From anthonykozar at sbcglobal.net  Mon Oct  3 13:14:54 2005
From: anthonykozar at sbcglobal.net (Anthony Kozar)
Date: Mon, 03 Oct 2005 16:14:54 -0400
Subject: [CM] Trouble using tunings
Message-ID: <BF67097D.11FD5%anthonykozar@sbcglobal.net>

Hello, 

I am having trouble getting tunings to work with note names containing
octave specifiers.  I am using CM 2.3.4 in MCL 4.3.1.

After loading the example file scales.cm, I tried the following statements:

? (hertz 'a4)
440.0
? (hertz 'a4 :in #!meantone)
> Error: a4 is not a note in #<tuning: "meantone">.
> While executing: #<standard-method tuning-note->hertz (tuning t t)>
> Type Command-. to abort.
See the Restarts? menu item for further choices.
1 > 
Aborted
? (hertz 'a :in #!meantone)
218.48034899614788


As you can see, the hertz function works fine with the standard scale or
with the meantone scale without the octave specifier.

It seems to me that this used to work because I have an audio file rendered
with Csound using the Just Intonation scale from Common Music (the same
version of CM too!).  I just do not have the exact CM code that I used to
create it anymore.

Thanks in advance for any help!

Anthony Kozar
anthonykozar AT sbcglobal DOT net
http://akozar.spymac.net/



From taube at uiuc.edu  Mon Oct  3 16:44:03 2005
From: taube at uiuc.edu (Rick Taube)
Date: Mon, 3 Oct 2005 18:44:03 -0500
Subject: [CM] Trouble using tunings
In-Reply-To: <BF67097D.11FD5%anthonykozar@sbcglobal.net>
References: <BF67097D.11FD5%anthonykozar@sbcglobal.net>
Message-ID: <f8badd62bc5ed9a150c2666b734fe1da@uiuc.edu>

Hi -- I dont have MCL anymore so if I can't build 2.3.4 in openmcl im 
not sure if ill be able to debug it.
FWIW it does work in 2.7.0:

? (new tuning :name 'meantone :octaves '(-1 10)
       :cents '((0 c) (86 cs) (192 d) (312 ds) (384 e) (504 f)(578 fs)
                (696 g) (771 gs) (888 a) (1008 as) (1082 b)
                1200))
#<tuning "meantone">
? (hertz 'a4 :in #&meantone)
436.96066
?



On Oct 3, 2005, at 3:14 PM, Anthony Kozar wrote:

> Hello,
>
> I am having trouble getting tunings to work with note names containing
> octave specifiers.  I am using CM 2.3.4 in MCL 4.3.1.
>
> After loading the example file scales.cm, I tried the following 
> statements:
>
> ? (hertz 'a4)
> 440.0
> ? (hertz 'a4 :in #!meantone)
>> Error: a4 is not a note in #<tuning: "meantone">.
>> While executing: #<standard-method tuning-note->hertz (tuning t t)>
>> Type Command-. to abort.
> See the Restarts? menu item for further choices.
> 1 >
> Aborted
> ? (hertz 'a :in #!meantone)
> 218.48034899614788
>
>
> As you can see, the hertz function works fine with the standard scale 
> or
> with the meantone scale without the octave specifier.
>
> It seems to me that this used to work because I have an audio file 
> rendered
> with Csound using the Just Intonation scale from Common Music (the same
> version of CM too!).  I just do not have the exact CM code that I used 
> to
> create it anymore.
>
> Thanks in advance for any help!
>
> Anthony Kozar
> anthonykozar AT sbcglobal DOT net
> http://akozar.spymac.net/
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From anthonykozar at sbcglobal.net  Mon Oct  3 23:42:14 2005
From: anthonykozar at sbcglobal.net (Anthony Kozar)
Date: Tue, 04 Oct 2005 02:42:14 -0400
Subject: [CM] Trouble using tunings
In-Reply-To: <f8badd62bc5ed9a150c2666b734fe1da@uiuc.edu>
Message-ID: <BF679C86.11FE9%anthonykozar@sbcglobal.net>

Thanks for looking into this Rick.  I suppose that I should have updated to
CM 2.4.x by now, but I got CM/CLM/CMN all into the same image several years
ago and have just stuck with it.

