About threading with Csound and Python, you shouldn't worry about the Python 
global interpreter lock, because the Csound performance thread is spawned 
inside Csound's C code by Csound, not by Python.

It makes no difference about PortAudio, you are correct, the only fix is to 
fix the PortAudio configuration.

Regards,
Mike

----- Original Message ----- 
From: "Chuckk Hubbard" 
To: 
Sent: Saturday, May 30, 2009 5:56 AM
Subject: [Csnd] Re: Re: Pared down example of Python-Csound Windows problem


> Hi Mike.
> Thanks for the info. The reason for running Csound in a separate
> process is that, in my program, I want the option of running the audio
> engine (Csound) at a higher priority than the GUI. The example doesn't
> show this since the GUI doesn't do much, but in my sequencer there's a
> line that moves across the score showing the current location, and I
> want to avoid potential dropouts from that.
>
> I've heard of the possibility, with other languages, of having threads
> with assigned priorities, but as I understand it Python's Global
> Interpreter Lock or some such feature prevents that.
>
> If it's from the portaudio dll, I'm guessing running Csound inside the
> same process wouldn't make a difference, though, no? It still has to
> end some time.
>
> Actually one option I'm considering is to have Csound running
> constantly, and just update instruments and options on the fly; that
> would allow the user to hear notes as they enter them, not just when
> they hit play...
>
> -Chuckk
>
> On Fri, May 29, 2009 at 6:28 PM,   wrote:
>> I tried your example with my current build of Csound on Windows XP, and I
>> get the same behavior you report.
>>
>> The problem looks to me like it is caused in the PortAudio DLL, which 
>> means
>> that the access violation cannot be handled by the program that loaded 
>> the
>> DLL.
>>
>> In all likelihood, this is because my build of PortAudio is not correct. 
>> I
>> do not recall this problem occurring in the previous version of Csound,
>> which had a different version of my build of PortAudio.
>>
>> I tried using perf.InputMessage('e 1') to stop Csound, and that didn't 
>> work,
>> the result was just the same message box.
>>
>> Then I tried installing a signal handler to handle the crash instead of
>> getting the message box, and that also didn't work, the result was just 
>> the
>> same message box.
>>
>> I still don't get why you have Csound running in a separate process -- is 
>> it
>> just because of this crash?
>>
>> Regards,
>> Mike
>> ----- Original Message ----- From: "Chuckk Hubbard"
>> 
>> To: "Csound List" 
>> Sent: Friday, May 29, 2009 5:30 PM
>> Subject: [Csnd] Pared down example of Python-Csound Windows problem
>>
>>
>>> I posted a different version of this before and it seemed there was no
>>> problem on Vista; I since found that someone else who was trying to
>>> run my program on XP (which is what I have) was having the same error
>>> message as me.
>>> It doesn't interfere with the operation of anything, and in fact only
>>> happens right when the child process is closing; but one must click on
>>> OK on the error message in order to proceed, which gets to be like
>>> listening to the faucet drip all night.
>>>
>>> For this example, check the audio options just to be sure, and then
>>> hit PLAY followed by STOP after a second or two. That's when I get the
>>> error.
>>>
>>> I've commented pretty thoroughly, and I'm pretty sure the problem
>>> happens in either the play() or stop() function of csdproblem.py, or
>>> in the marked if clause in csdproblemaudio.py.
>>>
>>> The error reads:
>>> """The instruction at "0x7c9104fa" referenced memory at "0xfffffffd".
>>> The memory could not be "read".
>>> Click on OK to terminate the program"""
>>> The window says python.exe Application error, but I believe it would
>>> read this anyway, as no Csound process actually runs.
>>>
>>> I and the other fellow are both using Csound Windows binaries for 5.10
>>> Can anyone else running XP confirm this?
>>>
>>> Thanks a lot.
>>> -Chuckk
>>>
>>> --
>>> http://www.badmuthahubbard.com
>>>
>>>
>>> Send bugs reports to this list.
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
>
> -- 
> http://www.badmuthahubbard.com
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe 
> csound"
> 

Okay.
But at any rate, if the PortAudio DLL is corrected, there's no need to
wait for another Csound build, it's dynamically linked, no?

