I guess it's the one in framework, I think that's the one blue tries  
to access?

Thanks

P

On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:

> yes, there might be a link problem there. Did you try the  
> lib_csnd.dylib inside the framework or the one in /usr/local/lib?
>
> ----- Original Message -----
> From: Peiman Khosravi 
> Date: Thursday, May 20, 2010 8:23 pm
> Subject: [Csnd] lib_csnd.dylib doubles problem
> To: csound@lists.bath.ac.uk
>
>
> -----------------------------------------------------------
> |
>
>> Hello,
>
>
>> I think there is still something wrong with lib_csnd.dylib doubles  
>> on OS X as blue crashes when run in api mode. >
>> Thanks>
>> Peiman |
> -----------------------------------------------------------
>
>
>
> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
> National University of Ireland, Maynooth
>
>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

I've had problems getting the API working on OSX myself. It could be
that the lib_csnd.dylib in /usr/local/lib is the issue, but it might
also be the symlink in the java lib dir. I don't remember exactly
where it is, but I know the installer was symlinking the 32 bit
version instead of the doubles version (don't remember if that
changed). I've had a note to look further into this but haven't found
the time. If you figure it out, would love to know!

On 5/20/10, Peiman Khosravi  wrote:
> I don't know if it's normal but I have these in /usr/local/lib
>
> lib_csnd.dylib			
> lib_csnd64.dylib		
> libcsnd.dylib			
> libcsnd64.dylib	
>
> And only lib_csnd.dylib in CsoundLib64.framework
>
> And I get this line in blue's crash report:
>
> Thread 23 Crashed:
> 0   lib_csnd.dylib                	0x51266738
> CsoundCallbackWrapper::SetMessageCallback() + 24
>
>
> On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:
>
>> yes, there might be a link problem there. Did you try the
>> lib_csnd.dylib inside the framework or the one in /usr/local/lib?
>>
>> ----- Original Message -----
>> From: Peiman Khosravi 
>> Date: Thursday, May 20, 2010 8:23 pm
>> Subject: [Csnd] lib_csnd.dylib doubles problem
>> To: csound@lists.bath.ac.uk
>>
>>
>> -----------------------------------------------------------
>> |
>>
>>> Hello,
>>
>>
>>> I think there is still something wrong with lib_csnd.dylib doubles
>>> on OS X as blue crashes when run in api mode. >
>>> Thanks>
>>> Peiman |
>> -----------------------------------------------------------
>>
>>
>>
>> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
>> National University of Ireland, Maynooth
>>
>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>> "unsubscribe csound"
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>             https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>
>

Hi Victor,

Thanks very much for your time in looking into this!

I'll have to look further into what is going on, as blue uses Csound
and not CppSound, and it also uses Java threads and does not use
CsoundPerformanceThread.  I think I'll finally have some time to look
into this today and will try to see if I can get anywhere.

