Hi Jake,

I took a look and reproduced the odd timing results.  The printouts
show different results:

Move plate 1 from 1 to 3 at 0.100000
Move plate 2 from 1 to 2 at 0.200000
Move plate 1 from 3 to 2 at 0.300000
Move plate 3 from 1 to 3 at 0.400000
Move plate 1 from 2 to 1 at 0.500000
Move plate 2 from 2 to 3 at 0.600000
Move plate 1 from 1 to 3 at 0.700000
Move plate 4 from 1 to 2 at 0.800000
Move plate 1 from 3 to 2 at 0.900000
Move plate 2 from 3 to 1 at 1.000000

which is expected as the $DURATION is set to .1.  With CS6 though, the
output shows:

Move plate 0 from 1 to 2 at 0.000000
Move plate 1 from 1 to 3 at 0.100000
Move plate 0 from 2 to 3 at 0.100000
Move plate 2 from 1 to 2 at 0.300000
Move plate 0 from 3 to 1 at 0.300000
Move plate 1 from 3 to 2 at 0.400000
Move plate 0 from 1 to 2 at 0.400000
Move plate 3 from 1 to 3 at 0.700000
Move plate 0 from 2 to 3 at 0.700000
Move plate 1 from 2 to 1 at 0.800000
Move plate 0 from 3 to 1 at 0.800000

which has a plate 0 moving in between each of the ones that would show
up in CS5's output. With the UDO code, that printf with plate 0 should
never happen, as the if block explicitly checks for i_n != 0 before
going into the block with printf.

I took a look at the output from using "--verbose -I" and that lead to
me fix a couple of other bugs, but I'm still not sure what is going on
with this bug.  I'm a bit stumped at this point, but I get the bad
suspicion some memory is getting shared with the recursive UDO calls,
but it's just a hunch.  Perhaps others might see what the problem is.

Thanks,
steven

On Sat, Apr 13, 2013 at 4:24 PM, Jacob Joaquin  wrote:
> Hey Devs,
>
> I started rendering my personal backlog of Csound files with C6 and plan to
> post anything that isn't working as expected. I've spent a lot of time
> poking and prodding Csound and have collected a sizable library of tests, so
> hopefully this will be of some use.
>
> I found issue with this script, Sine Waves of Hanoi:
> https://gist.github.com/jacobjoaquin/5378797
>
> It plays back properly in C5, but not C6. The timing of the events sounds
> like it's quantized funny.
>
> Best,
> jake
>
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

I am looking at this too, but could this issue be to do with rates and how if-then-else deals with it? I noticed that if I put
a print i_n after the if i_n == 0, I never get a i_n = 0, which is to be expected. The thing is, event is k-rate and so is printf, so I am wondering whether
there is confusion going on there.

Victor
On 13 Apr 2013, at 23:45, Steven Yi wrote:

> Hi Jake,
> 
> I took a look and reproduced the odd timing results.  The printouts
> show different results:
> 
> Move plate 1 from 1 to 3 at 0.100000
> Move plate 2 from 1 to 2 at 0.200000
> Move plate 1 from 3 to 2 at 0.300000
> Move plate 3 from 1 to 3 at 0.400000
> Move plate 1 from 2 to 1 at 0.500000
> Move plate 2 from 2 to 3 at 0.600000
> Move plate 1 from 1 to 3 at 0.700000
> Move plate 4 from 1 to 2 at 0.800000
> Move plate 1 from 3 to 2 at 0.900000
> Move plate 2 from 3 to 1 at 1.000000
> 
> which is expected as the $DURATION is set to .1.  With CS6 though, the
> output shows:
> 
> Move plate 0 from 1 to 2 at 0.000000
> Move plate 1 from 1 to 3 at 0.100000
> Move plate 0 from 2 to 3 at 0.100000
> Move plate 2 from 1 to 2 at 0.300000
> Move plate 0 from 3 to 1 at 0.300000
> Move plate 1 from 3 to 2 at 0.400000
> Move plate 0 from 1 to 2 at 0.400000
> Move plate 3 from 1 to 3 at 0.700000
> Move plate 0 from 2 to 3 at 0.700000
> Move plate 1 from 2 to 1 at 0.800000
> Move plate 0 from 3 to 1 at 0.800000
> 
> which has a plate 0 moving in between each of the ones that would show
> up in CS5's output. With the UDO code, that printf with plate 0 should
> never happen, as the if block explicitly checks for i_n != 0 before
> going into the block with printf.
> 
> I took a look at the output from using "--verbose -I" and that lead to
> me fix a couple of other bugs, but I'm still not sure what is going on
> with this bug.  I'm a bit stumped at this point, but I get the bad
> suspicion some memory is getting shared with the recursive UDO calls,
> but it's just a hunch.  Perhaps others might see what the problem is.
> 
> Thanks,
> steven
> 
> On Sat, Apr 13, 2013 at 4:24 PM, Jacob Joaquin  wrote:
>> Hey Devs,
>> 
>> I started rendering my personal backlog of Csound files with C6 and plan to
>> post anything that isn't working as expected. I've spent a lot of time
>> poking and prodding Csound and have collected a sizable library of tests, so
>> hopefully this will be of some use.
>> 
>> I found issue with this script, Sine Waves of Hanoi:
>> https://gist.github.com/jacobjoaquin/5378797
>> 
>> It plays back properly in C5, but not C6. The timing of the events sounds
>> like it's quantized funny.
>> 
>> Best,
>> jake
>> 
>> ------------------------------------------------------------------------------
>> Precog is a next-generation analytics platform capable of advanced
>> analytics on semi-structured data. The platform includes APIs for building
>> apps and a phenomenal toolset for data science. Developers can use
>> our toolset for easy data analysis & visualization. Get a free account!
>> http://www2.precog.com/precogplatform/slashdotnewsletter
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>> 
> 
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel

Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie




------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

In fact, when we make printf printf_i, the printout inconsistencies disappear. I am sure this is a rate thing. I am wondering if this orc exploited some bug in the original code that has now been put right.
On 14 Apr 2013, at 00:07, Victor Lazzarini wrote:

> I am looking at this too, but could this issue be to do with rates and how if-then-else deals with it? I noticed that if I put
> a print i_n after the if i_n == 0, I never get a i_n = 0, which is to be expected. The thing is, event is k-rate and so is printf, so I am wondering whether
> there is confusion going on there.
> 
> Victor
> On 13 Apr 2013, at 23:45, Steven Yi wrote:
> 
>> Hi Jake,
>> 
>> I took a look and reproduced the odd timing results.  The printouts
>> show different results:
>> 
>> Move plate 1 from 1 to 3 at 0.100000
>> Move plate 2 from 1 to 2 at 0.200000
>> Move plate 1 from 3 to 2 at 0.300000
>> Move plate 3 from 1 to 3 at 0.400000
>> Move plate 1 from 2 to 1 at 0.500000
>> Move plate 2 from 2 to 3 at 0.600000
>> Move plate 1 from 1 to 3 at 0.700000
>> Move plate 4 from 1 to 2 at 0.800000
>> Move plate 1 from 3 to 2 at 0.900000
>> Move plate 2 from 3 to 1 at 1.000000
>> 
>> which is expected as the $DURATION is set to .1.  With CS6 though, the
>> output shows:
>> 
>> Move plate 0 from 1 to 2 at 0.000000
>> Move plate 1 from 1 to 3 at 0.100000
>> Move plate 0 from 2 to 3 at 0.100000
>> Move plate 2 from 1 to 2 at 0.300000
>> Move plate 0 from 3 to 1 at 0.300000
>> Move plate 1 from 3 to 2 at 0.400000
>> Move plate 0 from 1 to 2 at 0.400000
>> Move plate 3 from 1 to 3 at 0.700000
>> Move plate 0 from 2 to 3 at 0.700000
>> Move plate 1 from 2 to 1 at 0.800000
>> Move plate 0 from 3 to 1 at 0.800000
>> 
>> which has a plate 0 moving in between each of the ones that would show
>> up in CS5's output. With the UDO code, that printf with plate 0 should
>> never happen, as the if block explicitly checks for i_n != 0 before
>> going into the block with printf.
>> 
>> I took a look at the output from using "--verbose -I" and that lead to
>> me fix a couple of other bugs, but I'm still not sure what is going on
>> with this bug.  I'm a bit stumped at this point, but I get the bad
>> suspicion some memory is getting shared with the recursive UDO calls,
>> but it's just a hunch.  Perhaps others might see what the problem is.
>> 
>> Thanks,
>> steven
>> 
>> On Sat, Apr 13, 2013 at 4:24 PM, Jacob Joaquin  wrote:
>>> Hey Devs,
>>> 
>>> I started rendering my personal backlog of Csound files with C6 and plan to
>>> post anything that isn't working as expected. I've spent a lot of time
>>> poking and prodding Csound and have collected a sizable library of tests, so
>>> hopefully this will be of some use.
>>> 
>>> I found issue with this script, Sine Waves of Hanoi:
>>> https://gist.github.com/jacobjoaquin/5378797
>>> 
>>> It plays back properly in C5, but not C6. The timing of the events sounds
>>> like it's quantized funny.
>>> 
>>> Best,
>>> jake
>>> 
>>> ------------------------------------------------------------------------------
>>> Precog is a next-generation analytics platform capable of advanced
>>> analytics on semi-structured data. The platform includes APIs for building
>>> apps and a phenomenal toolset for data science. Developers can use
>>> our toolset for easy data analysis & visualization. Get a free account!
>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>> 
>> 
>> ------------------------------------------------------------------------------
>> Precog is a next-generation analytics platform capable of advanced
>> analytics on semi-structured data. The platform includes APIs for building
>> apps and a phenomenal toolset for data science. Developers can use
>> our toolset for easy data analysis & visualization. Get a free account!
>> http://www2.precog.com/precogplatform/slashdotnewsletter
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
> 
> Dr Victor Lazzarini
> Senior Lecturer
> Dept. of Music
> NUI Maynooth Ireland
> tel.: +353 1 708 3545
> Victor dot Lazzarini AT nuim dot ie
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel

Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie




------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

I am pretty sure it is to do with rates. If I add a printk2 to the 'else'  block:

opcode Hanoi, i, iiii
i_n, ia, ic, ib xin
if i_n != 0 then
ifoo Hanoi i_n - 1, ia, ib, ic
event "i", $Sine, gk_c, 1, i_n, ia, ic
printf "Move plate %i from %i to %i at %f\n", 1, i_n, ia, ic, gk_c
ifoo Hanoi i_n - 1, ib, ic, ia
else
printk2 gk_c
gk_c = gk_c + $DURATION
endif
xout 1
endop