-Chuckk

On Sat, May 30, 2009 at 2:07 PM,   wrote:
> About threading with Csound and Python, you shouldn't worry about the Python
> global interpreter lock, because the Csound performance thread is spawned
> inside Csound's C code by Csound, not by Python.
>
> It makes no difference about PortAudio, you are correct, the only fix is to
> fix the PortAudio configuration.
>
> Regards,
> Mike
>
> ----- Original Message ----- From: "Chuckk Hubbard"
> 
> To: 
> Sent: Saturday, May 30, 2009 5:56 AM
> Subject: [Csnd] Re: Re: Pared down example of Python-Csound Windows problem
>
>
>> Hi Mike.
>> Thanks for the info. The reason for running Csound in a separate
>> process is that, in my program, I want the option of running the audio
>> engine (Csound) at a higher priority than the GUI. The example doesn't
>> show this since the GUI doesn't do much, but in my sequencer there's a
>> line that moves across the score showing the current location, and I
>> want to avoid potential dropouts from that.
>>
>> I've heard of the possibility, with other languages, of having threads
>> with assigned priorities, but as I understand it Python's Global
>> Interpreter Lock or some such feature prevents that.
>>
>> If it's from the portaudio dll, I'm guessing running Csound inside the
>> same process wouldn't make a difference, though, no? It still has to
>> end some time.
>>
>> Actually one option I'm considering is to have Csound running
>> constantly, and just update instruments and options on the fly; that
>> would allow the user to hear notes as they enter them, not just when
>> they hit play...
>>
>> -Chuckk
>>
>> On Fri, May 29, 2009 at 6:28 PM,   wrote:
>>>
>>> I tried your example with my current build of Csound on Windows XP, and I
>>> get the same behavior you report.
>>>
>>> The problem looks to me like it is caused in the PortAudio DLL, which
>>> means
>>> that the access violation cannot be handled by the program that loaded
>>> the
>>> DLL.
>>>
>>> In all likelihood, this is because my build of PortAudio is not correct.
>>> I
>>> do not recall this problem occurring in the previous version of Csound,
>>> which had a different version of my build of PortAudio.
>>>
>>> I tried using perf.InputMessage('e 1') to stop Csound, and that didn't
>>> work,
>>> the result was just the same message box.
>>>
>>> Then I tried installing a signal handler to handle the crash instead of
>>> getting the message box, and that also didn't work, the result was just
>>> the
>>> same message box.
>>>
>>> I still don't get why you have Csound running in a separate process -- is
>>> it
>>> just because of this crash?
>>>
>>> Regards,
>>> Mike
>>> ----- Original Message ----- From: "Chuckk Hubbard"
>>> 
>>> To: "Csound List" 
>>> Sent: Friday, May 29, 2009 5:30 PM
>>> Subject: [Csnd] Pared down example of Python-Csound Windows problem
>>>
>>>
>>>> I posted a different version of this before and it seemed there was no
>>>> problem on Vista; I since found that someone else who was trying to
>>>> run my program on XP (which is what I have) was having the same error
>>>> message as me.
>>>> It doesn't interfere with the operation of anything, and in fact only
>>>> happens right when the child process is closing; but one must click on
>>>> OK on the error message in order to proceed, which gets to be like
>>>> listening to the faucet drip all night.
>>>>
>>>> For this example, check the audio options just to be sure, and then
>>>> hit PLAY followed by STOP after a second or two. That's when I get the
>>>> error.
>>>>
>>>> I've commented pretty thoroughly, and I'm pretty sure the problem
>>>> happens in either the play() or stop() function of csdproblem.py, or
>>>> in the marked if clause in csdproblemaudio.py.
>>>>
>>>> The error reads:
>>>> """The instruction at "0x7c9104fa" referenced memory at "0xfffffffd".
>>>> The memory could not be "read".
>>>> Click on OK to terminate the program"""
>>>> The window says python.exe Application error, but I believe it would
>>>> read this anyway, as no Csound process actually runs.
>>>>
>>>> I and the other fellow are both using Csound Windows binaries for 5.10
>>>> Can anyone else running XP confirm this?
>>>>
>>>> Thanks a lot.
>>>> -Chuckk
>>>>
>>>> --
>>>> http://www.badmuthahubbard.com
>>>>
>>>>
>>>> Send bugs reports to this list.
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>>> csound"
>>>>
>>>
>>>
>>>
>>> Send bugs reports to this list.
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> --
>> http://www.badmuthahubbard.com
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>



-- 
http://www.badmuthahubbard.com

It is dynamically linked, but it is possible the binary interface has changed.

You can always try an earlier Portaudio DLL, or one from another
project. Just make the filename and location of the new file the same
as the filename and location of the old one.