Looking into the problem some more myself, I have traced the error to
retrieving the note name from the scale's internal hash table.  For some
reason, (gethash 'a4 (scale-table #!meantone)) returns nil nil, even though
running the following command clearly shows a key a4 in the table:

(maphash #'(lambda (k v) (format t "~A = ~A~%" k v))
         (scale-table #!meantone))

result:
110 = (#S(sd :note d8 :keynum 110 :octave 9 :class 2 :name d :accidental
nil))
r = (#S(sd :note r :keynum -1 :octave nil :class nil :name nil :accidental
nil))
62 = (#S(sd :note d4 :keynum 62 :octave 5 :class 2 :name d :accidental nil))
[.....]
a4 = (#S(sd :note a4 :keynum 69 :octave 5 :class 9 :name a :accidental nil))
[.....]

Interestingly, (gethash 'r (scale-table #!meantone)) does work, but none of
the other note names that I have tried do.

I do not really understand how gethash could be misbehaving if that is
really what is happening.  (My MCL image was saved right after loading CM
AFAIK).

*shrug*

...

OH!  Just discovered that the symbols in the hash table seem to have extra
characters.  'a4 is in there as '\a4 and 'gs1 as '|gs1| for example.  Does
this make any sense??

Anthony Kozar
anthonykozar AT sbcglobal DOT net
http://akozar.spymac.net/


Rick Taube wrote on 10/3/05 7:44 PM:

> Hi -- I dont have MCL anymore so if I can't build 2.3.4 in openmcl im
> not sure if ill be able to debug it.
> FWIW it does work in 2.7.0:



From taube at uiuc.edu  Tue Oct  4 03:46:38 2005
From: taube at uiuc.edu (Rick Taube)
Date: Tue, 4 Oct 2005 05:46:38 -0500
Subject: [CM] Trouble using tunings
In-Reply-To: <BF679C86.11FE9%anthonykozar@sbcglobal.net>
References: <BF679C86.11FE9%anthonykozar@sbcglobal.net>
Message-ID: <ad8d7270cca3ca97b27683de71563e27@uiuc.edu>

> OH!  Just discovered that the symbols in the hash table seem to have 
> extra
> characters.  'a4 is in there as '\a4 and 'gs1 as '|gs1| for example.  
> Does
> this make any sense??
>
> Anthony Kozar
> anthonykozar AT sbcglobal DOT net
> http://akozar.spymac.net/


this is some sort of case-sensitivity issue - somehow the note-name 
symbols are lower-case but the REPL is in normal uppercase insensitive. 
did you do anything to the case-sensitivity of the Lisp reader when you 
saved the image?




From anthonykozar at sbcglobal.net  Tue Oct  4 11:03:43 2005
From: anthonykozar at sbcglobal.net (Anthony Kozar)
Date: Tue, 04 Oct 2005 14:03:43 -0400
Subject: [CM] Trouble using tunings
In-Reply-To: <ad8d7270cca3ca97b27683de71563e27@uiuc.edu>
Message-ID: <BF683C3E.11FFC%anthonykozar@sbcglobal.net>

Well, it appears that at some point I changed the value of the *print-case*
variable from :upcase to :downcase (which I think looks nicer).  I would
never have imagined that the printing behavior would affect anything
important, but changing it back does fix the problem!

Basically, if *print-case* is not :upcase when a tuning is defined, then the
'a4, 'c2, etc. symbols will not work later, no matter what the value of
*print-case* currently is.  *print-case* can be changed to :downcase after
defining any tunings and it appears that things continue to work correctly.

