I'm trying to use csnd's SetInputValueCallback in Java. The protype
looks like this:

SetInputValueCallback(SWIGTYPE_p_f_p_CSOUND__p_q_const__char_p_double__void s)

Any idea on how I can use it? In C I would do something like this:

void inputValueCallback(CSOUND *csound, const char *channelName, MYFLT *value){}
...
csoundSetInputValueCallback(csound, inputValueCallback);

Any ideas?

Wipe that, I just found the JavaCallbackWrapper class. If I don't
figure it out I'll get back to you!

On 16 January 2013 22:45, Rory Walsh  wrote:
> I'm trying to use csnd's SetInputValueCallback in Java. The protype
> looks like this:
>
> SetInputValueCallback(SWIGTYPE_p_f_p_CSOUND__p_q_const__char_p_double__void s)
>
> Any idea on how I can use it? In C I would do something like this:
>
> void inputValueCallback(CSOUND *csound, const char *channelName, MYFLT *value){}
> ...
> csoundSetInputValueCallback(csound, inputValueCallback);
>
> Any ideas?

I hope some Java programmers out there can help. I want to set up a
value callback function that will call a method in my class when it is
called by Csound. I've set up something like this in my main class,
where csound is the Csound object:

callbackWrapper = new CallbackWrapper(csound);
callbackWrapper.SetOutputValueCallback();


In my callback wrapper class I have the following function:

void OutputValueCallback(String _chan, double _val){
//do something
}

What I would like is to able to register one my my main class
functions as the callback function to be called so I can get my method
to do things to the members of my base class. In the C/C++ interface I
can pass a function to SetOutputValueCallback(), I guess this isn't
the case in Java?

Hi Rory,

You could sub-class CallbackWrapper or also use an anonymous class
doing something like:

callbackWrapper = new CallbackWrapper(csound) {
void OutputValueCallback(String _chan, double _val){
   myOtherClass.doSomething(_chan, _val);
}
};
callbackWrapper.SetOutputValueCallback();

I use a subclass in blue here:

https://sourceforge.net/p/bluemusic/blue/ci/a14b51b0bf09cc363c4d7564f84651e09f00a745/tree/blue-ui-core/src/blue/ui/core/render/BlueCallbackWrapper.java

for message callback.

Hope that helps!
steven

On Wed, Jan 16, 2013 at 11:45 PM, Rory Walsh  wrote:
> to do things to the members of my base class. In the C/C++ interface I

Thanks Steven!

On 17 January 2013 12:16, Steven Yi  wrote:
> Hi Rory,
>
> You could sub-class CallbackWrapper or also use an anonymous class
> doing something like:
>
> callbackWrapper = new CallbackWrapper(csound) {
> void OutputValueCallback(String _chan, double _val){
>    myOtherClass.doSomething(_chan, _val);
> }
> };
> callbackWrapper.SetOutputValueCallback();
>
> I use a subclass in blue here:
>
> https://sourceforge.net/p/bluemusic/blue/ci/a14b51b0bf09cc363c4d7564f84651e09f00a745/tree/blue-ui-core/src/blue/ui/core/render/BlueCallbackWrapper.java
>
> for message callback.
>
> Hope that helps!
> steven
>
> On Wed, Jan 16, 2013 at 11:45 PM, Rory Walsh  wrote:
>> to do things to the members of my base class. In the C/C++ interface I
>
>
> 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"
>

Thanks for your help on these ongoing java issue Steven. I have those
callbacks in place now, all working well. My own biggest issue is in
mixing the two java interfaces, csnd and androidCsound. One is
strictly doubles, the other strictly floats. Am I correct in saying
that as it stands users cannot use the same Java code they wrote using
csnd with Android? Wel, I know for somethings it's quite possible, but
not everything?



On 17 January 2013 13:17, Rory Walsh  wrote:
> Thanks Steven!
>
> On 17 January 2013 12:16, Steven Yi  wrote:
>> Hi Rory,
>>
>> You could sub-class CallbackWrapper or also use an anonymous class
>> doing something like:
>>
>> callbackWrapper = new CallbackWrapper(csound) {
>> void OutputValueCallback(String _chan, double _val){
>>    myOtherClass.doSomething(_chan, _val);
>> }
>> };
>> callbackWrapper.SetOutputValueCallback();
>>
>> I use a subclass in blue here:
>>
>> https://sourceforge.net/p/bluemusic/blue/ci/a14b51b0bf09cc363c4d7564f84651e09f00a745/tree/blue-ui-core/src/blue/ui/core/render/BlueCallbackWrapper.java
>>
>> for message callback.
>>
>> Hope that helps!
>> steven
>>
>> On Wed, Jan 16, 2013 at 11:45 PM, Rory Walsh  wrote:
>>> to do things to the members of my base class. In the C/C++ interface I
>>
>>
>> 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"
>>

