Re: [Rosegarden-user] Crucial: Help needed with Rosegarden Crashing:
Thanks, Rich. :) On 8/11/20 12:46 AM, r...@hydrophones.com wrote: Hi guys, I totally agree with what Mr. Jones has said. It works for me too. Been doing that since early days decades ago. Stay safe all. Best regards, Rich Hey, Michael! Long time no hear but yeah 80 hour weeks are killers... Ideas for our original poster to possibly produce an alternative 'backup' recording. If recording audio, start a jack_record command line session with the "-jt" option. When you start Rosegarden recording, jack_record will also be recording the audio. If recording MIDI, have an amidi command line session running with the "-r" option followed by a filename? That will record a file containing same midi events as Rosegarden. If Rosegarden crashes, it won't affect what those have recorded unless it takes down the whole computer. On August 10, 2020 6:05:22 PM HST, "D. Michael McIntyre" wrote: On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: Rosegarden crashes deleting my recording in the process. This is painful. Those ideas wont come around again. Lost forever. I very definitely feel your pain here. It happened to me more times than I want to remember. So much creativity right out the window. One of the most frustrating things we used to have was this stupid composition of preset length. You're playing along, in the pocket, really coming up with great ideas, and then WHAM, it just stops recording! I got that to go away a long time ago, but the memories of all those notes still linger. Anyway, for the most part, Rosegarden is vastly more stable than it used to be. Thanks in great measure to Ted Felix coming along, and making diligent and methodical efforts to improve this sort of thing. He was generally very successful, but I guess we've experienced some kind of regression. All you can really do is start using a debug build from an environment that allows core dumps. The usual thing is to start running from development source, from a debug environment. It's not absolutely necessary to run from development source, but it will help you in the long run. If the crash is repeatable (we hope it is!) and somebody (Ted Felix most likely) fixes it, then you will want to be in a position to test the result. Here are the instructions from the wiki: How to get devel source: https://www.rosegardenmusic.com/wiki/development_from_svn How to get a stack trace for a crash First, make sure you are running a version of rosegarden that was built with debugging turned on. -DCMAKE_BUILD_TYPE=Debug Without debugging, there will be no symbols in the binary, and the backtrace will be useless. You'll likely need to build rosegarden on your own to get a debug version. Instructions can be found here: Using the Eclipse IDE to work on Rosegarden Open a terminal window, and check to ensure that applications will be able to produce core dumps. The exact command and syntax may vary from shell to shell, but for bash it is ulimit -a: $ ulimit -a core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 20 ... The above example is quite typical for an end-user desktop system. Having the “core file size†set to 0 prevents the creation of very large core dump files in unexpected places, and is generally a good thing. However, this also prevents you from generating a stack trace. You need to change the limit to something larger than 0, but many systems prevent you from setting this to unlimited, so we suggest $ ulimit -c 100 Now start a debug version of rosegarden from the command line, and reproduce the crash. You should now have a core file in your current directory. The core file is either named core or core.. Run gdb: $ gdb rosegarden Then once you get the gdb prompt, use the command 'bt' to get the stack trace, and mail it to the authors or to the Rosegarden development mailing list, or include it in a bug report. Anyway, I hope this helps. I haven't been involved with development for quite a number of years now, and I'm semi-retired. I still monitor things, and still care. It's just my life went in a different direction, I guess. Especially in terms of working hours. When I was most passionate about Rosegarden, and most active, I had the luxury of making a decent living working only 45 hours per week. Now it's closer to 80 with my commute. Anyhoo brother, if I could turn back the hands of time and save your ideas from getting obliterated, I would. I really hate it for you. So much so that you inspired me to write the longest message I've posted in probably going on 10 years. -- D. Michael McIntyre -- David W. Jones gn...@hawaii.rr.com authenticity, honesty, community http://dancingtreefrog.com "My password is the last 8 digits of π." ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe
Re: [Rosegarden-user] Crucial: Help needed with Rosegarden Crashing:
Well, I didn't roll my own program like you did... ;) Your pad control idea sounds good. Maybe looking at how other programs implement 'midi learn' would enable you to add something to your program that people could use to 'learn' the MIDI commands their particular pad uses?Does it also let you change between clips? On 8/11/20 9:25 AM, liebre...@grossmann-venter.com wrote: That is in fact exactly what I ended up doing. I wrote a program to record midi commands and is busy to integrate it to a pad controller so I have the pad controller directly at my keyboard. The pads are the recorded clop[s and first pad is always open to be recorded. When flashing it records when not it is saved and everything is moved down the 64 pads and pad one is free again etc. I finished it last night after I reverse engineered the pad controller midi commands. With the lookup table the implementation was easy. I then dont use Rosegarden for recording at all, and just use it for editing. I had to do it as this is the third time I lost important recordings. There is no better editor than Rosegarden for me. If it wasnt for the recording issue I would give it 100% I will make a video of a crash so you see it is real. I will also make a video of my solution. I wont mind making my program available but it will only work for one pad controller, so it is not really generally useful as you would need that pad controller else it will be clunky mouse work and i dont want to support software. Anyway you are spot on with your suggestion. It is exactly what I ended up doing. On 2020-08-11 03:35, David W. Jones wrote: Hey, Michael! Long time no hear but yeah 80 hour weeks are killers... Ideas for our original poster to possibly produce an alternative 'backup' recording. If recording audio, start a jack_record command line session with the "-jt" option. When you start Rosegarden recording, jack_record will also be recording the audio. If recording MIDI, have an amidi command line session running with the "-r" option followed by a filename? That will record a file containing same midi events as Rosegarden. If Rosegarden crashes, it won't affect what those have recorded unless it takes down the whole computer. On August 10, 2020 6:05:22 PM HST, "D. Michael McIntyre" wrote: On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: -- David W. Jones gn...@hawaii.rr.com authenticity, honesty, community http://dancingtreefrog.com "My password is the last 8 digits of π." ___ 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] Crucial: Help needed with Rosegarden Crashing:
Hmm, never heard of MX Linux. And I used to run MEPIS ages ago. I presume whatever QT library MX Linux is using is the one that comes from their distro. How does their QT library compare to whatever the current QT version is? Is your MX Linux 32-bit or 64-bit? Just wondering, don't know if it's important or not. I use Debian 10. It looks like MX is based on Debian 9. So there might be QT bug fixes Debian 10 has that Debian 9 doesn't. Also, Debian 10 has RG 18.12 right now; that might have bug fixes missing from RG17.x. I have a development snapshot (not a recent one, I'm sure) of RG 20.06. It looks and works much better on my 4K display here than 18.12 does. But I haven't tried recording any audio with it. I have had crashes in the past with Rosegarden, but never while recording audio or MIDI. Those were also on my old i7/16GB RAM. Current machine is an i9/64GB RAM. I mostly work in the score view and the main window. Might not trigger the crashes you encounter. You mentioned crashes when you record about 30 minutes. Maybe what's happening is your system is set to record to tmp folder? On my system, by default, it pointed /tmp at RAM, not disk. That makes things fast, but at the time I only had 16GB of RAM and /tmp was only giving about 2GB or so. That maybe Audacity-specific, not RG. When I changed /tmp to point to ~/tmp, suddenly the amount I could record jumped to 100+ hours. Actually, I think RG records to a subfolder in your home folder? So maybe that doesn't apply. Hope that helps or inspires more ideas! On 8/11/20 9:40 AM, liebre...