed a fix [r15764] for a rather serious pitchbend bug.  At 
playback from the middle of a segment, a pitchbend of 0 (-8192) would be 
sent out on every channel.  This would shift the pithces downward 
depending on the instrument's response to pitchbend.


  If you've been running from svn, please update to the latest so that 
you do not run into this bug.


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] JACK MIDI... Why?

2020-03-22 Thread Ted Felix

  Next question

  Why do you want/need JACK MIDI?  (As with LV2 I may never get to it, 
but some justification will help focus the effort if I do.)  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


Re: [Rosegarden-user] LV2... Why?

2020-03-12 Thread Ted Felix

On 3/12/20 3:25 AM, Will Godfrey wrote:

People want it so that they can use various effects and instrument plugins,
some of which don't have stand-alone versions.


  Thanks, Will.

  BTW, have you been working with the latest svn?  Some recent changes 
might affect you.  Hopefully you have time to run it through the wringer.


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] LV2... Why?

2020-03-11 Thread Ted Felix
  Just wondering why folks want LV2 support.  Not that I'll ever get to 
it, but some justification will help focus the effort.  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] Testing Request: Rulers and PBSD

2020-02-09 Thread Ted Felix
  If anyone has some time for testing, I've just finished a review and 
cleanup of the "PitchBendSequenceDialog" (PBSD) and fixed numerous 
memory leaks in the rulers.  Grab the latest svn and give it a whirl.


  Some testing ideas in case you've not used these features...

  In the Matrix and Notation editors, the rulers can be enabled via the 
view menu:


  - View > Rulers > Show Pitch Bend Ruler
  - View > Rulers > Add Control Ruler > Volume Controller

  The rulers can be tested by doing some editing within them.  Add 
points, select points, move points, delete points, and anything else you 
can think of.


  The PBSD is used to insert pitch bend and other controller sequences 
(e.g. volume).  It can be reached through two menu items in the Matrix 
and Notation editors:


  - Controllers > Insert Pitch Bend Sequence...
- A note needs to be selected for this to work.

  - Controllers > Insert Controller Sequence...
- A note needs to be selected and a controller ruler.

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


Re: [Rosegarden-user] Looping

2020-01-02 Thread Ted Felix

On 8/19/19 4:46 AM, axs203 . wrote:
Rosegarden requires a shift key ( from memory) to be held and the actual 
size above the ruler to drag out a loop range is quite narrow 
vertically.


  I've opened feature request #481 to track this.  Looks pretty easy to 
fix.


https://sourceforge.net/p/rosegarden/feature-requests/481/

Ted.


___
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] ROSEGARDEN 19.12, codename "Yesterday" RELEASED

2019-12-15 Thread Ted Felix

On 12/13/19 4:40 PM, Will Godfrey wrote:

Just found time to try this, but get a build failure:

src/CMakeFiles/rosegarden.dir/gui/application/main.cpp.o 
/home/will/Software/rosegarden-15660/src/gui/application/main.cpp:
In function ‘void convert(const
QStringList&)’: 
/home/will/Software/rosegarden-15660/src/gui/application/main.cpp:356:10:
error: ‘doc’ was not declared in this scope ok = doc.openDocument( ^~~


  That's very odd.  doc is defined just a few lines prior to this at 
347.  This makes me think RosegardenDocument isn't defined somehow.  But 
that would have caused an error on line 347.


  Try expanding into an empty directory and building from there. 
Capture all the build steps and output in case there's a clue hiding there.


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] ROSEGARDEN 19.12, codename "Yesterday" RELEASED

2019-12-11 Thread Ted Felix

The Rosegarden team is proud to announce the release of version 19.12
of Rosegarden, a MIDI sequencer that features a rich understanding of
music notation along with basic support for digital audio.

  http://www.rosegardenmusic.com/

Included in this release...

= Bug Fixes =

  * Fix synth plugin Send Test Note button (r15502)
  * Plugin Editor: Fix Send Test Note not sustaining (r15511)
  * Treat Left + Right mouse buttons as Middle button in Matrix and
Notation editors (r15522)
  * Matrix: Fix no update on delete.  Bug #1569.  (r15616)
  * Fix missing metronome beat when expanding end during recording
(r15621)
  * Style: Make QLineEdit text gray when disabled (r15631)
  * Fix transposing keys by interval (r15632)
  * RMW: Reduce calls to leave/enterActionState() (r15641)