Right now, the Android Csound lib is compiled with MYFLT=float. If you
write desktop java code with MYFLT=float, then it will be compatible.
If you use double, then it will not be compatible.  We went with
floats build for Android initially as we thought it'd be more
performant for ARM, but maybe we should check performance again with a
double's build, as it would be nice to use doubles consistently across
platforms.  (Note, this situation is probably the same for iOS vs. OSX
Desktop).


On Thu, Jan 17, 2013 at 6:39 PM, Rory Walsh  wrote:
> somethings

As you know, my issue is related to the current work being done to
Csoundo. I have to assume that users are using a doubles build of
Csound for their desktops as it's the defacto install on all desktop
systems. As I understand, MYFLT is a preprocessor directive so one
couldn't for example write an API function that would force the use of
floats because the MYFLT type can't be changed after compilation?
Which brings me to my next assumption, that even if I build and
distribute a float version of cnsd it would still be incompatible with
any double version of Csound users already have installed? It seems
that the only time I get caught is when I have to use SWIG types as
parameters. I can use SetChannel() without a problem across both
interfaces, and it is defined in csnd as
SetChannel(String string, double d)
One solution is for me not to use the CsoundMYFLTArray class and just
go it manually. If need be I can just get Csound to tell me when it's
safe to update tables. But I don't want to do this only to get caught
out somewhere else down the line. Do you have any ideas?

On Jan 17, 2013, at 2:29 PM, Rory Walsh  wrote:

> I have to assume that users are using a doubles build of
> Csound for their desktops as it's the defacto install on all desktop
> systems. 

Rory, could you please clarify this? I'm running an out-of-the-box installation of Csound on my desktop Mac Pro computer, with OSX 10.8.2. This is the printout I see from QTCsound:

" Csound version 5.18.03 (float samples) Oct  6 2012"

Since I have no idea how to specify floats vs. doubles, I'm sure I did nothing to cause Csound to operate with floats rather than doubles. 

Does this mean that some versions of Csound are NOT defaulting to doubles, or does it mean that I'm failing to specify some option?

best,
Lou Cohen
www.jolc.net
www.opensound.org

As far as I know OSX comes with both versions of the library, Victor
will be able to confirm. I'm not sure why CsoundQT would default to
using the float version, or whether it's a conscious decision on the
part of CsoundsQT to use it.

On 17 January 2013 20:31, Louis Cohen  wrote:
>
> On Jan 17, 2013, at 2:29 PM, Rory Walsh  wrote:
>
>> I have to assume that users are using a doubles build of
>> Csound for their desktops as it's the defacto install on all desktop
>> systems.
>
> Rory, could you please clarify this? I'm running an out-of-the-box installation of Csound on my desktop Mac Pro computer, with OSX 10.8.2. This is the printout I see from QTCsound:
>
> " Csound version 5.18.03 (float samples) Oct  6 2012"
>
> Since I have no idea how to specify floats vs. doubles, I'm sure I did nothing to cause Csound to operate with floats rather than doubles.
>
> Does this mean that some versions of Csound are NOT defaulting to doubles, or does it mean that I'm failing to specify some option?
>
> best,
> Lou Cohen
> www.jolc.net
> www.opensound.org
>
>
>
>
>
>
> 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 Rory,

As mentioned, the issue is that the android build is built on floats. One solution that will work here is if we do a doubles build of Csound for Android. I'd need time to do a build and investigate performance differences, as well as if there are any code changes necessary. I'd like to avoid having to do multiple android builds if necessary (which could be tricky, and necessary if performance as a big impact).

Another option would be on your side to create a specific Csoundo for Android that works with float build. I assume then that a user working with processing then would be able to take their code and rebuild on Android and get going. Might be a good option too as you could make an eclipse android library project thg has all the processing customizations and is linked with CsoundAndroid, and perhaps we could even distribute that with he CsoundAndroid release.

Steven

I don't think having several android builds is a god idea either. And
there isn't really much pint I guess in investigate doubles, it's
probably not going to run as fast as float.