Thanks!
steven


On Fri, May 21, 2010 at 7:55 AM, Victor Lazzarini
 wrote:
> And I bet this is because CppSound is derived from two base classes and I
> don't think Java likes that.
> Victor
> On 21 May 2010, at 12:39, Victor Lazzarini wrote:
>
> Further findings:
> The problem I identified was in CsoundEditor.java: it seems that a
> combination of the CppSound class and CsoundPerformanceThread is causing the
> problem. However, using the Csound class and CsoundPerformanceThread is
> perfectly fine. I have adapted CSDPlayer to use it and it's OK.
> So if blue is using CppSound, then the problem might be there.
> Regards
> Victor
>
> On 21 May 2010, at 12:16, Victor Lazzarini wrote:
>
> More diagnostics:
> it seems that the CsoundPerformanceThread class is causing crashes on Java.
> CSDPlayer uses its own java threading and does not rely on that class, that
> is one of the reasons it works.
> Victor
> On 21 May 2010, at 11:56, Victor Lazzarini wrote:
>
> Ok, just to complement this discussion. I built the CSDPlayer.java example
> from examples/java  and tested it (the same program, no re-build was needed)
> with both floats and doubles (by changing the symlink in
> JavaVM.framework/libraries). Both versions ran without any problems, as I
> would expect.
> I think it should not matter whether the link is to floats or doubles,
> unless you are assuming specific MYFLT sizes in the java code, wrapping
> pointers or something like that. The same java compiled code should work
> with both. So I would expect blue to work with the installation as it is (I
> mean make a sound and not crash). I can change the installation to doubles
> (as I will probably do for python as well), but that should not matter as
> far as running the program is concerned.
> So I guess there might be a problem elsewhere. The Java module seems to be
> OK.
> Victor
>
>
> On 21 May 2010, at 11:27, Peiman Khosravi wrote:
>
> Just to confirm, I tried changing it too and blue still crashes.
> P
> On 21 May 2010, at 11:20, Victor Lazzarini wrote:
>
> I had a quick look this morning and what I found was: the csound module
> installed in JavaVM.framework/Resources/Libraries is a symlink to the 32-bit
> framework. I have changed this to point to the 64-bit framework, but blue is
> still crashing
> 21/05/2010 11:08:40 com.apple.launchd[119] ([0x0-0x42a42a].blue[39891])
> Exited with exit code: 138
> 21/05/2010 11:09:30 [0x0-0x42f42f].blue[39994] Invalid memory access of
> location 0x4 eip=0x5170b708
> I checked the lib_csnd.jnilib  and they are correct (both lib_csnd and
> CsoundLib are 64bit).
> Perhaps Steven might be able to cast some light onto this.
> Victor
>
> On 21 May 2010, at 10:58, Peiman Khosravi wrote:
>
> Thanks very much.
> Peiman
> On 21 May 2010, at 10:27, Victor Lazzarini wrote:
>
> yes, the link should be to lib_csnd.dylib inside CsoundLIb64 framework. But
> Steven is right to say
> that the Java lib that is installed is linked to the 32 bit framework. I'll
> see if I can sort all of this out for
> the next release.
> Victor
>
> On 21 May 2010, at 00:47, Peiman Khosravi wrote:
>
> I don't know if it's normal but I have these in /usr/local/lib
> lib_csnd.dylib
> lib_csnd64.dylib
> libcsnd.dylib
> libcsnd64.dylib
> And only lib_csnd.dylib in CsoundLib64.framework
> And I get this line in blue's crash report:
> Thread 23 Crashed:
> 0   lib_csnd.dylib                 0x51266738
> CsoundCallbackWrapper::SetMessageCallback() + 24
>
> On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:
>
> yes, there might be a link problem there. Did you try the lib_csnd.dylib
> inside the framework or the one in /usr/local/lib?
>
> ----- Original Message -----
> From: Peiman Khosravi 
> Date: Thursday, May 20, 2010 8:23 pm
> Subject: [Csnd] lib_csnd.dylib doubles problem
> To: csound@lists.bath.ac.uk
>
>
> -----------------------------------------------------------
> |
>
> Hello,
>
>
> I think there is still something wrong with lib_csnd.dylib doubles on OS X
> as blue crashes when run in api mode. >
>
> Thanks>
>
> Peiman |
>
> -----------------------------------------------------------
>
>
>
> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
> National University of Ireland, Maynooth
>
>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>
>
>
>
>
>
>
>
>
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

OK, if that is the case, then take a look at the CSDPlayer example in  
the sources. That works fine, so it should help you in your diagnosis.
On 21 May 2010, at 15:09, Steven Yi wrote:

