Jack 0.51 and PD 0.36, both from planet CCRMA, seem to work fine together.
However, since I was having problems with rosegaren-4 and Jack (frooze at least the UI), I built Jack 0.61.0 from the tarball at the Jack site.
Now rosegarden starts up fine, but PD doesn't.
Following is a shell session showing the problem. I just thought I'd post this in case anyone had any ideas.
Regards
Larry Troxler
~ ~ ~
root@dhcp-25-2 ~]$ jackd -R -d alsa -d hw:0 -r 44100 -p 512 -P & [1] 7518 jackd 0.61.0 Copyright 2001-2002 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details
[root@dhcp-25-2 ~]$ loading driver .. driver client name = [alsa_pcm] creating alsa driver ... hw:0|512|2|44100|swmon|rt delivering event 2 to IP client alsa_pcm delivering event 5 to IP client alsa_pcm starting engine
[root@dhcp-25-2 ~]$ pd -r 44100 -jack -outchannels 2 -audiobuf 20 -rt & [2] 7525 reading startup file: /root/.pdrc Testing for Jack cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) [root@dhcp-25-2 ~]$ using OSS audio interface priority 96 scheduling enabled. memory locking enabled. priority 98 scheduling enabled. memory locking enabled.
((((((((((((((()))))))))))))))
( scheme is here )
((((((((((((((()))))))))))))))
loading /root/pd/pdinit.scm loading /root/dl/siod/siod.scm ;; Optional Runtime Library for Release 3.0 done. loading /root/pd/pdsite.scm loading /root/pd/pdexampl.scm done. done. done.
@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@...the zexy external1.1...@
@ forum::f?r::uml?ute 2001 @
@ send me a 'help' message @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@
**** alsa_pcm: xrun of at least 31.762 msecs
audio I/O stuck... closing audio
~ ~ ~
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Music Technology Group - UPF, Barcelona http://www.iua.upf.es/mtg/eng/ AGNULA - A GNU Linux Audio distribution <www.agnula.org> DeMuDi - Debian Multimedia Distribution <www.demudi.org>
On Sat, 15 Mar 2003, Larry Troxler wrote:
Jack 0.51 and PD 0.36, both from planet CCRMA, seem to work fine together.
However, since I was having problems with rosegaren-4 and Jack (frooze at least the UI), I built Jack 0.61.0 from the tarball at the Jack site.
Now rosegarden starts up fine, but PD doesn't.
Following is a shell session showing the problem. I just thought I'd post this in case anyone had any ideas.
Regards
Larry Troxler
~ ~ ~
root@dhcp-25-2 ~]$ jackd -R -d alsa -d hw:0 -r 44100 -p 512 -P & [1] 7518 jackd 0.61.0 Copyright 2001-2002 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details
[root@dhcp-25-2 ~]$ loading driver .. driver client name = [alsa_pcm] creating alsa driver ... hw:0|512|2|44100|swmon|rt delivering event 2 to IP client alsa_pcm delivering event 5 to IP client alsa_pcm starting engine
[root@dhcp-25-2 ~]$ pd -r 44100 -jack -outchannels 2 -audiobuf 20 -rt & [2] 7525 reading startup file: /root/.pdrc Testing for Jack cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) [root@dhcp-25-2 ~]$ using OSS audio interface priority 96 scheduling enabled. memory locking enabled. priority 98 scheduling enabled. memory locking enabled.
((((((((((((((())))))))))))))) ( scheme is here ) ((((((((((((((()))))))))))))))loading /root/pd/pdinit.scm loading /root/dl/siod/siod.scm ;; Optional Runtime Library for Release 3.0 done. loading /root/pd/pdsite.scm loading /root/pd/pdexampl.scm done. done. done.
@@@@@@@@@@@@@@@@@@@@@@@@@@@@ @...the zexy external1.1...@ @ forum::f?r::uml?ute 2001 @ @ send me a 'help' message @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@**** alsa_pcm: xrun of at least 31.762 msecs
audio I/O stuck... closing audio
~ ~ ~
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
Hallo, guenter geiger hat gesagt: // guenter geiger wrote:
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ?
At least I had this problem, which also killed jackd. Temporarily I'm now using a version of Pd on my laptop, where jack_connect in s_linux.c return's 0 immediatly.
ciao
Frank Barknecht _ ______footils.org__
Hi I got latest Jack (yesterdays CVS) , alsa rc8a. Everithing works fine (ardour,freqtweak,tappir, pressgang) besides qjackconnect. I use pd_devel from cvs a couple of weeks ago. Tried to get the newer one, 2 days ago from CVS but it wasn't working wth JACK.I haven't tried today though. Hope that helps -- regards anton
On Tue, 18 Mar 2003, Antonis Galanopoulos wrote:
Hi I got latest Jack (yesterdays CVS) , alsa rc8a. Everithing works fine (ardour,freqtweak,tappir, pressgang) besides qjackconnect. I use pd_devel from cvs a couple of weeks ago. Tried to get the newer one, 2 days ago from CVS but it wasn't working wth JACK.I haven't tried today though. Hope that helps
No :(
It works on my system, and it seems to work on several other systems too. It is hard to say what went wrong without additonal information What does pd say when it starts ? Did you check out the devel_0_36 version ? Did you compile with --enable-jack ?