= New Features =

  * Matrix Editor: Implement new auto-scroll behavior (r15525 - r15544)
  * Segment Canvas: Implement new auto-scroll behavior
(r15546 - r15551)
  * Notation Editor: Implement new auto-scroll behavior
(r15553 - r1)
  * Add --convert command line option (r15607)
  * Add quaver triplet to standard quantizations (r15613)
  * Quantize: Add arbitrary quantization units.  Feature #477. (r15630)
  * Preferences: Allow dangerous operations during playback.
Feature #478. (r15643, r15646, r15647, r15651)
  * Add ability to enable metronome during count-in and enable/disable
during recording.  Feature #453. (r15649)

= Significant Code Cleanups =

  * Segment Parameters box cleanup (r15562 - r15597, r15604)
  * Replace FastVector with std::vector (r15609)
  * Quantize dialog cleanup (r15611 - r15614, r15617 - r15620,
r15623 - r15629)
  * Preferences cleanup (r15634 - r15640)

= Additional Contributors =

  * Hubert Figuiere
  * lman
  * David P. Sicilia

Ted.


___
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] Rosegarden and qsynth

2019-11-26 Thread Ted Felix

On 11/26/19 5:46 AM, Lorenzo Sutton wrote:
I think it *is* also worth mentioning that different CPU 'monitoring' 
tools provide different results:


  Absolutely.  Which is why I mentioned "top" specifically.

I do have a laptop with Intel's so-called 'hyperthreading' and was 
reading that readings about CPU usage etc. can be quite weird in this 
case..


  As a rule of thumb, a HT logical core is like having an additional 
15% of a full physical core.  Or for a 2 physical HT core, 4 logical 
core CPU you've got about 2.3 CPUs.  I'm pretty sure "top" reports a 
total of 400% CPU with all four logical cores in use.  Windows Task 
Manager reports a total of 100% for all four.  So a maxed out core in 
Windows looks like 25% on a four logical core machine.


  So, yeah, it's a mess.  I've spent a lot of time working with these 
numbers.  You get used to them.


Ted.


___
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] Rosegarden and qsynth

2019-11-25 Thread Ted Felix

On 11/22/19 6:39 PM, Will Godfrey wrote:

Exactly what setup have you got there if you don't mind me asking?


  1st gen Core i3-370M.  I'm using "top" to measure CPU usage.

  I go with low spec hardware to make performance issues like this 
easier to see.  This is how I was able to improve rg performance 
significantly.


  To me the important point is that while idling, fluidsynth uses a 
small fraction of the CPU that zyn and yosh use.  That means there is 
definitely room for improvement.


  It's entirely possible that the issue is related to the fact that I'm 
using whatever version Ubuntu 18.04 shipped with.  Maybe they built this 
with optimizations off?  Maybe these are older versions?


Ted.


___
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] Rosegarden and qsynth

2019-11-22 Thread Ted Felix

On 11/20/19 10:56 AM, Will Godfrey wrote:

There was one thing that grabbed my attention. That was when you say Zyn uses a
lot of CPU when idling.

I very much doubt that.


  I'm looking right at it.  I see an idling fluidsynth using 1%, jackd 
(which is never idling) using 2%, and an idling zynaddsubfx (3.0.3) 
using 35% cpu.


  Yoshimi 1.5.6 uses 20% cpu when idling.

Ted.


___
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] Rosegarden and qsynth

2019-11-19 Thread Ted Felix

On 11/18/19 6:15 PM, Tim Burgess via Rosegarden-user wrote:
If I start rosegarden, I can see messages go by and the Qsynth1 light 
flash, but no audio. When I exit rosegarden, I get a blare of sound (as 
if all notes sent are played at once).


  Sounds like the events aren't actually being consumed by fluidsynth 
and are piling up.  I see this behavior if I press keys on my keyboard 
before launching rg.  Then they all appear at once.



I'd appreciate any input/thoughts about how to resolve this.


  I don't know a lot about qsynth, but it sounds like it might be a 
good idea to run fluidsynth without qsynth to see if that helps at all. 
This is covered in my Linux MIDI Guide.


  Since you seem to be OK with building software, you will probably 
find my MIDI Guide interesting.  It covers almost everything and might 
get you to a new and more specific place where the trouble may lie:


  http://tedfelix.com/linux/linux-midi.html

Ted.


___
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] no midi out channel for zynaddsubfx

2019-08-14 Thread Ted Felix

On 8/14/19 5:23 PM, krsg...@trixtar.org wrote:

Regardless of whether the standalone Zynaddsubfx (not the dssi) gets connected 
in qjackctl manually or automatically. a usable channel for it no longer shows 
up in the Rosegarden 'Manage Midi Devices' window midi-channels list. Others 
like say qsynth or hydrogen still show up no problem just like before.