Thanks very much for the help.  And sorry to bother you over an obscure
"user error"!

Anthony Kozar
anthonykozar AT sbcglobal DOT net
http://akozar.spymac.net/


Rick Taube wrote on 10/4/05 6:46 AM:

> this is some sort of case-sensitivity issue - somehow the note-name
> symbols are lower-case but the REPL is in normal uppercase insensitive.
> did you do anything to the case-sensitivity of the Lisp reader when you
> saved the image?



From cycle.code.media at gmail.com  Tue Oct  4 19:50:04 2005
From: cycle.code.media at gmail.com (James Baker)
Date: Wed, 5 Oct 2005 13:50:04 +1100
Subject: [CM] Common Music and SPEAR
Message-ID: <428d2dc0510041950w10ecfe2x8417f377b4a71b80@mail.gmail.com>

hello,

after reading rick's paper on spectral composition and checking spear
out im very interested in spectral composition methods, but none of
the examples in said pdf seem to work.  am i doing something incorrect
or is the cm->spear connection not ready for public consumption just
yet ?

cheers,

james



From taube at uiuc.edu  Wed Oct  5 05:28:44 2005
From: taube at uiuc.edu (Rick Taube)
Date: Wed, 5 Oct 2005 07:28:44 -0500
Subject: [CM] Common Music and SPEAR
In-Reply-To: <428d2dc0510041950w10ecfe2x8417f377b4a71b80@mail.gmail.com>
References: <428d2dc0510041950w10ecfe2x8417f377b4a71b80@mail.gmail.com>
Message-ID: <1ffc60db0f1d3a5e80e33b4991aa87f6@uiuc.edu>

Ill try to find some time today to make sure the examples work, in any 
event they will only run in cm 2.7.0.
--rick
> hello,
>
> after reading rick's paper on spectral composition and checking spear
> out im very interested in spectral composition methods, but none of
> the examples in said pdf seem to work.  am i doing something incorrect
> or is the cm->spear connection not ready for public consumption just
> yet ?
>
> cheers,
>
> james
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From taube at uiuc.edu  Wed Oct  5 07:16:53 2005
From: taube at uiuc.edu (Rick Taube)
Date: Wed, 5 Oct 2005 09:16:53 -0500
Subject: [CM] Common Music and SPEAR
In-Reply-To: <428d2dc0510050531p42d07a77t9a1ed7612c4f9a33@mail.gmail.com>
References: <428d2dc0510041950w10ecfe2x8417f377b4a71b80@mail.gmail.com> <1ffc60db0f1d3a5e80e33b4991aa87f6@uiuc.edu> <428d2dc0510050531p42d07a77t9a1ed7612c4f9a33@mail.gmail.com>
Message-ID: <a2c04f35d2265f997d2602688ca41b2a@uiuc.edu>

Ive verified that the examples all work in openmcl (osx). Note that 
anything involving FOMUS will not work in Scheme. I am having some 
basic problems in Gauche even with loading the sources and im not sure 
why, ill have to upgrade my Gauche version before trying to debug 
things.


> thanks for that, i am using a fairly recent cvs checkout and gauche 
> 0.8.5.
>
> james
>
> On 10/5/05, Rick Taube <taube at uiuc.edu> wrote:
>> Ill try to find some time today to make sure the examples work, in any
>> event they will only run in cm 2.7.0.
>> --rick
>>> hello,
>>>
>>> after reading rick's paper on spectral composition and checking spear
>>> out im very interested in spectral composition methods, but none of
>>> the examples in said pdf seem to work.  am i doing something 
>>> incorrect
>>> or is the cm->spear connection not ready for public consumption just
>>> yet ?
>>>
>>> cheers,
>>>
>>> james
>>>
>>> _______________________________________________
>>> Cmdist mailing list
>>> Cmdist at ccrma.stanford.edu
>>> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>>
>>



