OK, so I've been working on the 5.12 packages. I've had one issue, and
I'd like to take the opportunity to ask a few questions:

1. The old parser is segfaulting on me. The new one seems to work OK.
Do I want to investigate the segfault, or should I just move to the
new parser? In other words, is the old parser deprecated?
2. The windows binaries are not shipping winsound or csound5gui, and
I've been thinking about dropping them too. Should I?
3. As a more general inquiry, I would like input on what features to
enable or disable. Here is the current (not-yet-released) scons line
I'm using. Should I disable/enable something? I've been thinking about
disabling portaudio and pulseaudio, and just default to alsa.

useDouble=1 usePortAudio=1 usePortMIDI=1 \
           useJack=1 useFLTK=1 buildCsound5GUI=1 buildWinsound=1 \
           generatePdf=0 useOSC=1 buildPythonOpcodes=1 \
           dynamicCsoundLibrary=1 buildPDClass=1 \
           buildDSSI=1 buildUtilities=1 buildTclcsound=1 \
           buildVirtual=1 buildInterfaces=1 buildJavaWrapper=1 \
           buildPythonWrapper=1 buildLuaWrapper=1 \
           buildCSEditor=1 buildNewParser=1 prefix=/usr buildCsoundAC=1 \
           useUDP=1 buildRelease=1 buildSDFT=1 buildImageOpcodes=1 \
           useGettext=1 LINKFLAGS='-Wl,-as-needed' SHLINKFLAGS='-Wl,-as-needed'

On Sun, Mar 28, 2010 at 13:49, Victor Lazzarini
 wrote:
> Many of the bugs were OSX-related. I believe the 5.12 sources are OK,
> but maybe John ffitch can say which tag is the most correct.
> On 28 Mar 2010, at 16:42, Felipe Sateler wrote:
>
>> On Fri, Mar 19, 2010 at 05:06, Andres Cabrera
>>  wrote:
>>> Hi,
>>>
>>> It might be important to use the fixes added for 5.12.3. Is there a
>>> source release for that package?
>>
>> Ping on this? I remember that 5.12 had a number of bugs that were then
>> fixed. Has the source release been updated?
>>
>> --
>>
>> Saludos,
>> Felipe Sateler
>>
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>



-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

On Thu, Apr 1, 2010 at 13:15, Felipe Sateler  wrote:
> OK, so I've been working on the 5.12 packages. I've had one issue, and
> I'd like to take the opportunity to ask a few questions:
>
> 1. The old parser is segfaulting on me. The new one seems to work OK.
> Do I want to investigate the segfault, or should I just move to the
> new parser? In other words, is the old parser deprecated?
> 2. The windows binaries are not shipping winsound or csound5gui, and
> I've been thinking about dropping them too. Should I?
> 3. As a more general inquiry, I would like input on what features to
> enable or disable. Here is the current (not-yet-released) scons line
> I'm using. Should I disable/enable something? I've been thinking about
> disabling portaudio and pulseaudio, and just default to alsa.
>
> useDouble=1 usePortAudio=1 usePortMIDI=1 \
>           useJack=1 useFLTK=1 buildCsound5GUI=1 buildWinsound=1 \
>           generatePdf=0 useOSC=1 buildPythonOpcodes=1 \
>           dynamicCsoundLibrary=1 buildPDClass=1 \
>           buildDSSI=1 buildUtilities=1 buildTclcsound=1 \
>           buildVirtual=1 buildInterfaces=1 buildJavaWrapper=1 \
>           buildPythonWrapper=1 buildLuaWrapper=1 \
>           buildCSEditor=1 buildNewParser=1 prefix=/usr buildCsoundAC=1 \
>           useUDP=1 buildRelease=1 buildSDFT=1 buildImageOpcodes=1 \
>           useGettext=1 LINKFLAGS='-Wl,-as-needed' SHLINKFLAGS='-Wl,-as-needed'

I forgot to add, I'm adding these options conditionally:

useALSA=1 -> disabled on non-linux ports of debian
gcc4opt=generic -> only on i386 and amd64

-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/li

We should investigate the segfault, the new parser is not yet ready  
for production in 5.12. Is this relating to the '/' and addition of  
input nchnls option?

Victor
On 1 Apr 2010, at 17:15, Felipe Sateler wrote:

> 1. The old parser is segfaulting on me. The new one seems to work OK.
> Do I want to investigate the segfault, or should I just move to the
> new parser? In other words, is the old parser deprecated?


------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Agreed on new parser, not yet ready for prime time.

As for pulseaudio and portaudio, I think having those would be good.
Wouldn't defaulting to alsa mean a modification to .csoundrc, or is
that in the code?