  After playing with this for a bit, I'm guessing that zyn is coming up 
in JACK MIDI mode.  Since rg doesn't support JACK MIDI, rg cannot see it.


  I have JACK audio set up and I started zyn from the command line like 
this and it worked:


zynaddsubfx -I ALSA -O JACK -a

  Actually, it appears as if zyn's defaults are exactly what I needed 
and all I need to do is this:


zynaddsubfx -a

  And finally, specifying the sample rate and buffer size to sync it up 
with how I run fluidsynth:


zynaddsubfx -I ALSA -O JACK -a -r 44100 -b 128

  The built-in help is actually pretty helpful:

zynaddsubfx --help

  If you would prefer to use the desktop icons to launch zyn, it 
appears that in Ubuntu, at least, the three desktop icons that are 
offered are not helpful.  The JACK one assumes JACK MIDI and Audio while 
the ALSA one assumes ALSA MIDI and Audio.  Not what I need anyway.  You 
can see the .desktop files here:


ls -l /usr/share/applications/zyn*

  You can make your own from one of these and modify the options to 
match your needs.


Ted.


___
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] Play/stop transport with space bar?

2019-08-14 Thread Ted Felix

On 8/14/19 6:48 AM, Arnaldo Pirrone wrote:
It would be great if we had the play/stop transport shortcut key on 
space instead of CTRL+ENTER, and the rec button on CTRL+R, or just R to 
start recording when the transport is rolling. Can it be done?


https://sourceforge.net/p/rosegarden/feature-requests/412/

Also, it would just be great if the default double click action on 
tracks opens the piano roll/matrix editor by default instead of the sheet.


  Go to Edit > Preferences... > General > Behavior > Double-click opens 
segment in.  You can select whichever of the three editors you prefer.


Ted.


___
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] 17.12 & 19.06 Studio>Audio-Mixer

2019-08-08 Thread Ted Felix

On 8/1/19 7:28 PM, krsg...@trixtar.org wrote:

When I launch the mixer with no file loaded into RG the channels or tracks are 
numbered but these number are very tiny (almost unreadable),


  The size is set in the code in AudioStrip.cpp, line 72.  Right now 
it's 6 point, and that's pretty tiny alright.



and they vanish altogether as soon as a file gets loaded


  The .rg file?  That sounds like something is odd in the file.

  What's being displayed is the name of the Instrument which you can 
set by selecting the track in the main window and then pressing the 
microscopic button to the left of the name in the "Instrument 
Parameters" box.  If you just want the numbers, 1-16, you can put those 
in there.  Though that might look strange in other parts of the UI. 
Maybe A1, A2, etc



Is this a bug, is it known, any fixes?


  This is working as designed for the most part.

  I suspect that we should probably go ahead and make the audio mixer a 
bit bigger to fill some sort of standard minimum monitor size.  Then we 
should be able to go with a slightly large point size for all the controls.


Ted.


___
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] (no subject)

2019-07-24 Thread Ted Felix

On 7/24/19 6:34 PM, Ted Felix wrote:
   We might want to make this a preference in case anyone wants to 
adjust it.  Anyone else interested in this feature?


  Thanks to Mr. Sweaters for seconding this one.  I've opened a new 
feature request to track:


https://sourceforge.net/p/rosegarden/feature-requests/479/

  Get in there and upvote with the "+" button if you're interested.

Ted.


___
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] (no subject)

2019-07-24 Thread Ted Felix

On 7/24/19 4:24 PM, "Jan Beneš III." wrote:

may I ask please - is there any way to make track bands wider or taller? I 
would like to see events in the track view better.


  Vertical size is fixed.  However, if you are building the code, you 
can experiment.  Take a look at TrackEditor::m_trackCellHeight which is 
set in TrackEditor::init() in src/gui/editors/segment/TrackEditor.cpp. 
I tried 50 and got some really tall tracks:


//m_trackCellHeight = std::min(fontMetrics.height() + 9, 24);
m_trackCellHeight = 50;

  Note that I also tried 100 and that caused trouble, so there are limits.

  We might want to make this a preference in case anyone wants to 
adjust it.  Anyone else interested in this feature?


Ted.


___
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] Not able to carry out some editing while playing back?

2019-07-22 Thread Ted Felix

On 7/22/19 3:05 AM, Lorenzo Sutton wrote:
I see your point, although I must admit it feels much less usable and 
'reactive' now, especially being able to place notes, move them etc. 