> Hi Victor,
>
> Thanks very much for your time in looking into this!
>
> I'll have to look further into what is going on, as blue uses Csound
> and not CppSound, and it also uses Java threads and does not use
> CsoundPerformanceThread.  I think I'll finally have some time to look
> into this today and will try to see if I can get anywhere.
>
> Thanks!
> steven
>
>
> On Fri, May 21, 2010 at 7:55 AM, Victor Lazzarini
>  wrote:
>> And I bet this is because CppSound is derived from two base classes  
>> and I
>> don't think Java likes that.
>> Victor
>> On 21 May 2010, at 12:39, Victor Lazzarini wrote:
>>
>> Further findings:
>> The problem I identified was in CsoundEditor.java: it seems that a
>> combination of the CppSound class and CsoundPerformanceThread is  
>> causing the
>> problem. However, using the Csound class and  
>> CsoundPerformanceThread is
>> perfectly fine. I have adapted CSDPlayer to use it and it's OK.
>> So if blue is using CppSound, then the problem might be there.
>> Regards
>> Victor
>>
>> On 21 May 2010, at 12:16, Victor Lazzarini wrote:
>>
>> More diagnostics:
>> it seems that the CsoundPerformanceThread class is causing crashes  
>> on Java.
>> CSDPlayer uses its own java threading and does not rely on that  
>> class, that
>> is one of the reasons it works.
>> Victor
>> On 21 May 2010, at 11:56, Victor Lazzarini wrote:
>>
>> Ok, just to complement this discussion. I built the CSDPlayer.java  
>> example
>> from examples/java  and tested it (the same program, no re-build  
>> was needed)
>> with both floats and doubles (by changing the symlink in
>> JavaVM.framework/libraries). Both versions ran without any  
>> problems, as I
>> would expect.
>> I think it should not matter whether the link is to floats or  
>> doubles,
>> unless you are assuming specific MYFLT sizes in the java code,  
>> wrapping
>> pointers or something like that. The same java compiled code should  
>> work
>> with both. So I would expect blue to work with the installation as  
>> it is (I
>> mean make a sound and not crash). I can change the installation to  
>> doubles
>> (as I will probably do for python as well), but that should not  
>> matter as
>> far as running the program is concerned.
>> So I guess there might be a problem elsewhere. The Java module  
>> seems to be
>> OK.
>> Victor
>>
>>
>> On 21 May 2010, at 11:27, Peiman Khosravi wrote:
>>
>> Just to confirm, I tried changing it too and blue still crashes.
>> P
>> On 21 May 2010, at 11:20, Victor Lazzarini wrote:
>>
>> I had a quick look this morning and what I found was: the csound  
>> module
>> installed in JavaVM.framework/Resources/Libraries is a symlink to  
>> the 32-bit
>> framework. I have changed this to point to the 64-bit framework,  
>> but blue is
>> still crashing
>> 21/05/2010 11:08:40 com.apple.launchd[119]  
>> ([0x0-0x42a42a].blue[39891])
>> Exited with exit code: 138
>> 21/05/2010 11:09:30 [0x0-0x42f42f].blue[39994] Invalid memory  
>> access of
>> location 0x4 eip=0x5170b708
>> I checked the lib_csnd.jnilib  and they are correct (both lib_csnd  
>> and
>> CsoundLib are 64bit).
>> Perhaps Steven might be able to cast some light onto this.
>> Victor
>>
>> On 21 May 2010, at 10:58, Peiman Khosravi wrote:
>>
>> Thanks very much.
>> Peiman
>> On 21 May 2010, at 10:27, Victor Lazzarini wrote:
>>
>> yes, the link should be to lib_csnd.dylib inside CsoundLIb64  
>> framework. But
>> Steven is right to say
>> that the Java lib that is installed is linked to the 32 bit  
>> framework. I'll
>> see if I can sort all of this out for
>> the next release.
>> Victor
>>
>> On 21 May 2010, at 00:47, Peiman Khosravi wrote:
>>
>> I don't know if it's normal but I have these in /usr/local/lib
>> lib_csnd.dylib
>> lib_csnd64.dylib
>> libcsnd.dylib
>> libcsnd64.dylib
>> And only lib_csnd.dylib in CsoundLib64.framework
>> And I get this line in blue's crash report:
>> Thread 23 Crashed:
>> 0   lib_csnd.dylib                 0x51266738
>> CsoundCallbackWrapper::SetMessageCallback() + 24
>>
>> On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:
>>
>> yes, there might be a link problem there. Did you try the  
>> lib_csnd.dylib
>> inside the framework or the one in /usr/local/lib?
>>
>> ----- Original Message -----
>> From: Peiman Khosravi 
>> Date: Thursday, May 20, 2010 8:23 pm
>> Subject: [Csnd] lib_csnd.dylib doubles problem
>> To: csound@lists.bath.ac.uk
>>
>>
>> -----------------------------------------------------------
>> |
>>
>> Hello,
>>
>>
>> I think there is still something wrong with lib_csnd.dylib doubles  
>> on OS X
>> as blue crashes when run in api mode. >
>>
>> Thanks>
>>
>> Peiman |
>>
>> -----------------------------------------------------------
>>
>>
>>
>> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
>> National University of Ireland, Maynooth
>>
>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>            https://sourceforge.net/tracker/? 
>> group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe
>> csound"
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

I haven't had a chance to look at the CSDPlayer example, as I'm still
having issues with getting the API to load at all.  I think there's an
issue having to do with 32bit/64bit versions of Java and trying to
load the jnilib (jnilib is i386, but for some reason running blue from
the app launcher keeps using the x86_64 version of Java).  I think I
need to figure out this issue first, then can work on the API issue.