.. I know there are already too many instructions to get this going. What do people think about makeing the devel_0_36 branch the main branch in CVS ? Also, we could autodetect the existance of the jack library, so the --enable-jack is not necessary anymore.
Greetings, Guenter
-- regards anton
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
No :(
Ok I ll try to be more analycal...
I got pd_devel from cvs 10 minutes ago, Still not working.
It works on my system, and it seems to work on several other systems too. It is hard to say what went wrong without additonal information What does pd say when it starts ?
thats what I get when it starts:
ear:/home/anton/Desktop/final_practical_project # pd -jack -rt -outchannels 8 /home/anton/Desktop/final_practical_project/controller2.pd reading startup file: /root/.pdrc Testing for Jack
*** libjack: WARNING! Use of function jack_set_buffer_size_callback() is deprecated! ***
jack: sample rate 48000/sec using JACK audio interface priority 98 scheduling enabled. memory locking enabled. priority 96 scheduling enabled. memory locking enabled.
and at the end: engine has shut down socket; thread exiting pd_gui: pd process exited
I also tried with simpler patches because the one I work uses 8 jack outputs. No difference however.
Did you check out the devel_0_36 version ?
Yes thats what I use for the past 4 months
Did you compile with --enable-jack ?
I did, I also tried --enable-jack and --enable-alsa together same results
.. I know there are already too many instructions to get this going. What do people think about makeing the devel_0_36 branch the main branch in CVS ?
As I said I already use the devel branch for the jack functionality. It has improved very much the last few months .
Also, we could autodetect the existance of the jack library, so the --enable-jack is not necessary anymore.
For the devel to compile, i need to manually create obj and bin directories and remove the Werror from the makefiles. Would be nice to have them fixed as well. Although it is easy to do that manualy, is a pain until you figure it out.
i hope now I am more helpfull Feel free to contact me if I can do anything more to help -- regards anton
On Tuesday 18 March 2003 13:53, Antonis Galanopoulos wrote:
For the devel to compile, i need to manually create obj and bin directories and remove the Werror from the makefiles. Would be nice to have them fixed as well. Although it is easy to do that manualy, is a pain until you figure it out.
Thanks Antonis for that info. It indeed probably saved me a lot of time, since I just now ran into these problems and had already seen your message.
Larry Troxler
On Tue, 18 Mar 2003, Antonis Galanopoulos wrote:
thats what I get when it starts:
ear:/home/anton/Desktop/final_practical_project # pd -jack -rt -outchannels 8 /home/anton/Desktop/final_practical_project/controller2.pd reading startup file: /root/.pdrc Testing for Jack
*** libjack: WARNING! Use of function jack_set_buffer_size_callback() is deprecated! ***
jack: sample rate 48000/sec using JACK audio interface priority 98 scheduling enabled. memory locking enabled. priority 96 scheduling enabled. memory locking enabled.
and at the end: engine has shut down socket; thread exiting pd_gui: pd process exited
I also tried with simpler patches because the one I work uses 8 jack outputs. No difference however.
Doesn't look good. It seems that pd just crashes. Have you tried starting pd without a patch and load the testtone ?
Also, at this stage I would probably need your system information, and contents of your .pdrc file.
I am still not able to reproduce it, anyone else has the same problem ?
Greetings,
Guenter
Did you check out the devel_0_36 version ?
Yes thats what I use for the past 4 months
Did you compile with --enable-jack ?
I did, I also tried --enable-jack and --enable-alsa together same results
.. I know there are already too many instructions to get this going. What do people think about makeing the devel_0_36 branch the main branch in CVS ?
As I said I already use the devel branch for the jack functionality. It has improved very much the last few months .
Also, we could autodetect the existance of the jack library, so the --enable-jack is not necessary anymore.
For the devel to compile, i need to manually create obj and bin directories and remove the Werror from the makefiles. Would be nice to have them fixed as well. Although it is easy to do that manualy, is a pain until you figure it out.
i hope now I am more helpfull Feel free to contact me if I can do anything more to help -- regards anton
Hi all,
Doesn't look good. It seems that pd just crashes. Have you tried starting pd without a patch and load the testtone ?
Got todays devel_0_36 from cvs pd -alsa works with my complex patch or without pd -jack also works with my complex patch or without pd -rt -jack or pd -rt doesnt work with my patch but works with others ( ie testtone) and othe simple patches my patch crashes with the usual very informative: engine has shut down socket; thread exiting pd_gui: pd process exited
Also, at this stage I would probably need your system information, and contents of your .pdrc file.