  Your way of working is completely valid and rg should support it. 
Unfortunately, much of the code doesn't handle this well and I really 
don't want us crashing and losing people's data.  Perhaps we should add 
an "edit during playback" option to the preferences.  It will warn that 
this practice may lead to instability, crashes, and data loss.  Then it 
will reverse the fixes that I put in.  Or at least a subset of them that 
you were using.


  I'll add a feature request...

Ted.


___
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] Segment Parameters Box

2019-06-26 Thread Ted Felix

On 6/26/19 8:17 PM, Ted Felix wrote:

On 6/26/19 4:05 PM, D. Michael McIntyre wrote:
That did the trick.  My only quibble now is that the label seems a few 
pixels short.
   I'll play around a bit and see if I can't give the whole thing a few 
pixels of breathing room.


  This looks pretty good:

m_label->setContentsMargins(4, 3, 4, 3);

  I'll go ahead and add that.

Ted.


___
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] Segment Parameters Box

2019-06-26 Thread Ted Felix

On 6/26/19 4:05 PM, D. Michael McIntyre wrote:
That did the trick.  My only quibble now is that the label seems a few 
pixels short.


  I noticed that as well.  It looks like the "Edit" button was 
expanding that row in the layout quite a bit vertically.  Now it's 
exactly the size that the text needs which feels rather claustrophobic.


  I'll play around a bit and see if I can't give the whole thing a few 
pixels of breathing room.


Ted.


___
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] Segment Parameters Box

2019-06-26 Thread Ted Felix

On 6/25/19 3:49 PM, D. Michael McIntyre wrote:
What seems questionable is that this label text no longer seems to be 
constrained,


  I took one more crack at this in [r15606] and I think this approach 
is best.  Essentially, the text no longer dictates the size (hint) of 
the control.  Now the layout is in control.  The label simply expands to 
fill the layout horizontally.


  That's what I wanted originally.  Hopefully it will work properly 
now.  give it a whirl.  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


Re: [Rosegarden-user] Segment Parameters Box

2019-06-25 Thread Ted Felix

On 6/25/19 3:49 PM, D. Michael McIntyre wrote:
What seems questionable is that this label text no longer seems to be 
constrained,


  Should be fixed in [r15604].

Ted.


___
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] Segment Parameters Box

2019-06-25 Thread Ted Felix

On 6/25/19 3:49 PM, D. Michael McIntyre wrote:
What seems questionable is that this label text no longer seems to be 
constrained, so if I click around and activate a segment with a stupidly 
long label, the entire parameter area side panel thing expands to 
accommodate this wide text.  I'm thinking that was an oversight rather 
than a design choice.


  Definitely.  Good catch.  I'll have a look.

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] Segment Parameters Box

2019-06-25 Thread Ted Felix
  Review and rewrite of the Segment Parameters box in the upper left is 
complete.  Please grab the latest svn and give it a good brutal testing 
for me.  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


Re: [Rosegarden-user] Auto-scrolling

2019-06-13 Thread Ted Felix

On 4/3/18 12:35 PM, Will Godfrey wrote:

If Ted can fix this he'll be my hero for life :)


  I'm now applying for the above position.  Sorry it took over a year...

  In svn you will now find that both the Matrix and the Notation 
editors behave like the segment canvas.  If you drag something outside 
of the viewport, a relatively gentle auto-scroll should begin.  The 
further you go outside, the faster the auto-scroll.  Horizontally 
anyway.  Vertical is fixed rate since rg doesn't do a lot vertically.


  If the window is close to the right edge of the monitor, the 
auto-scroll area might be moved into the viewport a little.


  Also, all of the rulers (standard, tempo, controller) should exhibit 
the same behavior.  In some places, this was completely missing, so if 
you've grumbled about a lack of auto-scroll anywhere in the UI, it 
should be there now.  If it isn't, it can be added very easily.


  As always, please test this to death and let me know if I missed 
anything.


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] ROSEGARDEN 19.06, codename "Xerxes" RELEASED

2019-06-05 Thread Ted Felix

The Rosegarden team is proud to announce the release of version 19.06
of Rosegarden, a MIDI sequencer that features a rich understanding of
music notation along with basic support for digital audio.

http://www.rosegardenmusic.com/

Highlights of 19.06 include cleanup and bug fixes.