On Fri, May 21, 2010 at 10:14 AM, Victor Lazzarini
 wrote:
> OK, if that is the case, then take a look at the CSDPlayer example in the
> sources. That works fine, so it should help you in your diagnosis.
> On 21 May 2010, at 15:09, Steven Yi wrote:
>
>> Hi Victor,
>>
>> Thanks very much for your time in looking into this!
>>
>> I'll have to look further into what is going on, as blue uses Csound
>> and not CppSound, and it also uses Java threads and does not use
>> CsoundPerformanceThread.  I think I'll finally have some time to look
>> into this today and will try to see if I can get anywhere.
>>
>> Thanks!
>> steven
>>
>>
>> On Fri, May 21, 2010 at 7:55 AM, Victor Lazzarini
>>  wrote:
>>>
>>> And I bet this is because CppSound is derived from two base classes and I
>>> don't think Java likes that.
>>> Victor
>>> On 21 May 2010, at 12:39, Victor Lazzarini wrote:
>>>
>>> Further findings:
>>> The problem I identified was in CsoundEditor.java: it seems that a
>>> combination of the CppSound class and CsoundPerformanceThread is causing
>>> the
>>> problem. However, using the Csound class and CsoundPerformanceThread is
>>> perfectly fine. I have adapted CSDPlayer to use it and it's OK.
>>> So if blue is using CppSound, then the problem might be there.
>>> Regards
>>> Victor
>>>
>>> On 21 May 2010, at 12:16, Victor Lazzarini wrote:
>>>
>>> More diagnostics:
>>> it seems that the CsoundPerformanceThread class is causing crashes on
>>> Java.
>>> CSDPlayer uses its own java threading and does not rely on that class,
>>> that
>>> is one of the reasons it works.
>>> Victor
>>> On 21 May 2010, at 11:56, Victor Lazzarini wrote:
>>>
>>> Ok, just to complement this discussion. I built the CSDPlayer.java
>>> example
>>> from examples/java  and tested it (the same program, no re-build was
>>> needed)
>>> with both floats and doubles (by changing the symlink in
>>> JavaVM.framework/libraries). Both versions ran without any problems, as I
>>> would expect.
>>> I think it should not matter whether the link is to floats or doubles,
>>> unless you are assuming specific MYFLT sizes in the java code, wrapping
>>> pointers or something like that. The same java compiled code should work
>>> with both. So I would expect blue to work with the installation as it is
>>> (I
>>> mean make a sound and not crash). I can change the installation to
>>> doubles
>>> (as I will probably do for python as well), but that should not matter as
>>> far as running the program is concerned.
>>> So I guess there might be a problem elsewhere. The Java module seems to
>>> be
>>> OK.
>>> Victor
>>>
>>>
>>> On 21 May 2010, at 11:27, Peiman Khosravi wrote:
>>>
>>> Just to confirm, I tried changing it too and blue still crashes.
>>> P
>>> On 21 May 2010, at 11:20, Victor Lazzarini wrote:
>>>
>>> I had a quick look this morning and what I found was: the csound module
>>> installed in JavaVM.framework/Resources/Libraries is a symlink to the
>>> 32-bit
>>> framework. I have changed this to point to the 64-bit framework, but blue
>>> is
>>> still crashing
>>> 21/05/2010 11:08:40 com.apple.launchd[119] ([0x0-0x42a42a].blue[39891])
>>> Exited with exit code: 138
>>> 21/05/2010 11:09:30 [0x0-0x42f42f].blue[39994] Invalid memory access of
>>> location 0x4 eip=0x5170b708
>>> I checked the lib_csnd.jnilib  and they are correct (both lib_csnd and
>>> CsoundLib are 64bit).
>>> Perhaps Steven might be able to cast some light onto this.
>>> Victor
>>>
>>> On 21 May 2010, at 10:58, Peiman Khosravi wrote:
>>>
>>> Thanks very much.
>>> Peiman
>>> On 21 May 2010, at 10:27, Victor Lazzarini wrote:
>>>
>>> yes, the link should be to lib_csnd.dylib inside CsoundLIb64 framework.
>>> But
>>> Steven is right to say
>>> that the Java lib that is installed is linked to the 32 bit framework.
>>> I'll
>>> see if I can sort all of this out for
>>> the next release.
>>> Victor
>>>
>>> On 21 May 2010, at 00:47, Peiman Khosravi wrote:
>>>
>>> I don't know if it's normal but I have these in /usr/local/lib
>>> lib_csnd.dylib
>>> lib_csnd64.dylib
>>> libcsnd.dylib
>>> libcsnd64.dylib
>>> And only lib_csnd.dylib in CsoundLib64.framework
>>> And I get this line in blue's crash report:
>>> Thread 23 Crashed:
>>> 0   lib_csnd.dylib                 0x51266738
>>> CsoundCallbackWrapper::SetMessageCallback() + 24
>>>
>>> On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:
>>>
>>> yes, there might be a link problem there. Did you try the lib_csnd.dylib
>>> inside the framework or the one in /usr/local/lib?
>>>
>>> ----- Original Message -----
>>> From: Peiman Khosravi 
>>> Date: Thursday, May 20, 2010 8:23 pm
>>> Subject: [Csnd] lib_csnd.dylib doubles problem
>>> To: csound@lists.bath.ac.uk
>>>
>>>
>>> -----------------------------------------------------------
>>> |
>>>
>>> Hello,
>>>
>>>
>>> I think there is still something wrong with lib_csnd.dylib doubles on OS
>>> X
>>> as blue crashes when run in api mode. >
>>>
>>> Thanks>
>>>
>>> Peiman |
>>>
>>> -----------------------------------------------------------
>>>
>>>
>>>
>>> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
>>> National University of Ireland, Maynooth
>>>
>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Okay, I've isolated the problem.  First problem is the 32bit/64bit.  I
know of a way to force 32bit Java to run on OS 10.6, but I don't think
it works on non-10.6 OSX, so am still searching for the solution for
that. (Emailed another mailing list for advice).  After running with
32bit Java, I was able to reproduce the problem.