On Thu, Apr 1, 2010 at 12:28 PM, Victor Lazzarini
 wrote:
> We should investigate the segfault, the new parser is not yet ready
> for production in 5.12. Is this relating to the '/' and addition of
> input nchnls option?
>
> Victor
> On 1 Apr 2010, at 17:15, Felipe Sateler wrote:
>
>> 1. The old parser is segfaulting on me. The new one seems to work OK.
>> Do I want to investigate the segfault, or should I just move to the
>> new parser? In other words, is the old parser deprecated?
>
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

OK, I'll look into the segfault.

As for defaulting to alsa, it means patching the code to automatically
load the alsa plugin instead of portaudio. The OLPC has already an
ifdef for this, in Top/csound.c round line 946.

On Thu, Apr 1, 2010 at 16:35, Steven Yi  wrote:
> Agreed on new parser, not yet ready for prime time.
>
> As for pulseaudio and portaudio, I think having those would be good.
> Wouldn't defaulting to alsa mean a modification to .csoundrc, or is
> that in the code?
>
> On Thu, Apr 1, 2010 at 12:28 PM, Victor Lazzarini
>  wrote:
>> We should investigate the segfault, the new parser is not yet ready
>> for production in 5.12. Is this relating to the '/' and addition of
>> input nchnls option?
>>
>> Victor
>> On 1 Apr 2010, at 17:15, Felipe Sateler wrote:
>>
>>> 1. The old parser is segfaulting on me. The new one seems to work OK.
>>> Do I want to investigate the segfault, or should I just move to the
>>> new parser? In other words, is the old parser deprecated?
>>
>>
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>



-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Some tests are failing: tests 25 and 32.

For test25 I get:

orch compiler:
	instr	1	
out inconsistent with global nchnls (2); replaced with outs
error:  insufficient required arguments, line 17:
  out asig
1 syntax errors in orchestra.  compilation invalid

But the test seems broken.

test32 yields a segfault. It appears to be related to the handling of
continuation backslashes. If I reformat test32.csd to avoid the \, it
works.

Anyone volunteer a fix? segfault is in line 586 of rdorch.c, but
trying to trace the logic has so far given me no illumination... What
I've gathered is that ST(linadr)[lincnt]; is returning NULL after line
13 in test32.csd. Maybe the parser is getting confused with real/fake
line numbers?



-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

This has been fixed in CVS, maybe you can  create a patch for your  
5.12 sources? Maybe there is a later tag with the fixes?
On 3 Apr 2010, at 23:07, Felipe Sateler wrote:

> Some tests are failing: tests 25 and 32.
>
> For test25 I get:
>
> orch compiler:
> 	instr	1	
> out inconsistent with global nchnls (2); replaced with outs
> error:  insufficient required arguments, line 17:
>  out asig
> 1 syntax errors in orchestra.  compilation invalid
>
> But the test seems broken.
>
> test32 yields a segfault. It appears to be related to the handling of
> continuation backslashes. If I reformat test32.csd to avoid the \, it
> works.
>
> Anyone volunteer a fix? segfault is in line 586 of rdorch.c, but
> trying to trace the logic has so far given me no illumination... What
> I've gathered is that ST(linadr)[lincnt]; is returning NULL after line
> 13 in test32.csd. Maybe the parser is getting confused with real/fake
> line numbers?
>
>
>
> -- 
>
> Saludos,
> Felipe Sateler
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel


------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

I see. However, the fix might be split into 2 commits (1.114 and 1.117
of rdorch.c). Does the fix need both commits, or just the last one?

On Sat, Apr 3, 2010 at 20:00, Victor Lazzarini  wrote:
> This has been fixed in CVS, maybe you can  create a patch for your
> 5.12 sources? Maybe there is a later tag with the fixes?
> On 3 Apr 2010, at 23:07, Felipe Sateler wrote:
>
>> Some tests are failing: tests 25 and 32.
>>
>> For test25 I get:
>>
>> orch compiler:
>>       instr   1
>> out inconsistent with global nchnls (2); replaced with outs
>> error:  insufficient required arguments, line 17:
>>  out asig
>> 1 syntax errors in orchestra.  compilation invalid
>>
>> But the test seems broken.
>>
>> test32 yields a segfault. It appears to be related to the handling of
>> continuation backslashes. If I reformat test32.csd to avoid the \, it
>> works.
>>
>> Anyone volunteer a fix? segfault is in line 586 of rdorch.c, but
>> trying to trace the logic has so far given me no illumination... What
>> I've gathered is that ST(linadr)[lincnt]; is returning NULL after line
>> 13 in test32.csd. Maybe the parser is getting confused with real/fake
>> line numbers?
>>
>>
>>
>> --
>>
>> Saludos,
>> Felipe Sateler
>>
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>



-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
https://lists.sourceforge.net/

Plus, the fix seems to have a typo (line 557, & instead of &&)?