I get this printout on csound 5:

new alloc for instr 1:
i1 0.00000
Move plate 1 from 1 to 3 at 0.100000
i1 0.10000
Move plate 2 from 1 to 2 at 0.200000
i1 0.20000
Move plate 1 from 3 to 2 at 0.300000
i1 0.30000
Move plate 3 from 1 to 3 at 0.400000
i1 0.40000
Move plate 1 from 2 to 1 at 0.500000
i1 0.50000
Move plate 2 from 2 to 3 at 0.600000
i1 0.60000
Move plate 1 from 1 to 3 at 0.700000
i1 0.70000
Move plate 4 from 1 to 2 at 0.800000
i1 0.80000
Move plate 1 from 3 to 2 at 0.900000
i1 0.90000

and this one on csound 6:

new alloc for instr 1:
Move plate 0 from 1 to 2 at 0.000000
 i1     0.00000
Move plate 1 from 1 to 3 at 0.100000
Move plate 0 from 2 to 3 at 0.100000
 i1     0.10000
 i1     0.20000
Move plate 2 from 1 to 2 at 0.300000
Move plate 0 from 3 to 1 at 0.300000
 i1     0.30000
Move plate 1 from 3 to 2 at 0.400000
Move plate 0 from 1 to 2 at 0.400000
 i1     0.40000
 i1     0.50000
 i1     0.60000
Move plate 3 from 1 to 3 at 0.700000
Move plate 0 from 2 to 3 at 0.700000
 i1     0.70000
Move plate 1 from 2 to 1 at 0.800000
Move plate 0 from 3 to 1 at 0.800000
 i1     0.80000
 i1     0.90000
Move plate 2 from 2 to 3 at 1.000000




On 14 Apr 2013, at 00:12, Victor Lazzarini wrote:

> In fact, when we make printf printf_i, the printout inconsistencies disappear. I am sure this is a rate thing. I am wondering if this orc exploited some bug in the original code that has now been put right.
> On 14 Apr 2013, at 00:07, Victor Lazzarini wrote:
> 
>> I am looking at this too, but could this issue be to do with rates and how if-then-else deals with it? I noticed that if I put
>> a print i_n after the if i_n == 0, I never get a i_n = 0, which is to be expected. The thing is, event is k-rate and so is printf, so I am wondering whether
>> there is confusion going on there.
>> 
>> Victor
>> On 13 Apr 2013, at 23:45, Steven Yi wrote:
>> 
>>> Hi Jake,
>>> 
>>> I took a look and reproduced the odd timing results.  The printouts
>>> show different results:
>>> 
>>> Move plate 1 from 1 to 3 at 0.100000
>>> Move plate 2 from 1 to 2 at 0.200000
>>> Move plate 1 from 3 to 2 at 0.300000
>>> Move plate 3 from 1 to 3 at 0.400000
>>> Move plate 1 from 2 to 1 at 0.500000
>>> Move plate 2 from 2 to 3 at 0.600000
>>> Move plate 1 from 1 to 3 at 0.700000
>>> Move plate 4 from 1 to 2 at 0.800000
>>> Move plate 1 from 3 to 2 at 0.900000
>>> Move plate 2 from 3 to 1 at 1.000000
>>> 
>>> which is expected as the $DURATION is set to .1.  With CS6 though, the
>>> output shows:
>>> 
>>> Move plate 0 from 1 to 2 at 0.000000
>>> Move plate 1 from 1 to 3 at 0.100000
>>> Move plate 0 from 2 to 3 at 0.100000
>>> Move plate 2 from 1 to 2 at 0.300000
>>> Move plate 0 from 3 to 1 at 0.300000
>>> Move plate 1 from 3 to 2 at 0.400000
>>> Move plate 0 from 1 to 2 at 0.400000
>>> Move plate 3 from 1 to 3 at 0.700000
>>> Move plate 0 from 2 to 3 at 0.700000
>>> Move plate 1 from 2 to 1 at 0.800000
>>> Move plate 0 from 3 to 1 at 0.800000
>>> 
>>> which has a plate 0 moving in between each of the ones that would show
>>> up in CS5's output. With the UDO code, that printf with plate 0 should
>>> never happen, as the if block explicitly checks for i_n != 0 before
>>> going into the block with printf.
>>> 
>>> I took a look at the output from using "--verbose -I" and that lead to
>>> me fix a couple of other bugs, but I'm still not sure what is going on
>>> with this bug.  I'm a bit stumped at this point, but I get the bad
>>> suspicion some memory is getting shared with the recursive UDO calls,
>>> but it's just a hunch.  Perhaps others might see what the problem is.
>>> 
>>> Thanks,
>>> steven
>>> 
>>> On Sat, Apr 13, 2013 at 4:24 PM, Jacob Joaquin  wrote:
>>>> Hey Devs,
>>>> 
>>>> I started rendering my personal backlog of Csound files with C6 and plan to
>>>> post anything that isn't working as expected. I've spent a lot of time
>>>> poking and prodding Csound and have collected a sizable library of tests, so
>>>> hopefully this will be of some use.
>>>> 
>>>> I found issue with this script, Sine Waves of Hanoi:
>>>> https://gist.github.com/jacobjoaquin/5378797
>>>> 
>>>> It plays back properly in C5, but not C6. The timing of the events sounds
>>>> like it's quantized funny.
>>>> 
>>>> Best,
>>>> jake
>>>> 
>>>> ------------------------------------------------------------------------------
>>>> Precog is a next-generation analytics platform capable of advanced
>>>> analytics on semi-structured data. The platform includes APIs for building
>>>> apps and a phenomenal toolset for data science. Developers can use
>>>> our toolset for easy data analysis & visualization. Get a free account!
>>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>>> _______________________________________________
>>>> Csound-devel mailing list
>>>> Csound-devel@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>> 
>>> 
>>> ------------------------------------------------------------------------------
>>> Precog is a next-generation analytics platform capable of advanced
>>> analytics on semi-structured data. The platform includes APIs for building
>>> apps and a phenomenal toolset for data science. Developers can use
>>> our toolset for easy data analysis & visualization. Get a free account!
>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>> 
>> Dr Victor Lazzarini
>> Senior Lecturer
>> Dept. of Music
>> NUI Maynooth Ireland
>> tel.: +353 1 708 3545
>> Victor dot Lazzarini AT nuim dot ie
>> 
>> 
>> 
>> 
>> ------------------------------------------------------------------------------
>> Precog is a next-generation analytics platform capable of advanced
>> analytics on semi-structured data. The platform includes APIs for building
>> apps and a phenomenal toolset for data science. Developers can use
>> our toolset for easy data analysis & visualization. Get a free account!
>> http://www2.precog.com/precogplatform/slashdotnewsletter
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
> 
> Dr Victor Lazzarini
> Senior Lecturer
> Dept. of Music
> NUI Maynooth Ireland
> tel.: +353 1 708 3545
> Victor dot Lazzarini AT nuim dot ie
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel

Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie




------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Hi Victor,

I think you may have not pulled from GIT before testing as I had
pushed a fix last night for the goto token to get the correct rate
(goto vs. kgoto). I still get the wrong timing here after that fix.
Just woke up here but am looking at it now (I'll be on IRC if you want
to discuss).

Thanks!
steven

On Sun, Apr 14, 2013 at 9:21 AM, Victor Lazzarini
 wrote:
> Maybe this explains it:
>
> 
>
> Isn't that the wrong opcode?
>
> Victor
> On 14 Apr 2013, at 01:26, Victor Lazzarini wrote:
>
> I think I got to the bottom of this now. The issue seems to boil down to
> this: if-else code like this
>
> instr 1
> i1 = 0
> if i1 == 0 then
> print i1
> else
> print i1+1
> endif
> endin
>
> in csound5:
>
> new alloc for instr 1:
> instr 1: i1 = 0.000
> B 0.000 .. 1.000 T 1.000 TT 1.000 M: 0.0 0.0
>
> in csound6:
>
> new alloc for instr 1:
> instr 1:  i1 = 0.000
> instr 1:  #i0 = 1.000
> B  0.000 ..  1.000 T  1.000 TT  1.000 M:      0.0      0.0
>
>
> There is a fall-through that is not there in Csound 5. I modified the hanoi
> UDO to use init-time code only (event_i and printf_i), but
> the code still did not work as in Csound 5. So I looked closer and found
> this difference.
>
>
> Dr Victor Lazzarini
> Senior Lecturer
> Dept. of Music
> NUI Maynooth Ireland
> tel.: +353 1 708 3545
> Victor dot Lazzarini AT nuim dot ie
>
>
>
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter_______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> Dr Victor Lazzarini
> Senior Lecturer
> Dept. of Music
> NUI Maynooth Ireland
> tel.: +353 1 708 3545
> Victor dot Lazzarini AT nuim dot ie
>
>
>
>
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Just some more analysis, the output from cs5 and cs6 witih --verbose
is listed below.  There's a difference in output between the two in
terms of how it's reported, but you can compare to see that the
opcodes that are found and used are the same.

CS5:

=.r argndxs: 13 2
=.r argndxs: 14 2
=.r argndxs: 15 3
=.r argndxs: 16 4
=.r argndxs: 18 3
init.k argndxs: 22 1
xin argndxs: -4 -5 -6 -7
!= argndxs: -8 -4 1
cngoto argndxs: -8 ***lbl
sub.ii argndxs: -9 -4 3
Hanoi argndxs: -10 -9 -5 -7 -6 1
event argndxs: 2013265922 4 22 3 -4 -5 -6
printf argndxs: 2013265923 3 -4 -5 -6 22
sub.ii argndxs: -11 -4 3
Hanoi argndxs: -10 -11 -7 -6 -5 1
goto argndxs: ***lbl
add.kk argndxs: 22 22 5
xout argndxs: 3
=.i argndxs: -5 -4
Hanoi argndxs: -6 -5 3 4 6 1
turnoff argndxs:
=.i argndxs: -3 5
=.i argndxs: -7 -3
=.i argndxs: -8 -4
sub.ii argndxs: -9 -5 3
div.ii argndxs: -10 -9 4
sub.ii argndxs: -11 -6 3
div.ii argndxs: -12 -11 4
sub.ii argndxs: -13 -7 7
linseg argndxs: -16 1 8 3 -13 3 8 1
mul.ka argndxs: -17 9 -16
sub.ii argndxs: -14 10 -8
mul.ii argndxs: -15 11 -14
oscil.ak argndxs: -18 -17 -15 3 12
line argndxs: -19 -10 -7 -12
sqrt.a argndxs: -20 -19
mul.aa argndxs: -21 -18 -20
sub.ka argndxs: -22 3 -19
sqrt.a argndxs: -23 -22
mul.aa argndxs: -24 -18 -23
outs argndxs: -21 -24

CS6:

=.r args: sr 44100
=.r args: kr 44100
=.r args: ksmps 1
=.r args: nchnls 2
=.r args: 0dbfs 1
init.k args: gk_c 0
xin args: i_n ia ic ib
!= args: #b0 i_n 0
cngoto args: #b0 __synthetic_301:
##sub.ii args: #i0 i_n 1
Hanoi args: ifoo #i0 ia ib ic 0
event args: "i" 2 gk_c 1 i_n ia ic
printf args: "Move plate %i from %i to %i at %f\n" 1 i_n ia ic gk_c
##sub.ii args: #i1 i_n 1
Hanoi args: ifoo #i1 ib ic ia 0
goto args: __synthetic_300:
##add.kk args: #k0 gk_c 0.1
=.k args: gk_c #k0
xout args: 1
=.i args: idiscs p4
Hanoi args: ifoo idiscs 1 2 3 0
turnoff args:
=.i args: p3 0.1
=.i args: idur p3
=.i args: i_n p4
##sub.ii args: #i2 p5 1
##div.ii args: #i3 #i2 2
=.i args: ia #i3
##sub.ii args: #i4 p6 1
##div.ii args: #i5 #i4 2
=.i args: ic #i5
##sub.ii args: #i6 idur 0.02
linseg args: aenv 0 0.01 1 #i6 1 0.01 0
##mul.ka args: #a0 0.707 aenv
##sub.ii args: #i7 10 i_n
##mul.ii args: #i8 440 #i7
##mul.ii args: #i9 -1 1
oscil.ak args: a1 #a0 #i8 1 #i9
line args: a2 ia idur ic
sqrt.a args: #a1 a2
##mul.aa args: #a2 a1 #a1
##sub.ka args: #a3 1 a2
sqrt.a args: #a4 #a3
##mul.aa args: #a5 a1 #a4
outs args: #a2 #a5


On Sun, Apr 14, 2013 at 10:18 AM, Steven Yi  wrote:
> Hi Victor,
>
> I think you may have not pulled from GIT before testing as I had
> pushed a fix last night for the goto token to get the correct rate
> (goto vs. kgoto). I still get the wrong timing here after that fix.
> Just woke up here but am looking at it now (I'll be on IRC if you want
> to discuss).
>
> Thanks!
> steven
>
> On Sun, Apr 14, 2013 at 9:21 AM, Victor Lazzarini
>  wrote:
>> Maybe this explains it:
>>
>> 
>>
>> Isn't that the wrong opcode?
>>
>> Victor
>> On 14 Apr 2013, at 01:26, Victor Lazzarini wrote:
>>
>> I think I got to the bottom of this now. The issue seems to boil down to
>> this: if-else code like this
>>
>> instr 1
>> i1 = 0
>> if i1 == 0 then
>> print i1
>> else
>> print i1+1
>> endif
>> endin
>>
>> in csound5:
>>
>> new alloc for instr 1:
>> instr 1: i1 = 0.000
>> B 0.000 .. 1.000 T 1.000 TT 1.000 M: 0.0 0.0
>>
>> in csound6:
>>
>> new alloc for instr 1:
>> instr 1:  i1 = 0.000
>> instr 1:  #i0 = 1.000
>> B  0.000 ..  1.000 T  1.000 TT  1.000 M:      0.0      0.0
>>
>>
>> There is a fall-through that is not there in Csound 5. I modified the hanoi
>> UDO to use init-time code only (event_i and printf_i), but
>> the code still did not work as in Csound 5. So I looked closer and found
>> this difference.
>>
>>
>> Dr Victor Lazzarini
>> Senior Lecturer
>> Dept. of Music
>> NUI Maynooth Ireland
>> tel.: +353 1 708 3545
>> Victor dot Lazzarini AT nuim dot ie
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Precog is a next-generation analytics platform capable of advanced
>> analytics on semi-structured data. The platform includes APIs for building
>> apps and a phenomenal toolset for data science. Developers can use
>> our toolset for easy data analysis & visualization. Get a free account!
>> http://www2.precog.com/precogplatform/slashdotnewsletter_______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>>
>> Dr Victor Lazzarini
>> Senior Lecturer
>> Dept. of Music
>> NUI Maynooth Ireland
>> tel.: +353 1 708 3545
>> Victor dot Lazzarini AT nuim dot ie
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Precog is a next-generation analytics platform capable of advanced
>> analytics on semi-structured data. The platform includes APIs for building
>> apps and a phenomenal toolset for data science. Developers can use
>> our toolset for easy data analysis & visualization. Get a free account!
>> http://www2.precog.com/precogplatform/slashdotnewsletter
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

OK, with your fixes, this code below works in the same way in Csound 5 and Csound 6, as I would expect. The original code, with its mixed rates is
harder to debug, but I think it is nothing to do with memory issues in recursive UDOs, but to do the the control-of-flow behaviour (which is different).

Sine Waves of Hanoi
Jacob Joaquin
June 3, 2010
jacobjoaquin@gmail.com
csoundblog.com


sr = 44100
kr = 44100
ksmps = 1
nchnls = 2
0dbfs = 1
#define Hanoi # 1 #
#define Sine # 2 #
#define DURATION # 0.1 #
#define DISCS # 10 #
gi_c init 0
opcode Hanoi, i, iiii
i_n, ia, ic, ib xin

if i_n != 0 then
ifoo Hanoi i_n - 1, ia, ib, ic
event_i "i", $Sine, gi_c, 1, i_n, ia, ic
ifoo Hanoi i_n - 1, ib, ic, ia
else
gi_c = gi_c + $DURATION
endif

xout 1
endop
instr $Hanoi
idiscs = p4
ifoo Hanoi idiscs, 1, 2, 3
turnoff
endin
instr $Sine
p3 = $DURATION
idur = p3
i_n = p4
ia = (p5 - 1) / 2
ic = (p6 - 1) / 2
aenv linseg 0, 0.01, 1, idur - 0.02, 1, 0.01, 0
a1 oscil 0.707 * aenv, 440 * ($DISCS - i_n), 1, -1
a2 line ia, idur, ic
outs a1 * sqrt(a2), a1 * sqrt(1 - a2)
endin