From taube at uiuc.edu  Thu Oct  6 05:20:55 2005
From: taube at uiuc.edu (Rick Taube)
Date: Thu, 6 Oct 2005 07:20:55 -0500
Subject: [CM] Common Music and SPEAR
In-Reply-To: <428d2dc0510050531p42d07a77t9a1ed7612c4f9a33@mail.gmail.com>
References: <428d2dc0510041950w10ecfe2x8417f377b4a71b80@mail.gmail.com> <1ffc60db0f1d3a5e80e33b4991aa87f6@uiuc.edu> <428d2dc0510050531p42d07a77t9a1ed7612c4f9a33@mail.gmail.com>
Message-ID: <4d0403b6698bd25d4f7757aa7635760b@uiuc.edu>

The spectral workshop examples now run in Gauche 0.8.5.  Since FOMUS 
does not work in Scheme I have added a scheme-only version of the 
examples (with FOMUS notes converted to MIDI events) to the tarball. 
You can download just the scheme example file from pinhead but it wont 
be there for long:

	http://pinhead.music.uiuc.edu/~hkt/examples.scm

I had to fix one small bug that prevented spear data from being 
imported in Scheme, the fix is in rev 1.20 of spectral.scm so you will 
need to do a cvs update before running the examples.
--rick


> thanks for that, i am using a fairly recent cvs checkout and gauche 
> 0.8.5.
>
> james
>
> On 10/5/05, Rick Taube <taube at uiuc.edu> wrote:
>> Ill try to find some time today to make sure the examples work, in any
>> event they will only run in cm 2.7.0.
>> --rick
>>> hello,
>>>
>>> after reading rick's paper on spectral composition and checking spear
>>> out im very interested in spectral composition methods, but none of
>>> the examples in said pdf seem to work.  am i doing something 
>>> incorrect
>>> or is the cm->spear connection not ready for public consumption just
>>> yet ?
>>>
>>> cheers,
>>>
>>> james
>>>
>>> _______________________________________________
>>> Cmdist mailing list
>>> Cmdist at ccrma.stanford.edu
>>> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>>
>>



From cycle.code.media at gmail.com  Thu Oct  6 05:25:58 2005
From: cycle.code.media at gmail.com (James Baker)
Date: Thu, 6 Oct 2005 23:25:58 +1100
Subject: [CM] Common Music and SPEAR
In-Reply-To: <4d0403b6698bd25d4f7757aa7635760b@uiuc.edu>
References: <428d2dc0510041950w10ecfe2x8417f377b4a71b80@mail.gmail.com>
	 <1ffc60db0f1d3a5e80e33b4991aa87f6@uiuc.edu>
	 <428d2dc0510050531p42d07a77t9a1ed7612c4f9a33@mail.gmail.com>
	 <4d0403b6698bd25d4f7757aa7635760b@uiuc.edu>
Message-ID: <428d2dc0510060525t761316d5ta3dace63765380ae@mail.gmail.com>

thanks again for taking the time, much appreciated!

james

On 10/6/05, Rick Taube <taube at uiuc.edu> wrote:
> The spectral workshop examples now run in Gauche 0.8.5.  Since FOMUS
> does not work in Scheme I have added a scheme-only version of the
> examples (with FOMUS notes converted to MIDI events) to the tarball.
> You can download just the scheme example file from pinhead but it wont
> be there for long:
>
>         http://pinhead.music.uiuc.edu/~hkt/examples.scm
>
> I had to fix one small bug that prevented spear data from being
> imported in Scheme, the fix is in rev 1.20 of spectral.scm so you will
> need to do a cvs update before running the examples.
> --rick
>
>
> > thanks for that, i am using a fairly recent cvs checkout and gauche
> > 0.8.5.
> >
> > james
> >
> > On 10/5/05, Rick Taube <taube at uiuc.edu> wrote:
> >> Ill try to find some time today to make sure the examples work, in any
> >> event they will only run in cm 2.7.0.
> >> --rick
> >>> hello,
> >>>
> >>> after reading rick's paper on spectral composition and checking spear
> >>> out im very interested in spectral composition methods, but none of
> >>> the examples in said pdf seem to work.  am i doing something
> >>> incorrect
> >>> or is the cm->spear connection not ready for public consumption just
> >>> yet ?
> >>>
> >>> cheers,
> >>>
> >>> james
> >>>
> >>> _______________________________________________
> >>> Cmdist mailing list
> >>> Cmdist at ccrma.stanford.edu
> >>> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
> >>
> >>
>
>