System info suse 8.1 gcc version 3.2 2.4.20-ck3 kernel with preempt and lowlat among others alsa 0.9.0rc8a pd_devel from cvs this morning jack cvs two days ago emu10k1 based soundcard
My .pdrc looks like that -lib /usr/local/lib/pd/externs/maxlib -lib /usr/local/lib/pd/externs/iemlib1 -lib /usr/local/lib/pd/externs/iemlib2 -lib /usr/local/lib/pd/externs/iem_t3_lib -lib /usr/local/lib/pd/externs/iem_mp3 -lib /usr/local/lib/pd/extra/xsample -path /usr/local/lib/pd/iemabs
However this works fine pd -outchannels 8 controller2.pd but this: pd -rt -outchannels 8 controller2.pd crashes. commenting the iem _mp3 allows me to start my patch with -rt but many externals cannot be found, for example the xsample~ which has nothing to do with iem_mp3 IMHO Ofcource I run it as root user
I am still not able to reproduce it, anyone else has the same problem ?
I dont know what to say. I looked at it for the past few hours. Maybe is something to do with my particular patch. But it works without realtime priority . I attach the patch, in case somebody whants to have a look. It uses a couple of Frank Barknecht's externals(sseg16,angriff ) and abstractions (pecycle). I stripped them down but still the same.
I will also try to upgrade to the latest 0.9.1 alsa when I get some time.
Another thing to say is that the same patch works great with an older cvs version of devel_0_36 with realtime.
-- regards anton
Hallo,
Antonis Galanopoulos hat gesagt: // Antonis Galanopoulos wrote:
Got todays devel_0_36 from cvs pd -alsa works with my complex patch or without pd -jack also works with my complex patch or without pd -rt -jack or pd -rt doesnt work with my patch but works with others ( ie testtone) and othe simple patches my patch crashes with the usual very informative: engine has shut down socket; thread exiting pd_gui: pd process exited
Can you see, if jackd ist still running? If not, then this is the problem I reported last thursday or friday: Pd -realtime can take too long to load a complex patch and a lot of libraries and this seems to just kill jackd.
I needed to get this running for my Pd-talk on the LAD conference and I still cannot use Pd with the Quattro on my notebook directly (-alsa), so I *needed* -jack on saturday running with a complex patch.
My solution was, to hack s_linux.c and disable the jack_connect functionality at all. Then the complex patch would load, and atfer everything is set up, I would jack_connect Pd by hand (or with a small shell script) I still would prefer a startup option like "-jack-connect 0/1/alsa_pcm_out_1".
I am starting jackd with jackstart and capabilities. I could figure, that Pd gets a higher priority than jackd and thus simply kills the daemon.
ciao
Frank Barknecht _ ______footils.org__
On Wednesday 19 March 2003 15:13, Frank Barknecht wrote:
Can you see, if jackd ist still running? If not, then this is the problem I reported last thursday or friday: Pd -realtime can take too long to load a complex patch and a lot of libraries and this seems to just kill jackd.
I needed to get this running for my Pd-talk on the LAD conference and I still cannot use Pd with the Quattro on my notebook directly (-alsa), so I *needed* -jack on saturday running with a complex patch.
My solution was, to hack s_linux.c and disable the jack_connect functionality at all. Then the complex patch would load, and atfer everything is set up, I would jack_connect Pd by hand (or with a small shell script) I still would prefer a startup option like "-jack-connect 0/1/alsa_pcm_out_1".
Well, first I have some good news from my camp - Paul Davis announced a new CVS commit a couple of days ago on the jack-devel list:
http://sourceforge.net/mailarchive/forum.php?thread_id=1853444&forum_id=...
... and this new version seems to generally solve most of the problems I was reporting on the list earlier. Now, running PD doesn't give any errors apart from the DIO error light being lit on the test patch.
But there is something strange going on that I can see, even before I begin to troubleshoot this:
I think I now understand the problem you (Frank) have, and the reason you hacked the jack_connect functionality. It seems (If I understand correctly, and that we are indeed talking about the same problem), that PD sees fit to connect its Jack client to the DAC jack client (in my case, ALSA) on startup.
At least, after I start Jack, then start PD, then run qjackconnect (which, BTW did _not_ need to be recompiled to use with this Jack snapshot I got - the Planet CCRMA binary works fine!), I see that the Jack PD clients (two of them- one for each channel I suppose) have been connected to the the JACK alsa clients. So it seems that PD does indeed connect itself on startup. Is there a way to disable this? Certainly I don't want it to connect to the soundcard through Jack, because if that were the case, there would be no point in using Jack at all. Perhaps I am missing some understanding about the Jack concept.
Is there a way to make more Jack clients avalable to PD, so that one audio output for example, goes to the DAC, and another one goes to an audio recorder? If so, (1) how is this specified on the PD command line, and (2) how would you decide from PD, which jack client audio input or output you want to connect to, in a given box? (For example, dac~ just takes a number of signal inputs equal to the number of output channels - how would you tell PD which Jack client the DAC stream should be sent to?)