#define Hanoi # 1 #
#define Sine # 2 #
#define DURATION # 0.1 #
#define DISCS # 10 #
f 1 0 [2 ^ 16] 10 1
i $Hanoi 0 1 $DISCS
e [(2 ^ $DISCS - 1) * $DURATION]



On 14 Apr 2013, at 10:40, Steven Yi wrote:

> Just some more analysis, the output from cs5 and cs6 witih --verbose
> is listed below.  There's a difference in output between the two in
> terms of how it's reported, but you can compare to see that the
> opcodes that are found and used are the same.
> 
> CS5:
> 
> =.r argndxs: 13 2
> =.r argndxs: 14 2
> =.r argndxs: 15 3
> =.r argndxs: 16 4
> =.r argndxs: 18 3
> init.k argndxs: 22 1
> xin argndxs: -4 -5 -6 -7
> != argndxs: -8 -4 1
> cngoto argndxs: -8 ***lbl
> sub.ii argndxs: -9 -4 3
> Hanoi argndxs: -10 -9 -5 -7 -6 1
> event argndxs: 2013265922 4 22 3 -4 -5 -6
> printf argndxs: 2013265923 3 -4 -5 -6 22
> sub.ii argndxs: -11 -4 3
> Hanoi argndxs: -10 -11 -7 -6 -5 1
> goto argndxs: ***lbl
> add.kk argndxs: 22 22 5
> xout argndxs: 3
> =.i argndxs: -5 -4
> Hanoi argndxs: -6 -5 3 4 6 1
> turnoff argndxs:
> =.i argndxs: -3 5
> =.i argndxs: -7 -3
> =.i argndxs: -8 -4
> sub.ii argndxs: -9 -5 3
> div.ii argndxs: -10 -9 4
> sub.ii argndxs: -11 -6 3
> div.ii argndxs: -12 -11 4
> sub.ii argndxs: -13 -7 7
> linseg argndxs: -16 1 8 3 -13 3 8 1
> mul.ka argndxs: -17 9 -16
> sub.ii argndxs: -14 10 -8
> mul.ii argndxs: -15 11 -14
> oscil.ak argndxs: -18 -17 -15 3 12
> line argndxs: -19 -10 -7 -12
> sqrt.a argndxs: -20 -19
> mul.aa argndxs: -21 -18 -20
> sub.ka argndxs: -22 3 -19
> sqrt.a argndxs: -23 -22
> mul.aa argndxs: -24 -18 -23
> outs argndxs: -21 -24
> 
> CS6:
> 
> =.r args: sr 44100
> =.r args: kr 44100
> =.r args: ksmps 1
> =.r args: nchnls 2
> =.r args: 0dbfs 1
> init.k args: gk_c 0
> xin args: i_n ia ic ib
> != args: #b0 i_n 0
> cngoto args: #b0 __synthetic_301:
> ##sub.ii args: #i0 i_n 1
> Hanoi args: ifoo #i0 ia ib ic 0
> event args: "i" 2 gk_c 1 i_n ia ic
> printf args: "Move plate %i from %i to %i at %f\n" 1 i_n ia ic gk_c
> ##sub.ii args: #i1 i_n 1
> Hanoi args: ifoo #i1 ib ic ia 0
> goto args: __synthetic_300:
> ##add.kk args: #k0 gk_c 0.1
> =.k args: gk_c #k0
> xout args: 1
> =.i args: idiscs p4
> Hanoi args: ifoo idiscs 1 2 3 0
> turnoff args:
> =.i args: p3 0.1
> =.i args: idur p3
> =.i args: i_n p4
> ##sub.ii args: #i2 p5 1
> ##div.ii args: #i3 #i2 2
> =.i args: ia #i3
> ##sub.ii args: #i4 p6 1
> ##div.ii args: #i5 #i4 2
> =.i args: ic #i5
> ##sub.ii args: #i6 idur 0.02
> linseg args: aenv 0 0.01 1 #i6 1 0.01 0
> ##mul.ka args: #a0 0.707 aenv
> ##sub.ii args: #i7 10 i_n
> ##mul.ii args: #i8 440 #i7
> ##mul.ii args: #i9 -1 1
> oscil.ak args: a1 #a0 #i8 1 #i9
> line args: a2 ia idur ic
> sqrt.a args: #a1 a2
> ##mul.aa args: #a2 a1 #a1
> ##sub.ka args: #a3 1 a2
> sqrt.a args: #a4 #a3
> ##mul.aa args: #a5 a1 #a4
> outs args: #a2 #a5
> 
> 
> On Sun, Apr 14, 2013 at 10:18 AM, Steven Yi  wrote:
>> Hi Victor,
>> 
>> I think you may have not pulled from GIT before testing as I had
>> pushed a fix last night for the goto token to get the correct rate
>> (goto vs. kgoto). I still get the wrong timing here after that fix.
>> Just woke up here but am looking at it now (I'll be on IRC if you want
>> to discuss).
>> 
>> Thanks!
>> steven
>> 
>> On Sun, Apr 14, 2013 at 9:21 AM, Victor Lazzarini
>>  wrote:
>>> Maybe this explains it:
>>> 
>>> 
>>> 
>>> Isn't that the wrong opcode?
>>> 
>>> Victor
>>> On 14 Apr 2013, at 01:26, Victor Lazzarini wrote:
>>> 
>>> I think I got to the bottom of this now. The issue seems to boil down to
>>> this: if-else code like this
>>> 
>>> instr 1
>>> i1 = 0
>>> if i1 == 0 then
>>> print i1
>>> else
>>> print i1+1
>>> endif
>>> endin
>>> 
>>> in csound5:
>>> 
>>> new alloc for instr 1:
>>> instr 1: i1 = 0.000
>>> B 0.000 .. 1.000 T 1.000 TT 1.000 M: 0.0 0.0
>>> 
>>> in csound6:
>>> 
>>> new alloc for instr 1:
>>> instr 1:  i1 = 0.000
>>> instr 1:  #i0 = 1.000
>>> B  0.000 ..  1.000 T  1.000 TT  1.000 M:      0.0      0.0
>>> 
>>> 
>>> There is a fall-through that is not there in Csound 5. I modified the hanoi
>>> UDO to use init-time code only (event_i and printf_i), but
>>> the code still did not work as in Csound 5. So I looked closer and found
>>> this difference.
>>> 
>>> 
>>> Dr Victor Lazzarini
>>> Senior Lecturer
>>> Dept. of Music
>>> NUI Maynooth Ireland
>>> tel.: +353 1 708 3545
>>> Victor dot Lazzarini AT nuim dot ie
>>> 
>>> 
>>> 
>>> ------------------------------------------------------------------------------
>>> Precog is a next-generation analytics platform capable of advanced
>>> analytics on semi-structured data. The platform includes APIs for building
>>> apps and a phenomenal toolset for data science. Developers can use
>>> our toolset for easy data analysis & visualization. Get a free account!
>>> http://www2.precog.com/precogplatform/slashdotnewsletter_______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>> 
>>> 
>>> Dr Victor Lazzarini
>>> Senior Lecturer
>>> Dept. of Music
>>> NUI Maynooth Ireland
>>> tel.: +353 1 708 3545
>>> Victor dot Lazzarini AT nuim dot ie
>>> 
>>> 
>>> 
>>> 
>>> ------------------------------------------------------------------------------
>>> Precog is a next-generation analytics platform capable of advanced
>>> analytics on semi-structured data. The platform includes APIs for building
>>> apps and a phenomenal toolset for data science. Developers can use
>>> our toolset for easy data analysis & visualization. Get a free account!
>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>> 
> 
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel

Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie




------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Hi Jake and All,

I was on IRC with John and Victor a while and we all worked through
this bug.  We determined the root cause having to do with a change
with how the performance functions are iterated through.  It caused a
bug in the case of recursive UDO's and how goto's would change the
next function to call.  It was also the root cause of recursive UDO's
recursing one extra time, that was  patched around previously.

The Hanoi Test now works.  It is likely that other projects using
recursive UDO's might have had issues too, which should all be fixed
up now.

Thanks!
steven


On Sun, Apr 14, 2013 at 12:19 AM, Victor Lazzarini
 wrote:
> I am pretty sure it is to do with rates. If I add a printk2 to the 'else'  block:
>
> opcode Hanoi, i, iiii
> i_n, ia, ic, ib xin
> if i_n != 0 then
> ifoo Hanoi i_n - 1, ia, ib, ic
> event "i", $Sine, gk_c, 1, i_n, ia, ic
> printf "Move plate %i from %i to %i at %f\n", 1, i_n, ia, ic, gk_c
> ifoo Hanoi i_n - 1, ib, ic, ia
> else
> printk2 gk_c
> gk_c = gk_c + $DURATION
> endif
> xout 1
> endop
>
> I get this printout on csound 5:
>
> new alloc for instr 1:
> i1 0.00000
> Move plate 1 from 1 to 3 at 0.100000
> i1 0.10000
> Move plate 2 from 1 to 2 at 0.200000
> i1 0.20000
> Move plate 1 from 3 to 2 at 0.300000
> i1 0.30000
> Move plate 3 from 1 to 3 at 0.400000
> i1 0.40000
> Move plate 1 from 2 to 1 at 0.500000
> i1 0.50000
> Move plate 2 from 2 to 3 at 0.600000
> i1 0.60000
> Move plate 1 from 1 to 3 at 0.700000
> i1 0.70000
> Move plate 4 from 1 to 2 at 0.800000
> i1 0.80000
> Move plate 1 from 3 to 2 at 0.900000
> i1 0.90000
>
> and this one on csound 6:
>
> new alloc for instr 1:
> Move plate 0 from 1 to 2 at 0.000000
>  i1     0.00000
> Move plate 1 from 1 to 3 at 0.100000
> Move plate 0 from 2 to 3 at 0.100000
>  i1     0.10000
>  i1     0.20000
> Move plate 2 from 1 to 2 at 0.300000
> Move plate 0 from 3 to 1 at 0.300000
>  i1     0.30000
> Move plate 1 from 3 to 2 at 0.400000
> Move plate 0 from 1 to 2 at 0.400000
>  i1     0.40000
>  i1     0.50000
>  i1     0.60000
> Move plate 3 from 1 to 3 at 0.700000
> Move plate 0 from 2 to 3 at 0.700000
>  i1     0.70000
> Move plate 1 from 2 to 1 at 0.800000
> Move plate 0 from 3 to 1 at 0.800000
>  i1     0.80000
>  i1     0.90000
> Move plate 2 from 2 to 3 at 1.000000
>
>
>
>
> On 14 Apr 2013, at 00:12, Victor Lazzarini wrote:
>
>> In fact, when we make printf printf_i, the printout inconsistencies disappear. I am sure this is a rate thing. I am wondering if this orc exploited some bug in the original code that has now been put right.
>> On 14 Apr 2013, at 00:07, Victor Lazzarini wrote:
>>
>>> I am looking at this too, but could this issue be to do with rates and how if-then-else deals with it? I noticed that if I put
>>> a print i_n after the if i_n == 0, I never get a i_n = 0, which is to be expected. The thing is, event is k-rate and so is printf, so I am wondering whether
>>> there is confusion going on there.
>>>
>>> Victor
>>> On 13 Apr 2013, at 23:45, Steven Yi wrote:
>>>
>>>> Hi Jake,
>>>>
>>>> I took a look and reproduced the odd timing results.  The printouts
>>>> show different results:
>>>>
>>>> Move plate 1 from 1 to 3 at 0.100000
>>>> Move plate 2 from 1 to 2 at 0.200000
>>>> Move plate 1 from 3 to 2 at 0.300000
>>>> Move plate 3 from 1 to 3 at 0.400000
>>>> Move plate 1 from 2 to 1 at 0.500000
>>>> Move plate 2 from 2 to 3 at 0.600000
>>>> Move plate 1 from 1 to 3 at 0.700000
>>>> Move plate 4 from 1 to 2 at 0.800000
>>>> Move plate 1 from 3 to 2 at 0.900000
>>>> Move plate 2 from 3 to 1 at 1.000000
>>>>
>>>> which is expected as the $DURATION is set to .1.  With CS6 though, the
>>>> output shows:
>>>>
>>>> Move plate 0 from 1 to 2 at 0.000000
>>>> Move plate 1 from 1 to 3 at 0.100000
>>>> Move plate 0 from 2 to 3 at 0.100000
>>>> Move plate 2 from 1 to 2 at 0.300000
>>>> Move plate 0 from 3 to 1 at 0.300000
>>>> Move plate 1 from 3 to 2 at 0.400000
>>>> Move plate 0 from 1 to 2 at 0.400000