From taube at uiuc.edu  Thu Oct  6 05:36:05 2005
From: taube at uiuc.edu (Rick Taube)
Date: Thu, 6 Oct 2005 07:36:05 -0500
Subject: [CM] Re: metalevel problem
In-Reply-To: <000001c5ca21$4d620c50$0300a8c0@home7vdrzzxq9o>
References: <000001c5ca21$4d620c50$0300a8c0@home7vdrzzxq9o>
Message-ID: <4c719a720d6ce39668cf544fdcdaf554@uiuc.edu>

Hi, the examples in the book are several years old now -- Im suprised 
that this is the only problem youve run into so far! Here is how to 
reset pitch bend values in  cm 2.7.0:

(define (clearbends)
   (process for i below 16
            output (new midi-pitch-bend :time (now)
                        :channel i :bend 0)))


In 2.7.0 ther3 are class objects for all midi message types in addtion 
to the "low level" messages that are created using function like 
make-pitch-bend.
Documentation for both levels can be found at:

	http://commonmusic.sf.net/doc/dict/midi-topic.html

For more info on tunings and modes you might also look in the tutorial:

	http://commonmusic.sourceforge.net/etc/tutorials/scales.lisp

and also in drew krauses code archive:

	http://www.wordecho.org/code/cmcode/cmusic.html

On Oct 5, 2005, at 9:54 PM, Denis Fitzpatrick wrote:

> Rick
>
> When running the following from Metalevel Ch. 15 (scales.cm)
>
> (define (clearbends )
>   (process for c below 16
>            output (new midi :time (now)
>                        :msg (make-pitch-bend c 0))))
>
> I get the following:
> *** - FIND-CLASS: MIDIMSG does not name a class
>
> I'm running cm 2.7.0 under xemacs on windows xp
>
> Everything has worked great to this point.  I'm enjoying metalevel 
> immensely
> - thanks.
>
> Denis Fitzpatrick
>



From dlphillips at woh.rr.com  Fri Oct  7 07:52:18 2005
From: dlphillips at woh.rr.com (Dave Phillips)
Date: Fri, 07 Oct 2005 10:52:18 -0400
Subject: [CM] error loading FOMUS in CM 2.7.0 CVS
Message-ID: <43468BA2.3060104@woh.rr.com>

Greetings:

  When loading FOMUS from cmio (which is really cool) :

; Error: (during macroexpansion)
; Error in FDEFINITION:  the function EVENT-PARTID is undefined.

; Error: (during macroexpansion)
; Error in FDEFINITION:  the function TIMESIG-PARTIDS is undefined.

; Error: (during macroexpansion)
; Error in FDEFINITION:  the function PART-PARTID is undefined.

; Error: (during macroexpansion)
; Error in FDEFINITION:  the function (SETF EVENT-PARTID) is undefined.

; Error: (during macroexpansion)
; Error in FDEFINITION:  the function (SETF TIMESIG-PARTIDS) is undefined.

; Error: (during macroexpansion)
; Error in FDEFINITION:  the function (SETF PART-PARTID) is undefined.

; Compilation unit finished.
;   6 errors

;; Loading #P"/home/dlphilp/fomus/classes.x86f".