= Bug Fixes =

  * Fix potential crashes during playback (r15422)
  * Fix crash when rescaling during playback (r15423)
  * Fix crash when splitting during playback (r15424)
  * Fix horizontal wheel scrolling on segment canvas
(Bug #1561, r15430)
  * Fix horizontal wheel scrolling on Matrix and Notation editors
(Bug #1561, r15432)
  * Enable High DPI Scaling (Bug #1565, r15444)
  * Roland-CM-64.rgd: Fix SN-U110-02 perc mappings (r15448)
  * Avoid using null pointers (Bug #1566, r15453)
  * Fix broken right edge auto scroll (Bug #1567, r15462, r15464)
  * Fix crash when adding a Track with no Devices (r15463)
  * Fix numerous memory leaks
(r15466, r15470-5, r15477, r15479, r15481, etc...)
  * Fix build with Qt 5.7, which doesn't have QSharedPointer(nullptr)
(r15476)
  * SegmentParameterBox: Fix invalid memory access (Bug #1568, r15492)
  * Fix Qt 5.7 QSharedPointer/nullptr build issue (r15493)

= New Features =

  * Add Ctrl+Space shortcut for record (r15482)

= Other people who contributed to Rosegarden development: =

  * Hubert Figuiere
  * Lucas

Ted.


___
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] Convert .rg to .mid with Command Line

2019-05-22 Thread Ted Felix

On 5/22/19 12:30 AM, David Sicilia wrote:

Thanks, I have just opened this ticket:

https://sourceforge.net/p/rosegarden/feature-requests/476/


  Great.  Thanks.  Providing a use case and rationale is really helpful 
in providing motivation for selecting this one next.


Ted.


___
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] Convert .rg to .mid with Command Line

2019-05-22 Thread Ted Felix

On 5/22/19 2:02 AM, david wrote:
It looks pretty clear. Not being Ted or a real coder, not sure how easy 
to do.


  Might require a touch of redesign here and there to make it clean and 
prevent it from popping up the GUI.


Want something really tough? How about a general RG scripting interface, 
like for Python? ;)


  I think the testing harness that David added might be a step in that 
direction.


Ted.


___
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] Convert .rg to .mid with Command Line

2019-05-21 Thread Ted Felix

On 5/21/19 10:01 AM, David Sicilia wrote:

Is there a way to use Rosegarden (or some other program) as a command
line tool to export  Rosegarden files (*.rg) to MIDI?


  No.


If such a feature doesn't exist, is it worth requesting it from the developers?


  All features are worth requesting.  Open a feature request on 
SourceForge:


  https://sourceforge.net/p/rosegarden/feature-requests/

  This is where I look for the next thing to work on.

Ted.


___
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] Crashes :(

2019-05-15 Thread Ted Felix

On 5/15/19 2:36 PM, Will Godfrey wrote:

My understanding is that nullptr is the 'modern' replacement for NULL, so I
can't see why it would fail.


  It is.

  In this case, the older version of QSharedPointer provides no way for 
the compiler to convert a nullptr (which is actually a nullptr_t) to a 
QSharedPointer which operator= requires.  It's a bit convoluted and only 
happens in rare situations.  QSharedPointer prior to Qt 5.8 is 
unfortunately one of those situations.


  Gorey details are attached in a small test program.


As a cross-check I tried changing a few old-style NULL with nullptr in
Yoshimi, and it compiled without complaint, so I'm quite lost now :(


  In most cases, nullptr is a drop-in replacement for NULL.

Ted.
// nullptr.cpp

// To Build:
// g++ -std=c++11 nullptr.cpp

// __cplusplus values:
//201103L = C++11
//199711L = C++98 (g++ 4.7 default)
#if __cplusplus >= 201103L
// This is C++11!
#else
#error Need to build with -std=c++11
#endif

class A
{
public:
explicit A(int *p) { m_p = p; }

// Ugh.  Avoid assignment operators if you can.
A =(const A )
{
m_p = rhs.m_p;
}

private:
int *m_p;
};

int main()
{
// nullptr is compatible with pointers in general.
int *p = nullptr;

// This is also OK.
A a(nullptr);

// This is not.  Because the ctor is explicit, we must 
// explicitly ask for the conversion for this to work.
//a = nullptr;

// Explicitly asking for the conversion works.
a = static_cast(nullptr);

// Adding a (non-explicit) ctor to A that takes a nullptr_t
// also fixes the problem.  Qt 5.8 adds this to QSharedPointer.

return 0;
}

___
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] Crashes :(

2019-05-15 Thread Ted Felix

On 5/15/19 2:36 PM, Will Godfrey wrote:

Just tried to compile this, and it fails with the report:

error: no match for ‘operator=’ (operand types are
‘QSharedPointer’ and ‘std::nullptr_t’) m_baseStyle =
nullptr;


  Should be fixed in [r15493].