> Another option would be on your side to create a specific Csoundo for
> Android that works with float build. I assume then that a user working with

I'm not sure I follow. At present it is a float build. At present we
don't use any eclipse project files at all. The main IDE IS
Processing. We go from there straight to the phone. It's savage :)

p.s. can you enlighten me as to why I don't get the same errors with
SetChannel()?

Don't make assumptions about the speed. Past tests have showed that on
some platforms, doubles are faster than floats; on other platforms
including Windows, floats are faster (15% or so) than doubles.

Try both and measure it.

Regards,
Mike

On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
> I don't think having several android builds is a god idea either. And
> there isn't really much pint I guess in investigate doubles, it's
> probably not going to run as fast as float.
>
>> Another option would be on your side to create a specific Csoundo for
>> Android that works with float build. I assume then that a user working with
>
> I'm not sure I follow. At present it is a float build. At present we
> don't use any eclipse project files at all. The main IDE IS
> Processing. We go from there straight to the phone. It's savage :)
>
> p.s. can you enlighten me as to why I don't get the same errors with
> SetChannel()?
>
>
> 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"
>



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

That's true Mike. I gues we could be surprised by the results.

On 17 January 2013 21:20, Michael Gogins  wrote:
> Don't make assumptions about the speed. Past tests have showed that on
> some platforms, doubles are faster than floats; on other platforms
> including Windows, floats are faster (15% or so) than doubles.
>
> Try both and measure it.
>
> Regards,
> Mike
>
> On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
>> I don't think having several android builds is a god idea either. And
>> there isn't really much pint I guess in investigate doubles, it's
>> probably not going to run as fast as float.
>>
>>> Another option would be on your side to create a specific Csoundo for
>>> Android that works with float build. I assume then that a user working with
>>
>> I'm not sure I follow. At present it is a float build. At present we
>> don't use any eclipse project files at all. The main IDE IS
>> Processing. We go from there straight to the phone. It's savage :)
>>
>> p.s. can you enlighten me as to why I don't get the same errors with
>> SetChannel()?
>>
>>
>> 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"
>>
>
>
>
> --
> Michael Gogins
> Irreducible Productions
> http://www.michael-gogins.com
> Michael dot Gogins at gmail dot com
>
>
> 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"
>

http://stackoverflow.com/questions/3310907/what-are-the-advantages-of-armv7-over-armv6-when-compiling-iphone-apps

Seems like they took out hardware double precision support for armv7. I think on Android we compile w/Neon. Still, testing required to verify for both platforms.

Rory: How is csoundo compiled to work with floats if it's used on the desktop? I'm a bit confused, but also on my cell phone. I'll be back online tomorrow when on campus and will be able to see what's going on better.

That's the thing. We have a member varialbe of type Csound in our
class, but we create an object of type AndroidCsound when the use is
running in Android mode. Once our csound member variable is created as
an AndroidCsound object everything else works fine on Android. I'm
still trying to get my head around it, but it means users only have to
swap out the following code to swap between desktop and android modes
when working with Csoundo in processing:

//Android mode
cs = new Csoundo(this, super.getApplicationContext());
//Java mode
cs = new Csoundo(this, "filename..csd");

Consider the following processing sketch. In it we can happily move
between devices by simply changing the constructor. It's pretty neat,
but I'm starting to suspect we are slightly lucky it works?

import csoundo.*;

Csoundo cs;
int cnt=0;
float x,y,amp;

void setup() {
    size(400, 400);
    //orientation(LANDSCAPE);
    background(0);
    frameRate(5);
    smooth();
    //Android Mode (comment out either)
    //cs = new Csoundo(this, super.getApplicationContext());
    //Java Mode
    cs = new Csoundo(this, "Random.csd");
    cs.run();
}

void draw() {
    x = random(width);
    y = random(height);
    amp = random(1);
    cs.setChn("pitch", y);
    cs.setChn("pan", x/width);
    cs.setChn("amp", amp);
    fill(255, (x/width) * 255, 0, (y/height) * 255);
    ellipse(x,y, 50*amp, 50*amp);
    fill(0, 0, 0, 10);
    rect(0, 0, width, height);
}