@grossmann-venter.com wrote: All just remember I am not blaming Rosegarden. This is not a rant or badmouthing Rosegarden. Rosegarden is free Magic software, unreplaceable for me. Since this doesnt happen to those who responded, there must be something locally here that triggers the crash and I think I know what the culprit might be. I really STRONGLY suspect QT to be the problem. If QT induces a crash and take Ropsegarden with it, then Rosegarden was stopped without a SIGINT and just crashed by QT. Since there was not at least a SIGINT, all data in memory will be lost and since it seems Rosegarden do not write to file and ljkeep the recording in memory only writing upon save button, all will be lost. I have had similar issues since QT was owned by Troltec decade or so ago. I actually stopped developing with QT long ago due to similar stability issues. I am pretty sure the source of the problem is some QT library. Rosegarden crashes when I press the Stop or pause buttons after a long recording so it could be a QT widget issue after a period of time. Otherwise My recordings are generally long. About 1/2 hour sessions. The crash only happens on long clips. Short clips dont crash. I will make a video of the crash and then see if it happens with a newer version. The version I currently use is the latest for my Distro, but I will upgrade to the newest unless there are unresolvable dependency problems. Rosegarden: 17.2 :~$ cat /etc/*lease* NAME="MX" VERSION="18 (Continuum)" ID="mx" VERSION_ID="18" PRETTY_NAME="MX 18 (Continuum)" ANSI_COLOR="0;34" HOME_URL="https://mxlinux.org; BUG_REPORT_URL="https://mxlinux.org; PRETTY_NAME="MX 18.3 Continuum" DISTRIB_ID=MX DISTRIB_RELEASE=18.3 DISTRIB_CODENAME=Continuum DISTRIB_DESCRIPTION="MX 18.3 Continuum" PRETTY_NAME="Debian GNU/Linux 9 (stretch)" NAME="Debian GNU/Linux" VERSION_ID="9" VERSION="9 (stretch)" VERSION_CODENAME=stretch ID=debian HOME_URL="https://www.debian.org/; SUPPORT_URL="https://www.debian.org/support; BUG_REPORT_URL="https://bugs.debian.org/; On 2020-08-11 04:42, MST wrote: Hi there! Of course this is sad and an unwanted behaviour of the software (if it is the software). But I want to point out that version 17.12 is also very old. Which system are you running Rosegarden 17.12 with? ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user -- David W. Jones gn...@hawaii.rr.com authenticity, honesty, community http://dancingtreefrog.com "My password is the last 8 digits of π." ___ 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] Crucial: Help needed with Rosegarden Crashing:
I appreciate your detailed instructions: It is really appreciated. Once all options are depleted and the most recent version does it too I will follow your debug guidelines to the T and save the procedure for future use. I mostly did embedded and numerical programming through my life and not much gui and higher level languages. I should be able to adapt fast and find the problem .. IF it is not something local here triggering it. I still suspect QT as the source of my problem. We'll see. You are all good people. I appreciate it much. On 2020-08-11 00:05, D. Michael McIntyre wrote: On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: Rosegarden crashes deleting my recording in the process. This is painful. Those ideas wont come around again. Lost forever. I very definitely feel your pain here. It happened to me more times than I want to remember. So much creativity right out the window. One of the most frustrating things we used to have was this stupid composition of preset length. You're playing along, in the pocket, really coming up with great ideas, and then WHAM, it just stops recording! I got that to go away a long time ago, but the memories of all those notes still linger. Anyway, for the most part, Rosegarden is vastly more stable than it used to be. Thanks in great measure to Ted Felix coming along, and making diligent and methodical efforts to improve this sort of thing. He was generally very successful, but I guess we've experienced some kind of regression. All you can really do is start using a debug build from an environment that allows core dumps. TRIMMED From original>> ___ 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] Crucial: Help needed with Rosegarden Crashing:
All just remember I am not blaming Rosegarden. This is not a rant or badmouthing Rosegarden. Rosegarden is free Magic software, unreplaceable for me. Since this doesnt happen to those who responded, there must be something locally here that triggers the crash and I think I know what the culprit might be. I really STRONGLY suspect QT to be the problem. If QT induces a crash and take Ropsegarden with it, then Rosegarden was stopped without a SIGINT and just crashed by QT. Since there was not at least a SIGINT, all data in memory will be lost and since it seems Rosegarden do not write to file and ljkeep the recording in memory only writing upon save button, all will be lost. I have had similar issues since QT was owned by Troltec decade or so ago. I actually stopped developing with QT long ago due to similar stability issues. I am pretty sure the source of the problem is some QT library. Rosegarden crashes when I press the Stop or pause buttons after a long recording so