Ted.


___
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] Crashes :(

2019-05-15 Thread Ted Felix

On 5/15/19 2:36 PM, Will Godfrey wrote:

Just tried to compile this, and it fails with the report:

error: no match for ‘operator=’ (operand types are
‘QSharedPointer’ and ‘std::nullptr_t’) m_baseStyle =
nullptr;


  Looks like you may have an older version of the Qt5 development 
stuff.  5.7 and prior do not support nullptr.  David fixed some of that 
in [r15476].  Looks like I need to be more careful.  Wonder if I can 
install the Qt 5.7 development stuff somehow so I can stay on top of this...


Ted.


___
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] Any way to continue a song with different studio?

2019-05-12 Thread Ted Felix

On 4/26/19 7:57 AM, Sami Jumppanen wrote:

3. Import song to the default empty studio project.


  Just pushed r15492 which was a memory access issue related to 
deleting segments and the segment parameter box.  This may have fixed a 
crash related to File > Merge that might have affected this particular case.


Ted.


___
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] Crashes :(

2019-05-12 Thread Ted Felix

On 5/3/19 5:06 AM, Will Godfrey wrote:

I haven't forgotten this - I'm still totally bogged down with work on Yoshimi.
There is a *lot* going on!


  Just pushed r15492 which was a memory access issue related to 
deleting segments and the segment parameter box.  This may have fixed 
the problem.


Ted.


___
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] Any way to continue a song with different studio?

2019-05-09 Thread Ted Felix

On 5/9/19 8:32 AM, Ted Felix wrote:
   This is a good point.  RG restricts the user to one synth per MIDI 
port.  That's unrealistic.  The concept of a synth needs representation 
in the Studio.  I'll open a feature request.


  Actually, upon further thought, you can create a Device for each 
synth and assign multiple Devices to the same port.  Assuming this works 
as it should, one can do daisy-chaining in this way.  It's a little 
clunky since the MIDI channels aren't enforced by rg and one could 
accidentally set a Track up to talk to a channel on the port that is 
incorrect for a specific Synth in the chain.


Ted.


___
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] Any way to continue a song with different studio?

2019-05-09 Thread Ted Felix

On 4/26/19 7:57 AM, Sami Jumppanen wrote:
I'm having problems with my changed hardware synth studio (physical MIDI 
ports) and continuing the old songs. I have found and tried different 
ways to migrate songs to new studio, but none of them seem to work well 
or at all.


  I would just create the studio from scratch and assign each track to 
the proper device/channel.  This might be a rather large undertaking, 
though.



2. Open the old song and then import a new studio from a newer RG file.


  I tested this with a very simple (one Device) setup.  It worked fine 
for me.  RG assigns all Tracks to the first Device it finds.  This is 
probably the best guess it can come up with.  So, then it should be a 
matter of assigning the Tracks to the proper Devices/Channels.


  You mentioned that with one of these approaches (2 or 3), you ended 
up only with "one port of the new MIDI device".  That certainly sounds 
like a bug that would be easy to track down and fix.  Was that this 
case?  Can you provide more details so I can try to reproduce this and 
fix it?



3. Import song to the default empty studio project.


  I'm not sure how you did this.  I tried doing this with File > Merge 
> Merge File...  Is that what you tried?  For me, it did nothing.  I 
ended up getting an exception thrown and caught in 
RosegardenApplication.  (I'm running a debug build, so that might be why 
I didn't crash.  Still, it was a std::bad_alloc which is probably fatal 
at some later point.)


  Was this the case that crashed for you?  Crashes are certainly cause 
for concern, so I'll have a closer look.  If you can give me more 
details of the crash you encountered, I'd appreciate it.  Would like to 
get it fixed before the 19.06 release if I can.


Device configuration should be detached from sequencer work, it makes 
very little sense anyway. Firstly mixing the terms device and port. One 
MIDI port can feed many devices (synthesizers) by daisy chaining (or 
other splitters/processors), so it's not rational to think about a MIDI 
port as a "device".


  This is a good point.  RG restricts the user to one synth per MIDI 
port.  That's unrealistic.  The concept of a synth needs representation 
in the Studio.  I'll open a feature request.


If no 
settings are found, so what? Just display patch numbers.


  See Feature Request #455:

https://sourceforge.net/p/rosegarden/feature-requests/455/

Ted.