Now, the error Victor posted earlier is the source of the problem:

0   lib_csnd.dylib                	0x51266738
CsoundCallbackWrapper::SetMessageCallback() + 24

Setting the message callback causes the crash. If I comment that out,
blue will run with the API without problems.  I don't know of a
solution for this yet, though this code works fine on other platforms,
so am thinking it may be tricky to figure out the answer for this one.

On Fri, May 21, 2010 at 2:49 PM, Steven Yi  wrote:
> I haven't had a chance to look at the CSDPlayer example, as I'm still
> having issues with getting the API to load at all.  I think there's an
> issue having to do with 32bit/64bit versions of Java and trying to
> load the jnilib (jnilib is i386, but for some reason running blue from
> the app launcher keeps using the x86_64 version of Java).  I think I
> need to figure out this issue first, then can work on the API issue.
>
> On Fri, May 21, 2010 at 10:14 AM, Victor Lazzarini
>  wrote:
>> OK, if that is the case, then take a look at the CSDPlayer example in the
>> sources. That works fine, so it should help you in your diagnosis.
>> On 21 May 2010, at 15:09, Steven Yi wrote:
>>
>>> Hi Victor,
>>>
>>> Thanks very much for your time in looking into this!
>>>
>>> I'll have to look further into what is going on, as blue uses Csound
>>> and not CppSound, and it also uses Java threads and does not use
>>> CsoundPerformanceThread.  I think I'll finally have some time to look
>>> into this today and will try to see if I can get anywhere.
>>>
>>> Thanks!
>>> steven
>>>
>>>
>>> On Fri, May 21, 2010 at 7:55 AM, Victor Lazzarini
>>>  wrote:
>>>>
>>>> And I bet this is because CppSound is derived from two base classes and I
>>>> don't think Java likes that.
>>>> Victor
>>>> On 21 May 2010, at 12:39, Victor Lazzarini wrote:
>>>>
>>>> Further findings:
>>>> The problem I identified was in CsoundEditor.java: it seems that a
>>>> combination of the CppSound class and CsoundPerformanceThread is causing
>>>> the
>>>> problem. However, using the Csound class and CsoundPerformanceThread is
>>>> perfectly fine. I have adapted CSDPlayer to use it and it's OK.
>>>> So if blue is using CppSound, then the problem might be there.
>>>> Regards
>>>> Victor
>>>>
>>>> On 21 May 2010, at 12:16, Victor Lazzarini wrote:
>>>>
>>>> More diagnostics:
>>>> it seems that the CsoundPerformanceThread class is causing crashes on
>>>> Java.
>>>> CSDPlayer uses its own java threading and does not rely on that class,
>>>> that
>>>> is one of the reasons it works.
>>>> Victor
>>>> On 21 May 2010, at 11:56, Victor Lazzarini wrote:
>>>>
>>>> Ok, just to complement this discussion. I built the CSDPlayer.java
>>>> example
>>>> from examples/java  and tested it (the same program, no re-build was
>>>> needed)
>>>> with both floats and doubles (by changing the symlink in
>>>> JavaVM.framework/libraries). Both versions ran without any problems, as I
>>>> would expect.
>>>> I think it should not matter whether the link is to floats or doubles,
>>>> unless you are assuming specific MYFLT sizes in the java code, wrapping
>>>> pointers or something like that. The same java compiled code should work