Hope this helps,
Mike

On 5/31/09, Chuckk Hubbard  wrote:
> Okay.
> But at any rate, if the PortAudio DLL is corrected, there's no need to
> wait for another Csound build, it's dynamically linked, no?
>
> -Chuckk
>
> On Sat, May 30, 2009 at 2:07 PM,   wrote:
>> About threading with Csound and Python, you shouldn't worry about the
>> Python
>> global interpreter lock, because the Csound performance thread is spawned
>> inside Csound's C code by Csound, not by Python.
>>
>> It makes no difference about PortAudio, you are correct, the only fix is
>> to
>> fix the PortAudio configuration.
>>
>> Regards,
>> Mike
>>
>> ----- Original Message ----- From: "Chuckk Hubbard"
>> 
>> To: 
>> Sent: Saturday, May 30, 2009 5:56 AM
>> Subject: [Csnd] Re: Re: Pared down example of Python-Csound Windows
>> problem
>>
>>
>>> Hi Mike.
>>> Thanks for the info. The reason for running Csound in a separate
>>> process is that, in my program, I want the option of running the audio
>>> engine (Csound) at a higher priority than the GUI. The example doesn't
>>> show this since the GUI doesn't do much, but in my sequencer there's a
>>> line that moves across the score showing the current location, and I
>>> want to avoid potential dropouts from that.
>>>
>>> I've heard of the possibility, with other languages, of having threads
>>> with assigned priorities, but as I understand it Python's Global
>>> Interpreter Lock or some such feature prevents that.
>>>
>>> If it's from the portaudio dll, I'm guessing running Csound inside the
>>> same process wouldn't make a difference, though, no? It still has to
>>> end some time.
>>>
>>> Actually one option I'm considering is to have Csound running
>>> constantly, and just update instruments and options on the fly; that
>>> would allow the user to hear notes as they enter them, not just when
>>> they hit play...
>>>
>>> -Chuckk
>>>
>>> On Fri, May 29, 2009 at 6:28 PM,   wrote:
>>>>
>>>> I tried your example with my current build of Csound on Windows XP, and
>>>> I
>>>> get the same behavior you report.
>>>>
>>>> The problem looks to me like it is caused in the PortAudio DLL, which
>>>> means
>>>> that the access violation cannot be handled by the program that loaded
>>>> the
>>>> DLL.
>>>>
>>>> In all likelihood, this is because my build of PortAudio is not correct.
>>>> I
>>>> do not recall this problem occurring in the previous version of Csound,
>>>> which had a different version of my build of PortAudio.
>>>>
>>>> I tried using perf.InputMessage('e 1') to stop Csound, and that didn't
>>>> work,
>>>> the result was just the same message box.
>>>>
>>>> Then I tried installing a signal handler to handle the crash instead of
>>>> getting the message box, and that also didn't work, the result was just
>>>> the
>>>> same message box.
>>>>
>>>> I still don't get why you have Csound running in a separate process --
>>>> is
>>>> it
>>>> just because of this crash?
>>>>
>>>> Regards,
>>>> Mike
>>>> ----- Original Message ----- From: "Chuckk Hubbard"
>>>> 
>>>> To: "Csound List" 
>>>> Sent: Friday, May 29, 2009 5:30 PM
>>>> Subject: [Csnd] Pared down example of Python-Csound Windows problem
>>>>
>>>>
>>>>> I posted a different version of this before and it seemed there was no
>>>>> problem on Vista; I since found that someone else who was trying to
>>>>> run my program on XP (which is what I have) was having the same error
>>>>> message as me.
>>>>> It doesn't interfere with the operation of anything, and in fact only
>>>>> happens right when the child process is closing; but one must click on
>>>>> OK on the error message in order to proceed, which gets to be like
>>>>> listening to the faucet drip all night.
>>>>>
>>>>> For this example, check the audio options just to be sure, and then
>>>>> hit PLAY followed by STOP after a second or two. That's when I get the
>>>>> error.
>>>>>
>>>>> I've commented pretty thoroughly, and I'm pretty sure the problem
>>>>> happens in either the play() or stop() function of csdproblem.py, or
>>>>> in the marked if clause in csdproblemaudio.py.
>>>>>
>>>>> The error reads:
>>>>> """The instruction at "0x7c9104fa" referenced memory at "0xfffffffd".
>>>>> The memory could not be "read".
>>>>> Click on OK to terminate the program"""
>>>>> The window says python.exe Application error, but I believe it would
>>>>> read this anyway, as no Csound process actually runs.
>>>>>
>>>>> I and the other fellow are both using Csound Windows binaries for 5.10
>>>>> Can anyone else running XP confirm this?
>>>>>
>>>>> Thanks a lot.
>>>>> -Chuckk
>>>>>
>>>>> --
>>>>> http://www.badmuthahubbard.com
>>>>>
>>>>>
>>>>> Send bugs reports to this list.
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>> "unsubscribe
>>>>> csound"
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to this list.
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>>> csound"
>>>>
>>>
>>>
>>>
>>> --
>>> http://www.badmuthahubbard.com
>>>
>>>
>>> Send bugs reports to this list.
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
>
> --
> http://www.badmuthahubbard.com
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"


-- 
Michael Gogins
Irreducible Productions
http://www.michael-gogins.com
Michael dot Gogins at gmail dot com