___
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] Crashes :(

2019-04-17 Thread Ted Felix

On 4/8/19 3:56 AM, Will Godfrey wrote:

I'm afraid I've still not had time to do anything with this - not fired up the
DAW at all lately - to much other stuff getting in the way :(


  I've updated the instructions for getting a stack trace on the wiki.

https://www.rosegardenmusic.com/wiki/frequently_asked_questions#how_to_get_a_stack_trace_for_a_crash

  I missed a few details before.  You'll need:

1. A binary built with debugging turned on.
2. ulimit set so that a core dump can be created.
3. Run from command line.
4. Use gdb's "bt" command to get the backtrace out of the core file.

  I've been completely unable to reproduce this crash.  The core dump 
and backtrace will be critical to tracking this down and fixing.


Ted.


___
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] Crashes :(

2019-04-10 Thread Ted Felix

On 4/8/19 3:56 AM, Will Godfrey wrote:

My guess is that the delete
operation has left an invalid pointer somewhere.


  There are definitely bad pointers left around in the code. 
Intentionally.  (!?)  See the comments in 
SequenceManager::segmentDeleted().  If, as I explain there, the memory 
manager decides to hand you a pointer to the same block of memory again 
(a very reasonable possibility), then we've got a crash in the making. 
This might explain why some see it and some don't.  Different memory 
manager behavior.



However, if *immediately* after doing the delete, I save the file it seems to
sort everything out and is then OK.


  Save might be scrambling the memory manager enough that the same 
pointer address doesn't get reused.


  I'll see if I can reproduce this given all the new info.  It would be 
nice to have confirmation that SequenceManager is the culprit.  If I 
can't reproduce, I'll have a closer look at fixing all the intentional 
dangling pointers.  It's something that needs to be done anyway.


Ted.


___
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] Identify / Change kayboard shortcut for record?

2019-04-02 Thread Ted Felix

On 4/1/19 4:15 PM, Lorenzo Sutton wrote:
The tooltip suggests that this is 'media record' which I assume must be 
some weird key on 'multimedia keyboards'.


  The other option is to use the space bar after starting playback. 
So...  Ctrl+Enter to begin playback, then spacebar to start recording.


  This seems less than ideal since it doesn't give you a metronome at 
the very beginning.


Is there a relatively easy way 
to change this? IIRC now keyboard shortcuts are hardcoded somewhere 
while once they used to be editable.


  data/rc/rosegardenmainwindow.rc contains the shortcuts for the main 
window.


  See feature request #412 for more details.  I just updated it.

https://sourceforge.net/p/rosegarden/feature-requests/412/

In this case I think the default should be changed to something more 
sane such as SHIFT+R or maybe CTRL+R (as the play shortcut is CTRL + 
Enter)...?


  Ctrl+R is used in notation.rc for Phrase > Triplet.

  Shift+R is used in the notation editor (notation.rc) to insert an F#.

  Shift+R is used in the matrix editor (matrix.rc) to insert an F# as well.

  I just tried Ctrl+R in rosegardenmainwindow.rc and we crash due to 
the conflict.


  Some possible options:

Ctrl+Space - This works fine, no conflicts.

Ctrl+E - This works fine, no conflicts.

Ctrl+R - We would have to move the notation editor's Ctrl+R to Ctrl+I. 
Anyone who has Ctrl+R memorized as "Triplet" will be very annoyed.


  I'm thinking Ctrl+Space is probably the way to go.  Any concerns?

Ted.


___
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] Bug tracking in general

2019-04-02 Thread Ted Felix

On 4/2/19 4:36 AM, Lorenzo Sutton wrote:
Is the Sourceforce bug tracker 
[https://sourceforge.net/p/rosegarden/bugs/] still the preferred way to 
report bugs? It's tempting to write here on the list :) but as much as 
sourceforge is not the prettiest I guess it's still the best practice 
for reference etc.?


  What seems to work best is to post to the user group first and get 
some thoughts, then open a feature request or bug report if the feedback 
doesn't solve the problem.


Ted.


___
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] Crashes :(

2019-03-28 Thread Ted Felix

On 3/28/19 6:24 PM, Will Godfrey wrote:

Not forgotten this - just busy firefighting yoshimi :(


  No problem.  I've been working my way through everything valgrind is 
showing me to see if I can find your crashes.  Nothing yet.  Lots of 
memory leaks, but nothing crashy.


Ted.


___
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] Crashes :(

2019-03-24 Thread Ted Felix

On 3/23/19 5:27 AM, Will Godfrey wrote:

Typically 3-4 minutes 8-12 tracks MIDI only.

Most commonly happens when splitting a track then deleting part of it. Actual
crash usually happens next time I go to 'Play'.