On 17 January 2013 23:08, Steven Yi  wrote:
> http://stackoverflow.com/questions/3310907/what-are-the-advantages-of-armv7-over-armv6-when-compiling-iphone-apps
>
> Seems like they took out hardware double precision support for armv7. I
> think on Android we compile w/Neon. Still, testing required to verify for
> both platforms.
>
> Rory: How is csoundo compiled to work with floats if it's used on the
> desktop? I'm a bit confused, but also on my cell phone. I'll be back online
> tomorrow when on campus and will be able to see what's going on better.
>
> On Jan 17, 2013 9:39 PM, "Rory Walsh"  wrote:
>>
>> That's true Mike. I gues we could be surprised by the results.
>>
>> On 17 January 2013 21:20, Michael Gogins  wrote:
>> > Don't make assumptions about the speed. Past tests have showed that on
>> > some platforms, doubles are faster than floats; on other platforms
>> > including Windows, floats are faster (15% or so) than doubles.
>> >
>> > Try both and measure it.
>> >
>> > Regards,
>> > Mike
>> >
>> > On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
>> >> I don't think having several android builds is a god idea either. And
>> >> there isn't really much pint I guess in investigate doubles, it's
>> >> probably not going to run as fast as float.
>> >>
>> >>> Another option would be on your side to create a specific Csoundo for
>> >>> Android that works with float build. I assume then that a user working
>> >>> with
>> >>
>> >> I'm not sure I follow. At present it is a float build. At present we
>> >> don't use any eclipse project files at all. The main IDE IS
>> >> Processing. We go from there straight to the phone. It's savage :)
>> >>
>> >> p.s. can you enlighten me as to why I don't get the same errors with
>> >> SetChannel()?
>> >>
>> >>
>> >> 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"
>> >>
>> >
>> >
>> >
>> > --
>> > Michael Gogins
>> > Irreducible Productions
>> > http://www.michael-gogins.com
>> > Michael dot Gogins at gmail dot com
>> >
>> >
>> > 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 Rory,

I took a look at Csoundo from GIT but I'm still not sure of a couple
things. It seems you distribute in the csoundo folder both csnd.jar
and csoundanroid.jar.  They should be used mutually exclusive of each
other, as they both contain swig generated classes for csound (you'll
likely get classloader issues and may get the wrong version of a
class). Did I understand that part of what is going on correctly?
Also, when you compile Csoundo.jar, is it compiled against csnd.jar or
csoundandroid.jar?  Again, this should be done only against one or the
other (as they contain classes with mostly the same names but
differences due to float/double).

steven



On Thu, Jan 17, 2013 at 11:18 PM, Rory Walsh  wrote:
> That's the thing. We have a member varialbe of type Csound in our
> class, but we create an object of type AndroidCsound when the use is
> running in Android mode. Once our csound member variable is created as
> an AndroidCsound object everything else works fine on Android. I'm
> still trying to get my head around it, but it means users only have to
> swap out the following code to swap between desktop and android modes