I remember reading about a jack~ box for PD - is this what you would use if you want multiple audio inputs/outputs?
I apologize in advance for the confused questions :-)
Larry
Hallo, Larry Troxler hat gesagt: // Larry Troxler wrote:
At least, after I start Jack, then start PD, then run qjackconnect (which, BTW did _not_ need to be recompiled to use with this Jack snapshot I got - the Planet CCRMA binary works fine!), I see that the Jack PD clients (two of them- one for each channel I suppose) have been connected to the the JACK alsa clients. So it seems that PD does indeed connect itself on startup. Is there a way to disable this?
Not currently, that's why I disabled jack_connect in s_linux.c at all.
Certainly I don't want it to connect to the soundcard through Jack, because if that were the case, there would be no point in using Jack at all. Perhaps I am missing some understanding about the Jack concept.
Well, something *must* connect to the soundcard in the end, otherwise no sound would be produced.
Is there a way to make more Jack clients avalable to PD, so that one audio output for example, goes to the DAC, and another one goes to an audio recorder? If so, (1) how is this specified on the PD command line,
$ pd -jack -channel 12 or $ pd -jack -inchannels 4 -outchannels 2
After that, you probably have a lot of jack_connect to do.
and (2) how would you decide from PD, which jack client audio input or output you want to connect to, in a given box? (For example, dac~ just takes a number of signal inputs equal to the number of output channels - how would you tell PD which Jack client the DAC stream should be sent to?)
dac~ takes the ids of output channels. So it works like this:
[dac~ 1] ==> pure_data_0:output0 [dac~ 2 4] ==> pure_data_0:output1, pure_data_0:output3 ...
ciao
Frank Barknecht _ ______footils.org__
Hi all
What do people think about makeing the devel_0_36 branch the main branch in CVS ?
that's an interesting question since it incorporates a few other issues:
- how does Miller handle cvs and how does he incorporate the devel branch in
his releases?
- how do we handle experimental changes in the cvs? (still other branches?)
- what about externals? i always considered my cvs contributions as
experimental (e.g. due to a compiler bug awaiting a workaround the vasp cvs version has not been functional for the last few days) but when there are built snapshots i (we?) will have to reconsider that.
hmmm, Thomas
On Tue, 18 Mar 2003, Thomas Grill wrote:
- how do we handle experimental changes in the cvs? (still other branches?)
- what about externals? i always considered my cvs contributions as
experimental (e.g. due to a compiler bug awaiting a workaround the vasp cvs version has not been functional for the last few days) but when there are built snapshots i (we?) will have to reconsider that.
If we start making releases and tagging them in CVS, then its ok that things don't always work in the main branch, as long as they compile. I think the code in any CVS should be considered under development. The releases are for the stable code. If necessary, we could have an experimental branch for things that are really experimental and likely to break things.
Quoting Eric Raymond "Release early and release often", I think its time to start making releases.
.hc
zen
\
\
\[D[D[D[D
On Tue, 18 Mar 2003, Thomas Grill wrote:
Hi all
What do people think about makeing the devel_0_36 branch the main branch in CVS ?
that's an interesting question since it incorporates a few other issues:
- how does Miller handle cvs and how does he incorporate the devel branch in
his releases?
He does not :(. I do not know how many patches get sent to Miller in order to incorporate them. We talked at the ICMC about the topic, and he said that it is a lot of work to check patches for correctness, which is understandable.
Maybe the CVS helps in making this process easier, because things get tested ...
The plan up to now is to have a place to incorporate patches easily, and whenever there is a new pd version, I just merge them ...
Of course, at some point this will be unpractical without folding back patches.
My approach to solve this problem up to now was to ignore it :)
- how do we handle experimental changes in the cvs? (still other branches?)
well, up to now it was all experimental. Making this the main branch would only change things for the user (I think noone checks out the original Miller version from CVS, which is currently the main branch, because this just gets downloaded from his site directly).
If we want to do something really experimental, which would break pd for a certain time, we have to do it in a separate branch, yes.
- what about externals? i always considered my cvs contributions as
experimental (e.g. due to a compiler bug awaiting a workaround the vasp cvs version has not been functional for the last few days) but when there are built snapshots i (we?) will have to reconsider that.
This really depends on how you manage your CVS subtree. If you have some sort of stable version and experimental version in CVS we could make built snapshots with the stable version only.
If it is too much of a hassle for you having branches in there, we could do the snapshots when you say it is ready.
Greetings,
Guenter
On Tuesday 18 March 2003 12:54, guenter geiger wrote:
On Tue, 18 Mar 2003, Antonis Galanopoulos wrote:
Hi I got latest Jack (yesterdays CVS) , alsa rc8a. Everithing works fine (ardour,freqtweak,tappir, pressgang) besides qjackconnect. I use pd_devel from cvs a couple of weeks ago. Tried to get the newer one, 2 days ago from CVS but it wasn't working wth JACK.I haven't tried today though. Hope that helps
No :(
It works on my system, and it seems to work on several other systems too. It is hard to say what went wrong without additonal information What does pd say when it starts ? Did you check out the devel_0_36 version ? Did you compile with --enable-jack ?