  Ok, thanks.  I'll see if I can reproduce with these sorts of edits.


Had a look at gdb in the past (ran away screaming) couldn't make a lot of sense
out of the information it presented.


  All I need is the backtrace which can be had by issuing the "bt" 
command in gdb against a core dump.  Might need to do some configuring 
to get your system to generate core dumps for rg.  E.g., run rg like 
this from the command line:


$ ulimit -c unlimited
$ rosegarden

  Then once you've got a core dump:

$ gdb rosegarden core

  And finally, use the "bt" (backtrace) command.

(gdb) bt

  Just need to copy/paste the output of the backtrace for me and I can 
do the rest.  Might not pinpoint the exact problem, but will certainly 
provide a data point to work with.



Some time back we had a segfault in
Yoshimi - gdb was no help at all and we found the problem by simply undoing
commits till we found the one where it changed :(


  A bisect is oftentimes a very useful tool in these situations.  But 
only if it's an easy-to-reproduce problem.  Running gdb on a core dump 
is best for intermittent or harder to reproduce issues.


Ted.


___
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] Crashes :(

2019-03-20 Thread Ted Felix

On 3/20/19 4:53 PM, Will Godfrey wrote:

There seems to be a memory allocation problem that is triggered by deleting
tracks/sections. This has been there for quite a long time, but it has only
been a real problem for me lately because I'm doing a lot of quite heavy
editing, and have to resort to doing very frequent saves.


  I recently saw a crash when deleting segments related to the undo 
history.  Couldn't reproduce it, though.  And it only happened on exit.



I'm not certain, but I get the impression that the very fact of doing these
saves seems to reduce the likelihood of the crashes.


  Memory issues are always tricky and can easily be exacerbated or 
mitigated by almost anything.



Currently using build 15419, but as I said, it's been there a long time - years.


  I'm not surprised.  Last time I ran valgrind against rg there were 
too many problems to even begin to track them all down.  I gave up 
because no one was complaining.  That appears to have changed.


  There are a number of ways you can help me track these down.

  First would be to characterize the editing you are doing.  Just a 
list of the sorts of things, like track deletes that seem to cause the 
issue.  Also, how big are the compositions?  Number of segments, tracks. 
 How many minutes?  Does it happen even if you never play the composition?


  Next, if you can analyze the core dumps for me when you crash, that 
could be a big help.  Though sometimes with memory issues, the stack 
traces are useless.  Have you worked with gdb and core dumps before?


Ted.


___
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] New Rosegarden Logo

2019-03-20 Thread Ted Felix

On 3/15/19 10:24 AM, Lorenzo Sutton wrote:
What would it look like inverted (i.e. white / transparent background) 
and coloured rose (but please remember the yellow/orange... :P )


  I was wondering that myself.  Don't think I can do it properly, 
though, the way I'm just fiddling with it in the GIMP.  I'll certainly 
talk to her about this and show her the app icons in GNOME/Ubuntu.


Ted.


___
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] New Rosegarden Logo

2019-03-17 Thread Ted Felix

On 3/16/19 2:02 PM, Chris Cannam wrote:

Website could do with a redesign too, not to mention tightening up the 
content...


  Website redesign is also planned.  Not sure she'll get to it, but 
it's on her mind.


Ted.


___
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] New Rosegarden Logo

2019-03-15 Thread Ted Felix

On 3/14/19 3:40 AM, Lorenzo Sutton wrote:
Looks good... I guess a 'square' concept, i.e. without the text which 
would be used for icons etc. would be useful ;)


  I added a tile mockup to the wiki page:

https://www.rosegardenmusic.com/wiki/rosegarden_logo

Ted.


___
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] New Rosegarden Logo

2019-03-15 Thread Ted Felix

On 3/14/19 2:56 PM, Dave Serls wrote:

   I can be as arbitrary and idiotic as the next fellow.
   Enlarge the elephant "G" a couple of font sizes and fill with pastel green.


  Ha!  Good stuff.

Ted.


___
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] New Rosegarden Logo

2019-03-15 Thread Ted Felix

On 3/14/19 3:40 AM, Lorenzo Sutton wrote:
Looks good... I guess a 'square' concept, i.e. without the text which 
would be used for icons etc. would be useful ;)


  I was thinking that too.  Not sure the "g" should be so prominent. 
However...  Elephant!


About the colour.. wondering if some reference to the old logo (yellow, 
orange tones) could remain.


  Of course.  I'll register your vote for this.

Ted.


___
Rosegarden-user mailing list
Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-user


  1   2   3   >