> when working with Csoundo in processing:
>
> //Android mode
> cs = new Csoundo(this, super.getApplicationContext());
> //Java mode
> cs = new Csoundo(this, "filename..csd");
>
> Consider the following processing sketch. In it we can happily move
> between devices by simply changing the constructor. It's pretty neat,
> but I'm starting to suspect we are slightly lucky it works?
>
> import csoundo.*;
>
> Csoundo cs;
> int cnt=0;
> float x,y,amp;
>
> void setup() {
>     size(400, 400);
>     //orientation(LANDSCAPE);
>     background(0);
>     frameRate(5);
>     smooth();
>     //Android Mode (comment out either)
>     //cs = new Csoundo(this, super.getApplicationContext());
>     //Java Mode
>     cs = new Csoundo(this, "Random.csd");
>     cs.run();
> }
>
> void draw() {
>     x = random(width);
>     y = random(height);
>     amp = random(1);
>     cs.setChn("pitch", y);
>     cs.setChn("pan", x/width);
>     cs.setChn("amp", amp);
>     fill(255, (x/width) * 255, 0, (y/height) * 255);
>     ellipse(x,y, 50*amp, 50*amp);
>     fill(0, 0, 0, 10);
>     rect(0, 0, width, height);
> }
>
>
>
> On 17 January 2013 23:08, Steven Yi  wrote:
>> http://stackoverflow.com/questions/3310907/what-are-the-advantages-of-armv7-over-armv6-when-compiling-iphone-apps
>>
>> Seems like they took out hardware double precision support for armv7. I
>> think on Android we compile w/Neon. Still, testing required to verify for
>> both platforms.
>>
>> Rory: How is csoundo compiled to work with floats if it's used on the
>> desktop? I'm a bit confused, but also on my cell phone. I'll be back online
>> tomorrow when on campus and will be able to see what's going on better.
>>
>> On Jan 17, 2013 9:39 PM, "Rory Walsh"  wrote:
>>>
>>> That's true Mike. I gues we could be surprised by the results.
>>>
>>> On 17 January 2013 21:20, Michael Gogins  wrote:
>>> > Don't make assumptions about the speed. Past tests have showed that on
>>> > some platforms, doubles are faster than floats; on other platforms
>>> > including Windows, floats are faster (15% or so) than doubles.
>>> >
>>> > Try both and measure it.
>>> >
>>> > Regards,
>>> > Mike
>>> >
>>> > On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
>>> >> I don't think having several android builds is a god idea either. And
>>> >> there isn't really much pint I guess in investigate doubles, it's
>>> >> probably not going to run as fast as float.
>>> >>
>>> >>> Another option would be on your side to create a specific Csoundo for
>>> >>> Android that works with float build. I assume then that a user working
>>> >>> with
>>> >>
>>> >> I'm not sure I follow. At present it is a float build. At present we
>>> >> don't use any eclipse project files at all. The main IDE IS
>>> >> Processing. We go from there straight to the phone. It's savage :)
>>> >>
>>> >> p.s. can you enlighten me as to why I don't get the same errors with
>>> >> SetChannel()?
>>> >>
>>> >>
>>> >> 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"
>>> >>
>>> >
>>> >
>>> >
>>> > --
>>> > Michael Gogins
>>> > Irreducible Productions
>>> > http://www.michael-gogins.com
>>> > Michael dot Gogins at gmail dot com
>>> >
>>> >
>>> > 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"
>