it could be a QT widget issue after a period of time. Otherwise My recordings are generally long. About 1/2 hour sessions. The crash only happens on long clips. Short clips dont crash. I will make a video of the crash and then see if it happens with a newer version. The version I currently use is the latest for my Distro, but I will upgrade to the newest unless there are unresolvable dependency problems. Rosegarden: 17.2 :~$ cat /etc/*lease* NAME="MX" VERSION="18 (Continuum)" ID="mx" VERSION_ID="18" PRETTY_NAME="MX 18 (Continuum)" ANSI_COLOR="0;34" HOME_URL="https://mxlinux.org; BUG_REPORT_URL="https://mxlinux.org; PRETTY_NAME="MX 18.3 Continuum" DISTRIB_ID=MX DISTRIB_RELEASE=18.3 DISTRIB_CODENAME=Continuum DISTRIB_DESCRIPTION="MX 18.3 Continuum" PRETTY_NAME="Debian GNU/Linux 9 (stretch)" NAME="Debian GNU/Linux" VERSION_ID="9" VERSION="9 (stretch)" VERSION_CODENAME=stretch ID=debian HOME_URL="https://www.debian.org/; SUPPORT_URL="https://www.debian.org/support; BUG_REPORT_URL="https://bugs.debian.org/; On 2020-08-11 04:42, MST wrote: Hi there! Of course this is sad and an unwanted behaviour of the software (if it is the software). But I want to point out that version 17.12 is also very old. Which system are you running Rosegarden 17.12 with? ___ 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] Crucial: Help needed with Rosegarden Crashing:
That is in fact exactly what I ended up doing. I wrote a program to record midi commands and is busy to integrate it to a pad controller so I have the pad controller directly at my keyboard. The pads are the recorded clop[s and first pad is always open to be recorded. When flashing it records when not it is saved and everything is moved down the 64 pads and pad one is free again etc. I finished it last night after I reverse engineered the pad controller midi commands. With the lookup table the implementation was easy. I then dont use Rosegarden for recording at all, and just use it for editing. I had to do it as this is the third time I lost important recordings. There is no better editor than Rosegarden for me. If it wasnt for the recording issue I would give it 100% I will make a video of a crash so you see it is real. I will also make a video of my solution. I wont mind making my program available but it will only work for one pad controller, so it is not really generally useful as you would need that pad controller else it will be clunky mouse work and i dont want to support software. Anyway you are spot on with your suggestion. It is exactly what I ended up doing. On 2020-08-11 03:35, David W. Jones wrote: Hey, Michael! Long time no hear but yeah 80 hour weeks are killers... Ideas for our original poster to possibly produce an alternative 'backup' recording. If recording audio, start a jack_record command line session with the "-jt" option. When you start Rosegarden recording, jack_record will also be recording the audio. If recording MIDI, have an amidi command line session running with the "-r" option followed by a filename? That will record a file containing same midi events as Rosegarden. If Rosegarden crashes, it won't affect what those have recorded unless it takes down the whole computer. On August 10, 2020 6:05:22 PM HST, "D. Michael McIntyre" wrote: On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: ___ 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] Crucial: Help needed with Rosegarden Crashing:
Hi guys, I totally agree with what Mr. Jones has said. It works for me too. Been doing that since early days decades ago. Stay safe all. Best regards, Rich > > > Hey, Michael! Long time no hear but yeah 80 hour weeks are killers... > > Ideas for our original poster to possibly produce an alternative 'backup' > recording. > > If recording audio, start a jack_record command line session with the > "-jt" option. When you start Rosegarden recording, jack_record will also > be recording the audio. > > If recording MIDI, have an amidi command line session running with the > "-r" option followed by a filename? That will record a file containing > same midi events as Rosegarden. > > If Rosegarden crashes, it won't affect what those have recorded unless it > takes down the whole computer. > > On August 10, 2020 6:05:22 PM HST, "D. Michael McIntyre" wrote: >>On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: >> >>> Rosegarden crashes deleting my recording in the process. >>> >>> This is painful. Those ideas wont come around again. Lost forever. >> >>I very definitely feel your pain here. It happened to me more times >>than >>I want to remember. So much creativity right out the window. One of the >>most frustrating things we used to have was this stupid composition of >>preset length. You're playing along, in the