Ok, I'm not Antonis, but I have a datapoint for you. This whole issue is frustrating, because everytime I rebuild Jack or Rosegarden or PD, it seems I get a different set of errors. Like I said before, I really don't have much more time to spend on this, especially since PD istself has disk streaming boxes that seems to work fine for me. But in case it's any help, here's a log:
~ ~ ~ begin paste from shell, comments inserted with ';' ~ ~ ~
; jack was built and installed from CVS a day ago, using the default branch.
; I nuked the old libjack and jack.h from planet CCRMA first.
[root@dhcp-25-2 ~]$ ls -l which jackd
-rwxr-xr-x 1 root root 125786 Mar 17 18:56 /usr/local/bin/jackd
[root@dhcp-25-2 ~]$ ldd which jackd
libjack.so.0 => /usr/local/lib/libjack.so.0 (0x40013000)
libm.so.6 => /lib/libm.so.6 (0x4002e000)
libdl.so.2 => /lib/libdl.so.2 (0x4004f000)
libpthread.so.0 => /lib/libpthread.so.0 (0x40052000)
libc.so.6 => /lib/libc.so.6 (0x40082000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
[root@dhcp-25-2 ~]$ cat which J
#!/bin/bash
jackd -R -d alsa -d hw:0 -r 44100 -p 512 -P &
[root@dhcp-25-2 ~]$ J [root@dhcp-25-2 ~]$ jackd 0.62.0 Copyright 2001-2002 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details
loading driver .. driver client name = [alsa_pcm] creating alsa driver ... hw:0|512|2|44100|swmon|rt delivering event 2 to IP client alsa_pcm delivering event 5 to IP client alsa_pcm starting engine
[root@dhcp-25-2 ~]$ pd -version
reading startup file: /root/.pdrc
Pd version 0.36-0
compiled 20:31:06 Mar 18 2003
[root@dhcp-25-2 ~]$ cat which PJ
#!/bin/sh
pd -r 44100 -jack -outchannels 2 -audiobuf 20 $1 $2 $3
[root@dhcp-25-2 ~]$ PJ -rt & [2] 1126 reading startup file: /root/.pdrc Testing for Jack [root@dhcp-25-2 ~]$ *** libjack: WARNING! Use of function jack_set_buffer_size_callback() is deprecated! ***
jack: sample rate 44100/sec jackd: pcm.c:5860: snd_pcm_mmap_commit: Assertion `pcm' failed. received signal 6 during shutdodelivering event 5 to IP client alsa_pcm cannot open specified fifo [/tmp/jack-ack-fifo-1117-1] for writing (No such device or address) bad status for client event handling (type = 5) got connect request cannot open specified fifo [/tmp/jack-ack-fifo-1117-1] for writing (No such device or address) bad status for client event handling (type = 5) delivering event 5 to IP client alsa_pcm delivering event 3 to IP client alsa_pcm got connect request cannot open specified fifo [/tmp/jack-ack-fifo-1117-1] for writing (No such device or address) bad status for client event handling (type = 5) delivering event 5 to IP client alsa_pcm delivering event 3 to IP client alsa_pcm using JACK audio interface priority 98 scheduling enabled. memory locking enabled. priority 96 scheduling enabled. memory locking enabled. watchdog: signaling pd... watchdog: signaling pd... jackd watchdog: timeout - killing jackd received signal 11 during shutdwatchdog: signaling pd... watchdog: signaling pd... watchdog: signaling pd... watchdog: signaling pd...
(and then I had to kill -9 it)
.. I know there are already too many instructions to get this going. What do people think about makeing the devel_0_36 branch the main branch in CVS ?
Can't answer that question, but maybe you can answer one for me - is devel_0_36 really the PD 0.37 development branch? If so, could I humbly ask why it's called devel_0_36? If not, does this mean that the latest PD official release was 0.35? If so, why is there a 0.36 on planet CCRMA? Does that mean the planet CCRMA is jumping the gun, and if so, which snapshot is used for their RPM? Sorry for my confusion, but I'm well, confused :-). I had thought that 0.36 was already released and not still in development. Maybe that's what I'm mistaken about.
Regards
Larry Troxler
Hallo, Larry Troxler hat gesagt: // Larry Troxler wrote:
Can't answer that question, but maybe you can answer one for me - is devel_0_36 really the PD 0.37 development branch? If so, could I humbly ask why it's called devel_0_36? If not, does this mean that the latest PD official release was 0.35? If so, why is there a 0.36 on planet CCRMA? Does that mean the planet CCRMA is jumping the gun, and if so, which snapshot is used for their RPM? Sorry for my confusion, but I'm well, confused :-). I had thought that 0.36 was already released and not still in development. Maybe that's what I'm mistaken about.