When users hit send to device in Processing only the csoundandroid
library gets sent to the device. And when I compile Csoundo I compile
with csnd.jar, as far as I can tell at least. So far we've not had any
issues?

On 18 January 2013 12:30, Steven Yi  wrote:
> Hi Rory,
>
> I took a look at Csoundo from GIT but I'm still not sure of a couple
> things. It seems you distribute in the csoundo folder both csnd.jar
> and csoundanroid.jar.  They should be used mutually exclusive of each
> other, as they both contain swig generated classes for csound (you'll
> likely get classloader issues and may get the wrong version of a
> class). Did I understand that part of what is going on correctly?
> Also, when you compile Csoundo.jar, is it compiled against csnd.jar or
> csoundandroid.jar?  Again, this should be done only against one or the
> other (as they contain classes with mostly the same names but
> differences due to float/double).
>
> steven
>
>
>
> On Thu, Jan 17, 2013 at 11:18 PM, Rory Walsh  wrote:
>> That's the thing. We have a member varialbe of type Csound in our
>> class, but we create an object of type AndroidCsound when the use is
>> running in Android mode. Once our csound member variable is created as
>> an AndroidCsound object everything else works fine on Android. I'm
>> still trying to get my head around it, but it means users only have to
>> swap out the following code to swap between desktop and android modes
>> when working with Csoundo in processing:
>>
>> //Android mode
>> cs = new Csoundo(this, super.getApplicationContext());
>> //Java mode
>> cs = new Csoundo(this, "filename..csd");
>>
>> Consider the following processing sketch. In it we can happily move
>> between devices by simply changing the constructor. It's pretty neat,
>> but I'm starting to suspect we are slightly lucky it works?
>>
>> import csoundo.*;
>>
>> Csoundo cs;
>> int cnt=0;
>> float x,y,amp;
>>
>> void setup() {
>>     size(400, 400);
>>     //orientation(LANDSCAPE);
>>     background(0);
>>     frameRate(5);
>>     smooth();
>>     //Android Mode (comment out either)
>>     //cs = new Csoundo(this, super.getApplicationContext());
>>     //Java Mode
>>     cs = new Csoundo(this, "Random.csd");
>>     cs.run();
>> }
>>
>> void draw() {
>>     x = random(width);
>>     y = random(height);
>>     amp = random(1);
>>     cs.setChn("pitch", y);
>>     cs.setChn("pan", x/width);
>>     cs.setChn("amp", amp);
>>     fill(255, (x/width) * 255, 0, (y/height) * 255);
>>     ellipse(x,y, 50*amp, 50*amp);
>>     fill(0, 0, 0, 10);
>>     rect(0, 0, width, height);
>> }
>>
>>
>>
>> On 17 January 2013 23:08, Steven Yi  wrote:
>>> http://stackoverflow.com/questions/3310907/what-are-the-advantages-of-armv7-over-armv6-when-compiling-iphone-apps
>>>
>>> Seems like they took out hardware double precision support for armv7. I
>>> think on Android we compile w/Neon. Still, testing required to verify for
>>> both platforms.
>>>
>>> Rory: How is csoundo compiled to work with floats if it's used on the
>>> desktop? I'm a bit confused, but also on my cell phone. I'll be back online
>>> tomorrow when on campus and will be able to see what's going on better.
>>>
>>> On Jan 17, 2013 9:39 PM, "Rory Walsh"  wrote:
>>>>
>>>> That's true Mike. I gues we could be surprised by the results.
>>>>
>>>> On 17 January 2013 21:20, Michael Gogins  wrote:
>>>> > Don't make assumptions about the speed. Past tests have showed that on
>>>> > some platforms, doubles are faster than floats; on other platforms
>>>> > including Windows, floats are faster (15% or so) than doubles.
>>>> >
>>>> > Try both and measure it.
>>>> >
>>>> > Regards,
>>>> > Mike
>>>> >
>>>> > On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
>>>> >> I don't think having several android builds is a god idea either. And
>>>> >> there isn't really much pint I guess in investigate doubles, it's
>>>> >> probably not going to run as fast as float.
>>>> >>
>>>> >>> Another option would be on your side to create a specific Csoundo for
>>>> >>> Android that works with float build. I assume then that a user working
>>>> >>> with
>>>> >>
>>>> >> I'm not sure I follow. At present it is a float build. At present we
>>>> >> don't use any eclipse project files at all. The main IDE IS
>>>> >> Processing. We go from there straight to the phone. It's savage :)
>>>> >>
>>>> >> p.s. can you enlighten me as to why I don't get the same errors with
>>>> >> SetChannel()?
>>>> >>
>>>> >>
>>>> >> 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"
>>>> >>
>>>> >
>>>> >
>>>> >
>>>> > --
>>>> > Michael Gogins
>>>> > Irreducible Productions
>>>> > http://www.michael-gogins.com
>>>> > Michael dot Gogins at gmail dot com
>>>> >
>>>> >
>>>> > 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, now I'm triply confused.  :P What's the issue again?  You have
csoundo/processing sketches working on both desktop and Android it
seems.