>>>> with both. So I would expect blue to work with the installation as it is
>>>> (I
>>>> mean make a sound and not crash). I can change the installation to
>>>> doubles
>>>> (as I will probably do for python as well), but that should not matter as
>>>> far as running the program is concerned.
>>>> So I guess there might be a problem elsewhere. The Java module seems to
>>>> be
>>>> OK.
>>>> Victor
>>>>
>>>>
>>>> On 21 May 2010, at 11:27, Peiman Khosravi wrote:
>>>>
>>>> Just to confirm, I tried changing it too and blue still crashes.
>>>> P
>>>> On 21 May 2010, at 11:20, Victor Lazzarini wrote:
>>>>
>>>> I had a quick look this morning and what I found was: the csound module
>>>> installed in JavaVM.framework/Resources/Libraries is a symlink to the
>>>> 32-bit
>>>> framework. I have changed this to point to the 64-bit framework, but blue
>>>> is
>>>> still crashing
>>>> 21/05/2010 11:08:40 com.apple.launchd[119] ([0x0-0x42a42a].blue[39891])
>>>> Exited with exit code: 138
>>>> 21/05/2010 11:09:30 [0x0-0x42f42f].blue[39994] Invalid memory access of
>>>> location 0x4 eip=0x5170b708
>>>> I checked the lib_csnd.jnilib  and they are correct (both lib_csnd and
>>>> CsoundLib are 64bit).
>>>> Perhaps Steven might be able to cast some light onto this.
>>>> Victor
>>>>
>>>> On 21 May 2010, at 10:58, Peiman Khosravi wrote:
>>>>
>>>> Thanks very much.
>>>> Peiman
>>>> On 21 May 2010, at 10:27, Victor Lazzarini wrote:
>>>>
>>>> yes, the link should be to lib_csnd.dylib inside CsoundLIb64 framework.
>>>> But
>>>> Steven is right to say
>>>> that the Java lib that is installed is linked to the 32 bit framework.
>>>> I'll
>>>> see if I can sort all of this out for
>>>> the next release.
>>>> Victor
>>>>
>>>> On 21 May 2010, at 00:47, Peiman Khosravi wrote:
>>>>
>>>> I don't know if it's normal but I have these in /usr/local/lib
>>>> lib_csnd.dylib
>>>> lib_csnd64.dylib
>>>> libcsnd.dylib
>>>> libcsnd64.dylib
>>>> And only lib_csnd.dylib in CsoundLib64.framework
>>>> And I get this line in blue's crash report:
>>>> Thread 23 Crashed:
>>>> 0   lib_csnd.dylib                 0x51266738
>>>> CsoundCallbackWrapper::SetMessageCallback() + 24
>>>>
>>>> On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:
>>>>
>>>> yes, there might be a link problem there. Did you try the lib_csnd.dylib
>>>> inside the framework or the one in /usr/local/lib?
>>>>
>>>> ----- Original Message -----
>>>> From: Peiman Khosravi 
>>>> Date: Thursday, May 20, 2010 8:23 pm
>>>> Subject: [Csnd] lib_csnd.dylib doubles problem
>>>> To: csound@lists.bath.ac.uk
>>>>
>>>>
>>>> -----------------------------------------------------------
>>>> |
>>>>
>>>> Hello,
>>>>
>>>>
>>>> I think there is still something wrong with lib_csnd.dylib doubles on OS
>>>> X
>>>> as blue crashes when run in api mode. >
>>>>
>>>> Thanks>
>>>>
>>>> Peiman |
>>>>
>>>> -----------------------------------------------------------
>>>>
>>>>
>>>>
>>>> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
>>>> National University of Ireland, Maynooth
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>>> csound"
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hello,