You're right, all this might be misleading. Correctly it is read as "devel_0_36 is the developer's branche based on official 0.36." Maybe pd-extended as it s called on the website is a better name.
ciao
Frank Barknecht _ ______footils.org__
Frank Barknecht said this at Wed, 19 Mar 2003 08:35:22 +0100:
You're right, all this might be misleading. Correctly it is read as "devel_0_36 is the developer's branche based on official 0.36." Maybe pd-extended as it s called on the website is a better name.
Hmm. I never intended it as such... I just wanted a more distinctive name to distinguish between my Mac binary distro and the official MSP binaries.
However, it seems like having the CVS snapshot under that name (thanks, Hans-Christoph) is attracting a few people who don't normally use CVS to obtain sources. See the emails from Eric (eskogen) and Tim Blechmann just in the past 12 hours...
So the name might work more generally, yes. Perhaps we'll need to be a bit more principled in how and when we take snapshots in the future, as they do seem to attract a lot of attention.
Cheers, adam
Can't answer that question, but maybe you can answer one for me - is devel_0_36 really the PD 0.37 development branch? If so, could I humbly ask why it's called devel_0_36? If not, does this mean that the latest PD official release was 0.35? If so, why is there a 0.36 on planet CCRMA?
Because the standard "miller" release (which is at this point 0.36, I think) does not include jack support yet. As one of my priorities is to have jack support I packaged cvs, AFAIK devel_0_36 means 0.36 plus tweaks and additions.
Does that mean the planet CCRMA is jumping the gun, and if so, which snapshot is used for their RPM?
Current cvs date of the pd in Planet CCRMA is 20030211.120137 (date.time in US Pacific zone).
-- Fernando
On Mon, 17 Mar 2003, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Oops, sorry. Just realized that the jack connection is setup before loading the libraries, so that can not be the problem. I am currently haveing problems too, using jack 0.51.0 from debian
its really strange,
Going to investigate further ..
Guenter
Guenter
Music Technology Group - UPF, Barcelona http://www.iua.upf.es/mtg/eng/ AGNULA - A GNU Linux Audio distribution <www.agnula.org> DeMuDi - Debian Multimedia Distribution <www.demudi.org>
On Sat, 15 Mar 2003, Larry Troxler wrote:
Jack 0.51 and PD 0.36, both from planet CCRMA, seem to work fine together.
However, since I was having problems with rosegaren-4 and Jack (frooze at least the UI), I built Jack 0.61.0 from the tarball at the Jack site.
Now rosegarden starts up fine, but PD doesn't.
Following is a shell session showing the problem. I just thought I'd post this in case anyone had any ideas.
Regards
Larry Troxler
~ ~ ~
root@dhcp-25-2 ~]$ jackd -R -d alsa -d hw:0 -r 44100 -p 512 -P & [1] 7518 jackd 0.61.0 Copyright 2001-2002 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details
[root@dhcp-25-2 ~]$ loading driver .. driver client name = [alsa_pcm] creating alsa driver ... hw:0|512|2|44100|swmon|rt delivering event 2 to IP client alsa_pcm delivering event 5 to IP client alsa_pcm starting engine
[root@dhcp-25-2 ~]$ pd -r 44100 -jack -outchannels 2 -audiobuf 20 -rt & [2] 7525 reading startup file: /root/.pdrc Testing for Jack cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) [root@dhcp-25-2 ~]$ using OSS audio interface priority 96 scheduling enabled. memory locking enabled. priority 98 scheduling enabled. memory locking enabled.
((((((((((((((())))))))))))))) ( scheme is here ) ((((((((((((((()))))))))))))))loading /root/pd/pdinit.scm loading /root/dl/siod/siod.scm ;; Optional Runtime Library for Release 3.0 done. loading /root/pd/pdsite.scm loading /root/pd/pdexampl.scm done. done. done.
@@@@@@@@@@@@@@@@@@@@@@@@@@@@ @...the zexy external1.1...@ @ forum::f?r::uml?ute 2001 @ @ send me a 'help' message @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@**** alsa_pcm: xrun of at least 31.762 msecs
audio I/O stuck... closing audio
~ ~ ~
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
On Monday 17 March 2003 11:40, guenter geiger wrote:
On Mon, 17 Mar 2003, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Oops, sorry. Just realized that the jack connection is setup before loading the libraries, so that can not be the problem. I am currently haveing problems too, using jack 0.51.0 from debian
its really strange,
Yes, it is strange, because jack 0.5.1.0 worked for me! (but wouldn't work with rosegarden, hence the need to upgrade it)
Larry Troxler
On Monday 17 March 2003 08:59, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Hello Guenter - I tried today's CVS of jack and have the same result. Which version of pd are you using? I'm using pd 0.36-0 directly from the CCRMA binary RPM. (I did try last night to compile PD from CVS but it seems the build is/was broken - failed at fiddle~).