On Fri, Jan 18, 2013 at 12:35 PM, Rory Walsh  wrote:
> When users hit send to device in Processing only the csoundandroid
> library gets sent to the device. And when I compile Csoundo I compile
> with csnd.jar, as far as I can tell at least. So far we've not had any
> issues?
>
> On 18 January 2013 12:30, Steven Yi  wrote:
>> Hi Rory,
>>
>> I took a look at Csoundo from GIT but I'm still not sure of a couple
>> things. It seems you distribute in the csoundo folder both csnd.jar
>> and csoundanroid.jar.  They should be used mutually exclusive of each
>> other, as they both contain swig generated classes for csound (you'll
>> likely get classloader issues and may get the wrong version of a
>> class). Did I understand that part of what is going on correctly?
>> Also, when you compile Csoundo.jar, is it compiled against csnd.jar or
>> csoundandroid.jar?  Again, this should be done only against one or the
>> other (as they contain classes with mostly the same names but
>> differences due to float/double).
>>
>> steven
>>
>>
>>
>> On Thu, Jan 17, 2013 at 11:18 PM, Rory Walsh  wrote:
>>> That's the thing. We have a member varialbe of type Csound in our
>>> class, but we create an object of type AndroidCsound when the use is
>>> running in Android mode. Once our csound member variable is created as
>>> an AndroidCsound object everything else works fine on Android. I'm
>>> still trying to get my head around it, but it means users only have to
>>> swap out the following code to swap between desktop and android modes
>>> when working with Csoundo in processing:
>>>
>>> //Android mode
>>> cs = new Csoundo(this, super.getApplicationContext());
>>> //Java mode
>>> cs = new Csoundo(this, "filename..csd");
>>>
>>> Consider the following processing sketch. In it we can happily move
>>> between devices by simply changing the constructor. It's pretty neat,
>>> but I'm starting to suspect we are slightly lucky it works?
>>>
>>> import csoundo.*;
>>>
>>> Csoundo cs;
>>> int cnt=0;
>>> float x,y,amp;
>>>
>>> void setup() {
>>>     size(400, 400);
>>>     //orientation(LANDSCAPE);
>>>     background(0);
>>>     frameRate(5);
>>>     smooth();
>>>     //Android Mode (comment out either)
>>>     //cs = new Csoundo(this, super.getApplicationContext());
>>>     //Java Mode
>>>     cs = new Csoundo(this, "Random.csd");
>>>     cs.run();
>>> }
>>>
>>> void draw() {
>>>     x = random(width);
>>>     y = random(height);
>>>     amp = random(1);
>>>     cs.setChn("pitch", y);
>>>     cs.setChn("pan", x/width);
>>>     cs.setChn("amp", amp);
>>>     fill(255, (x/width) * 255, 0, (y/height) * 255);
>>>     ellipse(x,y, 50*amp, 50*amp);
>>>     fill(0, 0, 0, 10);
>>>     rect(0, 0, width, height);
>>> }
>>>
>>>
>>>
>>> On 17 January 2013 23:08, Steven Yi  wrote:
>>>> http://stackoverflow.com/questions/3310907/what-are-the-advantages-of-armv7-over-armv6-when-compiling-iphone-apps
>>>>
>>>> Seems like they took out hardware double precision support for armv7. I
>>>> think on Android we compile w/Neon. Still, testing required to verify for
>>>> both platforms.
>>>>
>>>> Rory: How is csoundo compiled to work with floats if it's used on the
>>>> desktop? I'm a bit confused, but also on my cell phone. I'll be back online
>>>> tomorrow when on campus and will be able to see what's going on better.
>>>>
>>>> On Jan 17, 2013 9:39 PM, "Rory Walsh"  wrote:
>>>>>
>>>>> That's true Mike. I gues we could be surprised by the results.
>>>>>
>>>>> On 17 January 2013 21:20, Michael Gogins  wrote:
>>>>> > Don't make assumptions about the speed. Past tests have showed that on
>>>>> > some platforms, doubles are faster than floats; on other platforms
>>>>> > including Windows, floats are faster (15% or so) than doubles.
>>>>> >
>>>>> > Try both and measure it.
>>>>> >
>>>>> > Regards,
>>>>> > Mike
>>>>> >
>>>>> > On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
>>>>> >> I don't think having several android builds is a god idea either. And
>>>>> >> there isn't really much pint I guess in investigate doubles, it's
>>>>> >> probably not going to run as fast as float.
>>>>> >>
>>>>> >>> Another option would be on your side to create a specific Csoundo for
>>>>> >>> Android that works with float build. I assume then that a user working
>>>>> >>> with
>>>>> >>
>>>>> >> I'm not sure I follow. At present it is a float build. At present we
>>>>> >> don't use any eclipse project files at all. The main IDE IS
>>>>> >> Processing. We go from there straight to the phone. It's savage :)
>>>>> >>
>>>>> >> p.s. can you enlighten me as to why I don't get the same errors with
>>>>> >> SetChannel()?
>>>>> >>
>>>>> >>
>>>>> >> 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"
>>>>> >>
>>>>> >
>>>>> >
>>>>> >
>>>>> > --
>>>>> > Michael Gogins
>>>>> > Irreducible Productions
>>>>> > http://www.michael-gogins.com
>>>>> > Michael dot Gogins at gmail dot com
>>>>> >
>>>>> >
>>>>> > 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"
>