Just to mention that blue was working in api mode about one year ago  
when I used it extensively (also on OS X). So either something has  
changed in blue or in Csound since then. It might be worth trying it  
with the older version of blue to test.

P

On 21 May 2010, at 19:49, Steven Yi wrote:

> I haven't had a chance to look at the CSDPlayer example, as I'm still
> having issues with getting the API to load at all.  I think there's an
> issue having to do with 32bit/64bit versions of Java and trying to
> load the jnilib (jnilib is i386, but for some reason running blue from
> the app launcher keeps using the x86_64 version of Java).  I think I
> need to figure out this issue first, then can work on the API issue.
>
> On Fri, May 21, 2010 at 10:14 AM, Victor Lazzarini
>  wrote:
>> OK, if that is the case, then take a look at the CSDPlayer example  
>> in the
>> sources. That works fine, so it should help you in your diagnosis.
>> On 21 May 2010, at 15:09, Steven Yi wrote:
>>
>>> Hi Victor,
>>>
>>> Thanks very much for your time in looking into this!
>>>
>>> I'll have to look further into what is going on, as blue uses Csound
>>> and not CppSound, and it also uses Java threads and does not use
>>> CsoundPerformanceThread.  I think I'll finally have some time to  
>>> look
>>> into this today and will try to see if I can get anywhere.
>>>
>>> Thanks!
>>> steven
>>>
>>>
>>> On Fri, May 21, 2010 at 7:55 AM, Victor Lazzarini
>>>  wrote:
>>>>
>>>> And I bet this is because CppSound is derived from two base  
>>>> classes and I
>>>> don't think Java likes that.
>>>> Victor
>>>> On 21 May 2010, at 12:39, Victor Lazzarini wrote:
>>>>
>>>> Further findings:
>>>> The problem I identified was in CsoundEditor.java: it seems that a
>>>> combination of the CppSound class and CsoundPerformanceThread is  
>>>> causing
>>>> the
>>>> problem. However, using the Csound class and  
>>>> CsoundPerformanceThread is
>>>> perfectly fine. I have adapted CSDPlayer to use it and it's OK.
>>>> So if blue is using CppSound, then the problem might be there.
>>>> Regards
>>>> Victor
>>>>
>>>> On 21 May 2010, at 12:16, Victor Lazzarini wrote:
>>>>
>>>> More diagnostics:
>>>> it seems that the CsoundPerformanceThread class is causing  
>>>> crashes on
>>>> Java.
>>>> CSDPlayer uses its own java threading and does not rely on that  
>>>> class,
>>>> that
>>>> is one of the reasons it works.
>>>> Victor
>>>> On 21 May 2010, at 11:56, Victor Lazzarini wrote:
>>>>
>>>> Ok, just to complement this discussion. I built the CSDPlayer.java
>>>> example
>>>> from examples/java  and tested it (the same program, no re-build  
>>>> was
>>>> needed)
>>>> with both floats and doubles (by changing the symlink in
>>>> JavaVM.framework/libraries). Both versions ran without any  
>>>> problems, as I
>>>> would expect.
>>>> I think it should not matter whether the link is to floats or  
>>>> doubles,
>>>> unless you are assuming specific MYFLT sizes in the java code,  
>>>> wrapping
>>>> pointers or something like that. The same java compiled code  
>>>> should work
>>>> with both. So I would expect blue to work with the installation  
>>>> as it is
>>>> (I
>>>> mean make a sound and not crash). I can change the installation to
>>>> doubles