pocket, really coming up >>with great ideas, and then WHAM, it just stops recording! I got that to >>go away a long time ago, but the memories of all those notes still >>linger. >> >>Anyway, for the most part, Rosegarden is vastly more stable than it >>used >>to be. Thanks in great measure to Ted Felix coming along, and making >>diligent and methodical efforts to improve this sort of thing. He was >>generally very successful, but I guess we've experienced some kind of >>regression. >> >>All you can really do is start using a debug build from an environment >>that allows core dumps. >> >>The usual thing is to start running from development source, from a >>debug environment. It's not absolutely necessary to run from >>development >>source, but it will help you in the long run. If the crash is >>repeatable >>(we hope it is!) and somebody (Ted Felix most likely) fixes it, then >>you will want to be in a position to test the result. >> >>Here are the instructions from the wiki: >> >>How to get devel source: >>https://www.rosegardenmusic.com/wiki/development_from_svn >> >>How to get a stack trace for a crash >> >>First, make sure you are running a version of rosegarden that was built >> >>with debugging turned on. >> >>-DCMAKE_BUILD_TYPE=Debug >> >>Without debugging, there will be no symbols in the binary, and the >>backtrace will be useless. You'll likely need to build rosegarden on >>your own to get a debug version. Instructions can be found here: Using >>the Eclipse IDE to work on Rosegarden >> >>Open a terminal window, and check to ensure that applications will be >>able to produce core dumps. The exact command and syntax may vary from >>shell to shell, but for bash it is ulimit -a: >> >>$ ulimit -a >>core file size (blocks, -c) 0 >>data seg size (kbytes, -d) unlimited >>scheduling priority (-e) 20 >>... >> >>The above example is quite typical for an end-user desktop system. >>Having the âcore file sizeâ set to 0 prevents the creation of very >>large >>core dump files in unexpected places, and is generally a good thing. >>However, this also prevents you from generating a stack trace. You need >> >>to change the limit to something larger than 0, but many systems >>prevent >>you from setting this to unlimited, so we suggest >> >>$ ulimit -c 100 >> >>Now start a debug version of rosegarden from the command line, and >>reproduce the crash. You should now have a core file in your current >>directory. The core file is either named core or core.. >> >>Run gdb: >> >>$ gdb rosegarden >> >>Then once you get the gdb prompt, use the command 'bt' to get the stack >> >>trace, and mail it to the authors or to the Rosegarden development >>mailing list, or include it in a bug report. >> >> >>Anyway, I hope this helps. I haven't been involved with development for >> >>quite a number of years now, and I'm semi-retired. I still monitor >>things, and still care. It's just my life went in a different >>direction, >>I guess. Especially in terms of working hours. When I was most >>passionate about Rosegarden, and most active, I had the luxury of >>making >>a decent living working only 45 hours per week. Now it's closer to 80 >>with my commute. >> >>Anyhoo brother, if I could turn back the hands of time and save your >>ideas from getting obliterated, I would. I really hate it for you. So >>much so that you inspired me to write the longest message I've posted >>in >>probably going on 10 years. >> >>-- >>D. Michael McIntyre >> >> >>___ >>Rosegarden-user mailing list >>Rosegarden-user@lists.sourceforge.net - use the link below to
Re: [Rosegarden-user] Crucial: Help needed with Rosegarden Crashing:
Hi there! Of course this is sad and an unwanted behaviour of the software (if it is the software). But I want to point out that version 17.12 is also very old. Which system are you running Rosegarden 17.12 with? Gesendet: Dienstag, 11. August 2020 um 02:37 Uhr Von: liebre...@grossmann-venter.com An: rosegarden-user@lists.sourceforge.net Betreff: [Rosegarden-user] Crucial: Help needed with Rosegarden Crashing: Version 17.12 I reported this a while ago but did not get conclusive advice.It now start to hurt me badly. Just recorded a piano improvisation with amazing changes and when I clicked the stop button on rosegarden when recording the track Rosegarden crashes deleting my recording in the process. This is painful. Those ideas wont come around again. Lost forever. I dont blame Rosegarden for this, it is free software, but I want to use the opportunity to find the bug that causes this if a developer is willing to direct me. When Rosegarden works bug-free it is one of the most amazing pieces of software around. How do I switch on debug so I can catch the crash error so we can get to the bottom of this. ? Thanks ___ 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] Crucial: Help needed with Rosegarden Crashing:
Hey, Michael! Long time no hear but yeah 80 hour weeks are killers... Ideas for our original poster to possibly produce an alternative 'backup' recording. If recording audio, start a jack_record command line session with the "-jt" option. When you start Rosegarden recording, jack_record will also be recording the audio. If recording MIDI, have an amidi command line session running with the "-r" option followed by a filename? That will record a file containing same midi events as Rosegarden. If Rosegarden crashes, it won't affect what those have recorded unless it takes down the whole computer. On August 10, 2020 6:05:22 PM HST, "D. Michael McIntyre" wrote: >On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: > >> Rosegarden crashes deleting my recording in the process. >> >> This is painful. Those ideas wont come around again. Lost forever. > >I very definitely feel your pain here. It happened to me more times >than >I want to remember. So much creativity right out the window. One of the >most frustrating things we used to have was this stupid composition of >preset length. You're playing along, in the pocket, really coming up >with great ideas, and then WHAM, it just stops recording! I got that to >go away a long time ago, but the memories of all those notes still >linger. > >Anyway, for the most part, Rosegarden is vastly more stable than it >used >to be. Thanks in great measure to Ted Felix coming along, and making >diligent and methodical efforts to improve this sort of thing. He was >generally very successful, but I guess we've experienced some kind of >regression. > >All you can really do is start using a debug build from an environment >that allows core dumps. > >The usual thing is to start running from development source, from a >debug environment. It's not absolutely necessary to run from >development >source, but it will help you in the long run. If the crash is >repeatable >(we hope it is!) and somebody (Ted Felix most likely) fixes it, then >you will want to be in a position to test the result. > >Here are the instructions from the wiki: > >How to get devel source: >https://www.rosegardenmusic.com/wiki/development_from_svn > >How to get a stack trace for a crash > >First, make sure you are running a version of rosegarden that was built > >with debugging turned on. > >-DCMAKE_BUILD_TYPE=Debug > >Without debugging, there will be no symbols in the binary, and the >backtrace will be useless. You'll likely need to build rosegarden on >your own to get a debug version. Instructions can be found here: Using >the Eclipse IDE to work on Rosegarden > >Open a terminal window, and check to ensure that applications will be >able to produce core dumps. The exact command and syntax may vary from >shell to shell, but for bash it is ulimit -a: > >$ ulimit -a >core file size (blocks, -c) 0 >data seg size (kbytes, -d) unlimited >scheduling priority (-e) 20 >... > >The above example is quite typical for an end-user desktop system. >Having the “core file size” set to 0 prevents the creation of very >large >core dump files in unexpected places, and is generally a good thing. >However, this also prevents you from generating a stack trace. You need > >to change the limit to something larger than 0, but many systems >prevent >you from setting this to unlimited, so we suggest > >$ ulimit -c 100 > >Now start a debug version of rosegarden from the command line, and >reproduce the crash. You should now have a core file in your current >directory. The core file is either named core or core.. > >Run gdb: > >$ gdb rosegarden > >Then once you get the gdb prompt, use the command 'bt' to get the stack > >trace, and mail it to the authors or to the Rosegarden development >mailing list, or include it in a bug report. > > >Anyway, I hope this helps. I haven't been involved with development for > >quite a number of years now, and I'm semi-retired. I still monitor >things, and still care. It's just my life went in a different >direction, >I guess. Especially in terms of working hours. When I was most >passionate about Rosegarden, and most active, I had the luxury of >making >a decent living working only 45 hours per week. Now it's closer to 80 >with my commute. > >Anyhoo brother, if I could turn back the hands of time and save your >ideas from getting obliterated, I would. I really hate it for you. So >much so that you inspired me to write the longest message I've posted >in >probably going on 10 years. > >-- >D. Michael McIntyre > > >___ >Rosegarden-user mailing list >Rosegarden-user@lists.sourceforge.net - use the link below to >unsubscribe >https://lists.sourceforge.net/lists/listinfo/rosegarden-user --- David W. Jones gn...@hawaii.rr.com authenticity, honesty, community http://dancingtreefrog.com Sent from my Android device with F/LOSS K-9 Mail. ___