Execution of a form compiled with errors:
 (DEFMETHOD OBJ-PARTID ((EV EVENT-BASE))
   (EVENT-PARTID EV))
   [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR]

Restarts:
  0: [CONTINUE] Return NIL from load of 
#P"/home/dlphilp/fomus/classes.x86f".
  1:            Return NIL from load of "/home/dlphilp/fomus/load.lisp".
  2: [ABORT   ] Return to Top-Level.

Debug  (type H for help)

(C::DO-CALL #<Code Object "Top-Level Form" {5806C397}> 35 36 14 ...)
Source: Error finding source:
Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no longer 
exists:
  target:code/byte-interp.lisp.
0] 2
* (quit)
[dlphilp at localhost cm]$


  This is with CMUCL from Planet CCRMA's Red Hat 9.

  Btw, CLM, CMN, and PortMIDI all load from cmio without problems. 
Haven't tested the SC stuff at all yet. :(

Best,

dp



From taube at uiuc.edu  Fri Oct  7 09:00:19 2005
From: taube at uiuc.edu (Rick Taube)
Date: Fri, 7 Oct 2005 11:00:19 -0500
Subject: [CM] error loading FOMUS in CM 2.7.0 CVS
In-Reply-To: <43468BA2.3060104@woh.rr.com>
References: <43468BA2.3060104@woh.rr.com>
Message-ID: <f06d1ed8e3080e669c7117920baaa85e@uiuc.edu>

dont try rts with cmio!! you will lose bigtime in cmucl -- Im trying to 
get that going right now ill know later today or tomorrow if it works 
at all or not. I also have to commit some stuff that ive added/fixed 
today for cmucl/linux, will try to update the cvs tree  in an hour or 
so.

i didnt "announce" the new cmio yet because im still testing but since 
you brought it up: the cmio window has been completely rehauled, ive 
added pages for Fomus, Portmidi, SC, OSC Realtime.  i have to say it 
looks really cool in the gtk installed in planet ccrma.


By default the OSC page is set up to do SuperCollider real time but 
since todd hasnt mentioned this yet i wont say any more.



>  Btw, CLM, CMN, and PortMIDI all load from cmio without problems. 
> Haven't tested the SC stuff at all yet. :(




From taube at uiuc.edu  Fri Oct  7 12:44:01 2005
From: taube at uiuc.edu (Rick Taube)
Date: Fri, 7 Oct 2005 14:44:01 -0500
Subject: [CM] error loading FOMUS in CM 2.7.0 CVS
In-Reply-To: <43468BA2.3060104@woh.rr.com>
References: <43468BA2.3060104@woh.rr.com>
Message-ID: <a16f99d2a310b9e6ef951c4e402a9df7@uiuc.edu>

ive checked in the linux changes and it should now be ok to run cmio in 
parallel with rts in linux/cmucl. of course the performance of rts wont 
be very good in cmucl because it will run as a periodic task along with 
GTK (rather than as an os thread).  I actually added support for 
threaded rts to cmucl today but their internal-time-units-per-second is 
only 100ms, way to slow to be useful. perhaps if i can ever figure out 
semaphores i can get around this.

Anyway to try out rts in the cmio window, select portmidi or midishare 
page, click "open", then select the realtime page,  click "start" , 
enter your expr to sprout -- try "(bar (between 10 20))" for starters 
-- then you can press "compose" button in "realtime", ie whenever you 
want to launch a new bar process into rts. seems to work for me modulo 
all the handwaving in my first paragraph.

it will take a few hours for all he file changes to migrate to 
anonymous cvs. since lots of stuff has changed i would reccoment that 
you delete all you cm binaries before compiling:
	cd cm/bin
	rm -r cmu*



From dlphillips at woh.rr.com  Sun Oct  9 17:21:58 2005
From: dlphillips at woh.rr.com (Dave Phillips)
Date: Sun, 09 Oct 2005 20:21:58 -0400
Subject: [CM] error loading FOMUS in CM 2.7.0 CVS
In-Reply-To: <a16f99d2a310b9e6ef951c4e402a9df7@uiuc.edu>
References: <43468BA2.3060104@woh.rr.com> <a16f99d2a310b9e6ef951c4e402a9df7@uiuc.edu>
Message-ID: <4349B426.50001@woh.rr.com>

Hi Rick:

  Thanks for the update. I've been testing the FOMUS and RTS stuff a wee 
bit, it seems to work as advertised. The examples were run without 
error, even your example below ran well under CMUCL. I'll try to get 
some more intense testing done this week. Time is always in short supply 
around here...