Just FYI to list, I chatted with Rory and I think we've figured out
some things.  We'll experiment offlist for this and hopefully get this
sorted out. :)

On Fri, Jan 18, 2013 at 12:41 PM, Steven Yi  wrote:
> Okay, now I'm triply confused.  :P What's the issue again?  You have
> csoundo/processing sketches working on both desktop and Android it
> seems.
>
> On Fri, Jan 18, 2013 at 12:35 PM, Rory Walsh  wrote:
>> When users hit send to device in Processing only the csoundandroid
>> library gets sent to the device. And when I compile Csoundo I compile
>> with csnd.jar, as far as I can tell at least. So far we've not had any
>> issues?
>>
>> On 18 January 2013 12:30, Steven Yi  wrote:
>>> Hi Rory,
>>>
>>> I took a look at Csoundo from GIT but I'm still not sure of a couple
>>> things. It seems you distribute in the csoundo folder both csnd.jar
>>> and csoundanroid.jar.  They should be used mutually exclusive of each
>>> other, as they both contain swig generated classes for csound (you'll
>>> likely get classloader issues and may get the wrong version of a
>>> class). Did I understand that part of what is going on correctly?
>>> Also, when you compile Csoundo.jar, is it compiled against csnd.jar or
>>> csoundandroid.jar?  Again, this should be done only against one or the
>>> other (as they contain classes with mostly the same names but
>>> differences due to float/double).
>>>
>>> steven
>>>
>>>
>>>
>>> On Thu, Jan 17, 2013 at 11:18 PM, Rory Walsh  wrote:
>>>> That's the thing. We have a member varialbe of type Csound in our
>>>> class, but we create an object of type AndroidCsound when the use is
>>>> running in Android mode. Once our csound member variable is created as
>>>> an AndroidCsound object everything else works fine on Android. I'm
>>>> still trying to get my head around it, but it means users only have to
>>>> swap out the following code to swap between desktop and android modes
>>>> when working with Csoundo in processing:
>>>>
>>>> //Android mode
>>>> cs = new Csoundo(this, super.getApplicationContext());
>>>> //Java mode
>>>> cs = new Csoundo(this, "filename..csd");
>>>>
>>>> Consider the following processing sketch. In it we can happily move
>>>> between devices by simply changing the constructor. It's pretty neat,
>>>> but I'm starting to suspect we are slightly lucky it works?
>>>>
>>>> import csoundo.*;
>>>>
>>>> Csoundo cs;
>>>> int cnt=0;
>>>> float x,y,amp;
>>>>
>>>> void setup() {
>>>>     size(400, 400);
>>>>     //orientation(LANDSCAPE);
>>>>     background(0);
>>>>     frameRate(5);
>>>>     smooth();
>>>>     //Android Mode (comment out either)
>>>>     //cs = new Csoundo(this, super.getApplicationContext());
>>>>     //Java Mode
>>>>     cs = new Csoundo(this, "Random.csd");
>>>>     cs.run();
>>>> }
>>>>
>>>> void draw() {
>>>>     x = random(width);
>>>>     y = random(height);
>>>>     amp = random(1);
>>>>     cs.setChn("pitch", y);
>>>>     cs.setChn("pan", x/width);
>>>>     cs.setChn("amp", amp);
>>>>     fill(255, (x/width) * 255, 0, (y/height) * 255);
>>>>     ellipse(x,y, 50*amp, 50*amp);
>>>>     fill(0, 0, 0, 10);
>>>>     rect(0, 0, width, height);
>>>> }
>>>>
>>>>
>>>>
>>>> On 17 January 2013 23:08, Steven Yi  wrote:
>>>>> http://stackoverflow.com/questions/3310907/what-are-the-advantages-of-armv7-over-armv6-when-compiling-iphone-apps
>>>>>
>>>>> Seems like they took out hardware double precision support for armv7. I
>>>>> think on Android we compile w/Neon. Still, testing required to verify for
>>>>> both platforms.
>>>>>
>>>>> Rory: How is csoundo compiled to work with floats if it's used on the
>>>>> desktop? I'm a bit confused, but also on my cell phone. I'll be back online
>>>>> tomorrow when on campus and will be able to see what's going on better.
>>>>>
>>>>> On Jan 17, 2013 9:39 PM, "Rory Walsh"  wrote:
>>>>>>
>>>>>> That's true Mike. I gues we could be surprised by the results.
>>>>>>
>>>>>> On 17 January 2013 21:20, Michael Gogins  wrote:
>>>>>> > Don't make assumptions about the speed. Past tests have showed that on
>>>>>> > some platforms, doubles are faster than floats; on other platforms
>>>>>> > including Windows, floats are faster (15% or so) than doubles.
>>>>>> >
>>>>>> > Try both and measure it.
>>>>>> >
>>>>>> > Regards,
>>>>>> > Mike
>>>>>> >
>>>>>> > On Thu, Jan 17, 2013 at 4:05 PM, Rory Walsh  wrote:
>>>>>> >> I don't think having several android builds is a god idea either. And
>>>>>> >> there isn't really much pint I guess in investigate doubles, it's
>>>>>> >> probably not going to run as fast as float.
>>>>>> >>
>>>>>> >>> Another option would be on your side to create a specific Csoundo for
>>>>>> >>> Android that works with float build. I assume then that a user working
>>>>>> >>> with
>>>>>> >>
>>>>>> >> I'm not sure I follow. At present it is a float build. At present we
>>>>>> >> don't use any eclipse project files at all. The main IDE IS
>>>>>> >> Processing. We go from there straight to the phone. It's savage :)
>>>>>> >>
>>>>>> >> p.s. can you enlighten me as to why I don't get the same errors with
>>>>>> >> SetChannel()?
>>>>>> >>
>>>>>> >>
>>>>>> >> 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"
>>>>>> >>
>>>>>> >
>>>>>> >
>>>>>> >
>>>>>> > --
>>>>>> > Michael Gogins
>>>>>> > Irreducible Productions
>>>>>> > http://www.michael-gogins.com
>>>>>> > Michael dot Gogins at gmail dot com
>>>>>> >
>>>>>> >
>>>>>> > 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"
>>