>>>> Move plate 3 from 1 to 3 at 0.700000
>>>> Move plate 0 from 2 to 3 at 0.700000
>>>> Move plate 1 from 2 to 1 at 0.800000
>>>> Move plate 0 from 3 to 1 at 0.800000
>>>>
>>>> which has a plate 0 moving in between each of the ones that would show
>>>> up in CS5's output. With the UDO code, that printf with plate 0 should
>>>> never happen, as the if block explicitly checks for i_n != 0 before
>>>> going into the block with printf.
>>>>
>>>> I took a look at the output from using "--verbose -I" and that lead to
>>>> me fix a couple of other bugs, but I'm still not sure what is going on
>>>> with this bug.  I'm a bit stumped at this point, but I get the bad
>>>> suspicion some memory is getting shared with the recursive UDO calls,
>>>> but it's just a hunch.  Perhaps others might see what the problem is.
>>>>
>>>> Thanks,
>>>> steven
>>>>
>>>> On Sat, Apr 13, 2013 at 4:24 PM, Jacob Joaquin  wrote:
>>>>> Hey Devs,
>>>>>
>>>>> I started rendering my personal backlog of Csound files with C6 and plan to
>>>>> post anything that isn't working as expected. I've spent a lot of time
>>>>> poking and prodding Csound and have collected a sizable library of tests, so
>>>>> hopefully this will be of some use.
>>>>>
>>>>> I found issue with this script, Sine Waves of Hanoi:
>>>>> https://gist.github.com/jacobjoaquin/5378797
>>>>>
>>>>> It plays back properly in C5, but not C6. The timing of the events sounds
>>>>> like it's quantized funny.
>>>>>
>>>>> Best,
>>>>> jake
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Precog is a next-generation analytics platform capable of advanced
>>>>> analytics on semi-structured data. The platform includes APIs for building
>>>>> apps and a phenomenal toolset for data science. Developers can use
>>>>> our toolset for easy data analysis & visualization. Get a free account!
>>>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>>>> _______________________________________________
>>>>> Csound-devel mailing list
>>>>> Csound-devel@lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Precog is a next-generation analytics platform capable of advanced
>>>> analytics on semi-structured data. The platform includes APIs for building
>>>> apps and a phenomenal toolset for data science. Developers can use
>>>> our toolset for easy data analysis & visualization. Get a free account!
>>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>>> _______________________________________________
>>>> Csound-devel mailing list
>>>> Csound-devel@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>> Dr Victor Lazzarini
>>> Senior Lecturer
>>> Dept. of Music
>>> NUI Maynooth Ireland
>>> tel.: +353 1 708 3545
>>> Victor dot Lazzarini AT nuim dot ie
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Precog is a next-generation analytics platform capable of advanced
>>> analytics on semi-structured data. The platform includes APIs for building
>>> apps and a phenomenal toolset for data science. Developers can use
>>> our toolset for easy data analysis & visualization. Get a free account!
>>> http://www2.precog.com/precogplatform/slashdotnewsletter
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>> Dr Victor Lazzarini
>> Senior Lecturer
>> Dept. of Music
>> NUI Maynooth Ireland
>> tel.: +353 1 708 3545
>> Victor dot Lazzarini AT nuim dot ie
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Precog is a next-generation analytics platform capable of advanced
>> analytics on semi-structured data. The platform includes APIs for building
>> apps and a phenomenal toolset for data science. Developers can use
>> our toolset for easy data analysis & visualization. Get a free account!
>> http://www2.precog.com/precogplatform/slashdotnewsletter
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
> Dr Victor Lazzarini
> Senior Lecturer
> Dept. of Music
> NUI Maynooth Ireland
> tel.: +353 1 708 3545
> Victor dot Lazzarini AT nuim dot ie
>
>
>
>
> ------------------------------------------------------------------------------
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net