Larry Troxler
Nothing for nothing, but I'm having the same problems. I just attributed it to the fact that I was on a Crusoe processor. I'm using alsa and jack from Planet CCRMA. The sound card is ALi 5451. All jack apps have the same problem.
There was a hint from Paul Davis to the ALSA list rthat this might have something to do with ALSA. I'm too busy with other projects to figure out how to debug device drivers right now, but since I don't need jack until this summer I'm to let it be.
On Mon, 2003-03-17 at 19:08, Larry Troxler wrote:
On Monday 17 March 2003 08:59, guenter geiger wrote:
Hi Larr
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Hello Guenter - I tried today's CVS of jack and have the same result. Which version of pd are you using? I'm using pd 0.36-0 directly from the CCRMA binary RPM. (I did try last night to compile PD from CVS but it seems the build is/was broken - failed at fiddle~).
Larry Troxler
On Mon, 17 Mar 2003, Larry Troxler wrote:
On Monday 17 March 2003 08:59, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Hello Guenter - I tried today's CVS of jack and have the same result. Which version of pd are you using? I'm using pd 0.36-0 directly from the CCRMA binary RPM. (I did try last night to compile PD from CVS but it seems the build is/was broken - failed at fiddle~).
Yes, I am using the version from CVS. I resolved my problem, there was a bug in the jack packages that allowed me to have jackd 0.51 and libjack 0.50 isntalled.
I think your problem comes from binary incompatibilities, so you have to recompile.
Can you report why your compile fails with fiddle~ ?
Actually, if you are at the stage of compiling fiddle there should already be a pd in ../bin/pd Just try that and look what it does.
Greetings,
Guenter
On Tuesday 18 March 2003 05:29, guenter geiger wrote:
On Mon, 17 Mar 2003, Larry Troxler wrote:
On Monday 17 March 2003 08:59, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Hello Guenter - I tried today's CVS of jack and have the same result. Which version of pd are you using? I'm using pd 0.36-0 directly from the CCRMA binary RPM. (I did try last night to compile PD from CVS but it seems the build is/was broken - failed at fiddle~).
Yes, I am using the version from CVS. I resolved my problem, there was a bug in the jack packages that allowed me to have jackd 0.51 and libjack 0.50 isntalled.
I think your problem comes from binary incompatibilities, so you have to recompile.
If this means that Jack is still at the stage that I need to get the source code for every single application that I use that's jack-aware, I think I might need to cut my losses at this point. Who the H*#& has this much free time on their hands?
But to Jack's credit, I don't think anyone on their development team suggested it was ready for general use, so I can't complain.
Can you report why your compile fails with fiddle~ ?
Well unfortunately I didn't save the warnings (they were warnings that halted the build because we use -Werror).
One thing was, is that I didn't realize that the default CVS branch isn't actually the latest version. So I don't know if the fiddle build also fails on on the devel_36_0 version, since when I realized my mistake I checked out that version but hacked the makefile at the same time.
Actually, if you are at the stage of compiling fiddle there should already be a pd in ../bin/pd Just try that and look what it does.
Greetings,
Guenter
On Tue, 18 Mar 2003, Larry Troxler wrote:
If this means that Jack is still at the stage that I need to get the source code for every single application that I use that's jack-aware, I think I might need to cut my losses at this point. Who the H*#& has this much free time on their hands?
:) Exactly, this was the main reason I wanted to do demudi ... so I ended up haveing to deal with these problems all day long ...
But to Jack's credit, I don't think anyone on their development team suggested it was ready for general use, so I can't complain.
Yes, thats true .... And to be fair, it is almost there. At least you just have to recompile and you do not have to change the code like it happend with ALSA at the beginning.
Guenter
Can you report why your compile fails with fiddle~ ?
Well unfortunately I didn't save the warnings (they were warnings that halted the build because we use -Werror).
One thing was, is that I didn't realize that the default CVS branch isn't actually the latest version. So I don't know if the fiddle build also fails on on the devel_36_0 version, since when I realized my mistake I checked out that version but hacked the makefile at the same time.
Actually, if you are at the stage of compiling fiddle there should already be a pd in ../bin/pd Just try that and look what it does.
Greetings,
Guenter
Hello, just an update on this - I just discovered that qjackconnect (which is KDE's jack connect GUI) gives the same problem. So PD is not the problem here.
Larry Troxler
On Monday 17 March 2003 08:59, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Music Technology Group - UPF, Barcelona http://www.iua.upf.es/mtg/eng/ AGNULA - A GNU Linux Audio distribution <www.agnula.org> DeMuDi - Debian Multimedia Distribution <www.demudi.org>
On Sat, 15 Mar 2003, Larry Troxler wrote:
Jack 0.51 and PD 0.36, both from planet CCRMA, seem to work fine together.
However, since I was having problems with rosegaren-4 and Jack (frooze at least the UI), I built Jack 0.61.0 from the tarball at the Jack site.
Now rosegarden starts up fine, but PD doesn't.
Following is a shell session showing the problem. I just thought I'd post this in case anyone had any ideas.