On Sat, Apr 3, 2010 at 20:42, Felipe Sateler  wrote:
> I see. However, the fix might be split into 2 commits (1.114 and 1.117
> of rdorch.c). Does the fix need both commits, or just the last one?
>
> On Sat, Apr 3, 2010 at 20:00, Victor Lazzarini  wrote:
>> This has been fixed in CVS, maybe you can  create a patch for your
>> 5.12 sources? Maybe there is a later tag with the fixes?
>> On 3 Apr 2010, at 23:07, Felipe Sateler wrote:
>>
>>> Some tests are failing: tests 25 and 32.
>>>
>>> For test25 I get:
>>>
>>> orch compiler:
>>>       instr   1
>>> out inconsistent with global nchnls (2); replaced with outs
>>> error:  insufficient required arguments, line 17:
>>>  out asig
>>> 1 syntax errors in orchestra.  compilation invalid
>>>
>>> But the test seems broken.
>>>
>>> test32 yields a segfault. It appears to be related to the handling of
>>> continuation backslashes. If I reformat test32.csd to avoid the \, it
>>> works.
>>>
>>> Anyone volunteer a fix? segfault is in line 586 of rdorch.c, but
>>> trying to trace the logic has so far given me no illumination... What
>>> I've gathered is that ST(linadr)[lincnt]; is returning NULL after line
>>> 13 in test32.csd. Maybe the parser is getting confused with real/fake
>>> line numbers?
>>>
>>>
>>>
>>> --
>>>
>>> Saludos,
>>> Felipe Sateler
>>>
>>> ------------------------------------------------------------------------------
>>> Download Intel® Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>>
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>
>
>
> --
>
> Saludos,
> Felipe Sateler
>



-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
https://lists

OK, I've tested and both changes are needed. I really think that
commit 1.117 contains a typo, though.

On Sat, Apr 3, 2010 at 20:44, Felipe Sateler  wrote:
> Plus, the fix seems to have a typo (line 557, & instead of &&)?
>
> On Sat, Apr 3, 2010 at 20:42, Felipe Sateler  wrote:
>> I see. However, the fix might be split into 2 commits (1.114 and 1.117
>> of rdorch.c). Does the fix need both commits, or just the last one?
>>
>> On Sat, Apr 3, 2010 at 20:00, Victor Lazzarini  wrote:
>>> This has been fixed in CVS, maybe you can  create a patch for your
>>> 5.12 sources? Maybe there is a later tag with the fixes?
>>> On 3 Apr 2010, at 23:07, Felipe Sateler wrote:
>>>
>>>> Some tests are failing: tests 25 and 32.
>>>>
>>>> For test25 I get:
>>>>
>>>> orch compiler:
>>>>       instr   1
>>>> out inconsistent with global nchnls (2); replaced with outs
>>>> error:  insufficient required arguments, line 17:
>>>>  out asig
>>>> 1 syntax errors in orchestra.  compilation invalid
>>>>
>>>> But the test seems broken.
>>>>
>>>> test32 yields a segfault. It appears to be related to the handling of
>>>> continuation backslashes. If I reformat test32.csd to avoid the \, it
>>>> works.
>>>>
>>>> Anyone volunteer a fix? segfault is in line 586 of rdorch.c, but
>>>> trying to trace the logic has so far given me no illumination... What
>>>> I've gathered is that ST(linadr)[lincnt]; is returning NULL after line
>>>> 13 in test32.csd. Maybe the parser is getting confused with real/fake
>>>> line numbers?
>>>>
>>>>
>>>>
>>>> --
>>>>
>>>> Saludos,
>>>> Felipe Sateler
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Download Intel® Parallel Studio Eval
>>>> Try the new software tools for yourself. Speed compiling, find bugs
>>>> proactively, and fine-tune applications for parallel performance.
>>>> See why Intel Parallel Studio got high marks during beta.
>>>> http://p.sf.net/sfu/intel-sw-dev
>>>> _______________________________________________
>>>> Csound-devel mailing list
>>>> Csound-devel@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Download Intel® Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>
>>
>>
>> --
>>
>> Saludos,
>> Felipe Sateler
>>
>
>
>
> --
>
> Saludos,
> Felipe Sateler
>



-- 

Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
http