>>>> (as I will probably do for python as well), but that should not  
>>>> matter as
>>>> far as running the program is concerned.
>>>> So I guess there might be a problem elsewhere. The Java module  
>>>> seems to
>>>> be
>>>> OK.
>>>> Victor
>>>>
>>>>
>>>> On 21 May 2010, at 11:27, Peiman Khosravi wrote:
>>>>
>>>> Just to confirm, I tried changing it too and blue still crashes.
>>>> P
>>>> On 21 May 2010, at 11:20, Victor Lazzarini wrote:
>>>>
>>>> I had a quick look this morning and what I found was: the csound  
>>>> module
>>>> installed in JavaVM.framework/Resources/Libraries is a symlink to  
>>>> the
>>>> 32-bit
>>>> framework. I have changed this to point to the 64-bit framework,  
>>>> but blue
>>>> is
>>>> still crashing
>>>> 21/05/2010 11:08:40 com.apple.launchd[119]  
>>>> ([0x0-0x42a42a].blue[39891])
>>>> Exited with exit code: 138
>>>> 21/05/2010 11:09:30 [0x0-0x42f42f].blue[39994] Invalid memory  
>>>> access of
>>>> location 0x4 eip=0x5170b708
>>>> I checked the lib_csnd.jnilib  and they are correct (both  
>>>> lib_csnd and
>>>> CsoundLib are 64bit).
>>>> Perhaps Steven might be able to cast some light onto this.
>>>> Victor
>>>>
>>>> On 21 May 2010, at 10:58, Peiman Khosravi wrote:
>>>>
>>>> Thanks very much.
>>>> Peiman
>>>> On 21 May 2010, at 10:27, Victor Lazzarini wrote:
>>>>
>>>> yes, the link should be to lib_csnd.dylib inside CsoundLIb64  
>>>> framework.
>>>> But
>>>> Steven is right to say
>>>> that the Java lib that is installed is linked to the 32 bit  
>>>> framework.
>>>> I'll
>>>> see if I can sort all of this out for
>>>> the next release.
>>>> Victor
>>>>
>>>> On 21 May 2010, at 00:47, Peiman Khosravi wrote:
>>>>
>>>> I don't know if it's normal but I have these in /usr/local/lib
>>>> lib_csnd.dylib
>>>> lib_csnd64.dylib
>>>> libcsnd.dylib
>>>> libcsnd64.dylib
>>>> And only lib_csnd.dylib in CsoundLib64.framework
>>>> And I get this line in blue's crash report:
>>>> Thread 23 Crashed:
>>>> 0   lib_csnd.dylib                 0x51266738
>>>> CsoundCallbackWrapper::SetMessageCallback() + 24
>>>>
>>>> On 20 May 2010, at 21:07, Victor.Lazzarini@nuim.ie wrote:
>>>>
>>>> yes, there might be a link problem there. Did you try the  
>>>> lib_csnd.dylib
>>>> inside the framework or the one in /usr/local/lib?
>>>>
>>>> ----- Original Message -----
>>>> From: Peiman Khosravi 
>>>> Date: Thursday, May 20, 2010 8:23 pm
>>>> Subject: [Csnd] lib_csnd.dylib doubles problem
>>>> To: csound@lists.bath.ac.uk
>>>>
>>>>
>>>> -----------------------------------------------------------
>>>> |
>>>>
>>>> Hello,
>>>>
>>>>
>>>> I think there is still something wrong with lib_csnd.dylib  
>>>> doubles on OS
>>>> X
>>>> as blue crashes when run in api mode. >
>>>>
>>>> Thanks>
>>>>
>>>> Peiman |
>>>>
>>>> -----------------------------------------------------------
>>>>
>>>>
>>>>
>>>> Dr Victor Lazzarini, Senior Lecturer, Dept. of Music,
>>>> National University of Ireland, Maynooth
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe
>>>> csound"
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>           https://sourceforge.net/tracker/? 
>>> group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe
>> csound"
>>
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"