Regards
Larry Troxler
~ ~ ~
root@dhcp-25-2 ~]$ jackd -R -d alsa -d hw:0 -r 44100 -p 512 -P & [1] 7518 jackd 0.61.0 Copyright 2001-2002 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details
[root@dhcp-25-2 ~]$ loading driver .. driver client name = [alsa_pcm] creating alsa driver ... hw:0|512|2|44100|swmon|rt delivering event 2 to IP client alsa_pcm delivering event 5 to IP client alsa_pcm starting engine
[root@dhcp-25-2 ~]$ pd -r 44100 -jack -outchannels 2 -audiobuf 20 -rt & [2] 7525 reading startup file: /root/.pdrc Testing for Jack cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) [root@dhcp-25-2 ~]$ using OSS audio interface priority 96 scheduling enabled. memory locking enabled. priority 98 scheduling enabled. memory locking enabled.
((((((((((((((())))))))))))))) ( scheme is here ) ((((((((((((((()))))))))))))))loading /root/pd/pdinit.scm loading /root/dl/siod/siod.scm ;; Optional Runtime Library for Release 3.0 done. loading /root/pd/pdsite.scm loading /root/pd/pdexampl.scm done. done. done.
@@@@@@@@@@@@@@@@@@@@@@@@@@@@ @...the zexy external1.1...@ @ forum::f?r::uml?ute 2001 @ @ send me a 'help' message @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@**** alsa_pcm: xrun of at least 31.762 msecs
audio I/O stuck... closing audio
~ ~ ~
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
On Monday 17 March 2003 20:01, Larry Troxler wrote:
Hello, just an update on this - I just discovered that qjackconnect (which is KDE's jack connect GUI) gives the same problem. So PD is not the problem here.
Larry Troxler
I wonder if Jack changed it's API between version 0.5x and 0.6x? I'm having no luck now with any applications using jack 0.6x.
The Jack website is very vague on this subject - it says only that now with 0.61 the Jack API is stable. Does this mean that something changed between 0.51 an 0.61, or not?
Larry
On Monday 17 March 2003 08:59, guenter geiger wrote:
Hi Larry,
thanks for reporting the problem. Could it be that pd gets stuck because it loads all the libraries ? I have just tested the setup and pd works with the latest CVS version of jack .. I am a bit at a loss at this one :(
Guenter
Music Technology Group - UPF, Barcelona http://www.iua.upf.es/mtg/eng/ AGNULA - A GNU Linux Audio distribution <www.agnula.org> DeMuDi - Debian Multimedia Distribution <www.demudi.org>
On Sat, 15 Mar 2003, Larry Troxler wrote:
Jack 0.51 and PD 0.36, both from planet CCRMA, seem to work fine together.
However, since I was having problems with rosegaren-4 and Jack (frooze at least the UI), I built Jack 0.61.0 from the tarball at the Jack site.
Now rosegarden starts up fine, but PD doesn't.
Following is a shell session showing the problem. I just thought I'd post this in case anyone had any ideas.
Regards
Larry Troxler
~ ~ ~
root@dhcp-25-2 ~]$ jackd -R -d alsa -d hw:0 -r 44100 -p 512 -P & [1] 7518 jackd 0.61.0 Copyright 2001-2002 Paul Davis and others. jackd comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions; see the file COPYING for details
[root@dhcp-25-2 ~]$ loading driver .. driver client name = [alsa_pcm] creating alsa driver ... hw:0|512|2|44100|swmon|rt delivering event 2 to IP client alsa_pcm delivering event 5 to IP client alsa_pcm starting engine
[root@dhcp-25-2 ~]$ pd -r 44100 -jack -outchannels 2 -audiobuf 20 -rt & [2] 7525 reading startup file: /root/.pdrc Testing for Jack cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) cannot read connection request from client cannot complete new client connection process could not attach as client (duplicate client name?) [root@dhcp-25-2 ~]$ using OSS audio interface priority 96 scheduling enabled. memory locking enabled. priority 98 scheduling enabled. memory locking enabled.
((((((((((((((())))))))))))))) ( scheme is here ) ((((((((((((((()))))))))))))))loading /root/pd/pdinit.scm loading /root/dl/siod/siod.scm ;; Optional Runtime Library for Release 3.0 done. loading /root/pd/pdsite.scm loading /root/pd/pdexampl.scm done. done. done.
@@@@@@@@@@@@@@@@@@@@@@@@@@@@ @...the zexy external1.1...@ @ forum::f?r::uml?ute 2001 @ @ send me a 'help' message @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@**** alsa_pcm: xrun of at least 31.762 msecs
audio I/O stuck... closing audio
~ ~ ~
PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list
-
Adam Lindsay -
Antonis Galanopoulos -
Fernando Pablo Lopez-Lezcano -
Frank Barknecht -
guenter geiger -
Hans-Christoph Steiner -
jfm3 -
Larry Troxler -
Thomas Grill