Re: [Rosegarden-user] Crucial: Help needed with Rosegarden Crashing:
On 8/10/20 8:37 PM, liebre...@grossmann-venter.com wrote: Rosegarden crashes deleting my recording in the process. This is painful. Those ideas wont come around again. Lost forever. I very definitely feel your pain here. It happened to me more times than I want to remember. So much creativity right out the window. One of the most frustrating things we used to have was this stupid composition of preset length. You're playing along, in the pocket, really coming up with great ideas, and then WHAM, it just stops recording! I got that to go away a long time ago, but the memories of all those notes still linger. Anyway, for the most part, Rosegarden is vastly more stable than it used to be. Thanks in great measure to Ted Felix coming along, and making diligent and methodical efforts to improve this sort of thing. He was generally very successful, but I guess we've experienced some kind of regression. All you can really do is start using a debug build from an environment that allows core dumps. The usual thing is to start running from development source, from a debug environment. It's not absolutely necessary to run from development source, but it will help you in the long run. If the crash is repeatable (we hope it is!) and somebody (Ted Felix most likely) fixes it, then you will want to be in a position to test the result. Here are the instructions from the wiki: How to get devel source: https://www.rosegardenmusic.com/wiki/development_from_svn How to get a stack trace for a crash First, make sure you are running a version of rosegarden that was built with debugging turned on. -DCMAKE_BUILD_TYPE=Debug Without debugging, there will be no symbols in the binary, and the backtrace will be useless. You'll likely need to build rosegarden on your own to get a debug version. Instructions can be found here: Using the Eclipse IDE to work on Rosegarden Open a terminal window, and check to ensure that applications will be able to produce core dumps. The exact command and syntax may vary from shell to shell, but for bash it is ulimit -a: $ ulimit -a core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 20 ... The above example is quite typical for an end-user desktop system. Having the “core file size” set to 0 prevents the creation of very large core dump files in unexpected places, and is generally a good thing. However, this also prevents you from generating a stack trace. You need to change the limit to something larger than 0, but many systems prevent you from setting this to unlimited, so we suggest $ ulimit -c 100 Now start a debug version of rosegarden from the command line, and reproduce the crash. You should now have a core file in your current directory. The core file is either named core or core.. Run gdb: $ gdb rosegarden Then once you get the gdb prompt, use the command 'bt' to get the stack trace, and mail it to the authors or to the Rosegarden development mailing list, or include it in a bug report. Anyway, I hope this helps. I haven't been involved with development for quite a number of years now, and I'm semi-retired. I still monitor things, and still care. It's just my life went in a different direction, I guess. Especially in terms of working hours. When I was most passionate about Rosegarden, and most active, I had the luxury of making a decent living working only 45 hours per week. Now it's closer to 80 with my commute. Anyhoo brother, if I could turn back the hands of time and save your ideas from getting obliterated, I would. I really hate it for you. So much so that you inspired me to write the longest message I've posted in probably going on 10 years. -- D. Michael McIntyre ___ 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] Crucial: Help needed with Rosegarden Crashing:
Version 17.12 I reported this a while ago but did not get conclusive advice.It now start to hurt me badly. Just recorded a piano improvisation with amazing changes and when I clicked the stop button on rosegarden when recording the track Rosegarden crashes deleting my recording in the process. This is painful. Those ideas wont come around again. Lost forever. I dont blame Rosegarden for this, it is free software, but I want to use the opportunity to find the bug that causes this if a developer is willing to direct me. When Rosegarden works bug-free it is one of the most amazing pieces of software around. How do I switch on debug so I can catch the crash error so we can get to the bottom of this. ? Thanks ___ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user