  Linux, Planet CCRMA RH9, Common Music 2.7.0 CVS sources.

Best,

dp



Rick Taube wrote:

> ive checked in the linux changes and it should now be ok to run cmio 
> in parallel with rts in linux/cmucl. of course the performance of rts 
> wont be very good in cmucl because it will run as a periodic task 
> along with GTK (rather than as an os thread).  I actually added 
> support for threaded rts to cmucl today but their 
> internal-time-units-per-second is only 100ms, way to slow to be 
> useful. perhaps if i can ever figure out semaphores i can get around 
> this.
>
> Anyway to try out rts in the cmio window, select portmidi or midishare 
> page, click "open", then select the realtime page,  click "start" , 
> enter your expr to sprout -- try "(bar (between 10 20))" for starters 
> -- then you can press "compose" button in "realtime", ie whenever you 
> want to launch a new bar process into rts. seems to work for me modulo 
> all the handwaving in my first paragraph.
>
> it will take a few hours for all he file changes to migrate to 
> anonymous cvs. since lots of stuff has changed i would reccoment that 
> you delete all you cm binaries before compiling:
>     cd cm/bin
>     rm -r cmu*
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>



From taube at uiuc.edu  Mon Oct 10 07:12:43 2005
From: taube at uiuc.edu (Rick Taube)
Date: Mon, 10 Oct 2005 09:12:43 -0500
Subject: [CM] Re: CM + GTK2
In-Reply-To: <200510101913.11252.ahcnz@ihug.co.nz>
References: <200510101913.11252.ahcnz@ihug.co.nz>
Message-ID: <d8482798fd76305c0b715632b60ca672@uiuc.edu>

Hi Adam, you might be able to point cm's GTK FFI to your GTK libs by  
defining the variable cl-user::*gtk-libdir* to point to where your libs  
are stored before the file before you load cm:
	(defparameter cl-user::*gtk-libdir* "/path/to/gtk2/libs/")
      (load "/path/to/cm/src/cm.lisp")
But this assumes your GTK lib names are the same as what  
gtkffi-cmusbcl.lisp expects. otherwise you will have to edit the file  
and change the lib linking code.

on the other hand, if you are thinking of switching to fedora core  
anyway, consider just installing the latest Planet CCRMA distro -- I  
installed it without any glitches and its been a very stable  
environment -- all the music apps are already there and they work.

On Oct 10, 2005, at 1:13 AM, Adam Chennells wrote:

> Rick,
>
> Am trying to re-build latest CM.   [My first time from CVS].
> I have Cmucl on Mandrake 10 linux. Older CM works.
>
> It seems it can't find GTK2.
> But I believe I have a GTK2 on my system, see
> at the very bottom below.
>
> I'm hesitant to launch into an upgrade or install of the
> latest GTK2, for fear it would break many existing things.
>
> Perhaps I need only  lambda-to-GTK  installed?
> Or point install just the  libgtk-x11-2.0  source or
> libraries ?
>
> Or else, perhaps I can "upgrade" my GTK from
> the Mandrake site - painlessly I hope.
> The option of manually removing GTK and installing
> GTK2 doesn't look attractive to me just at present.
>
> Do you have any advice at this stage ?
>
> 	best,	Adam.
>
>
>
> OT.  Am considering switching to RH Fedora 3 or 4,
> so maybe that would ease the upgrade of GTK2.0.
> And I hope it will get Midishare running on a newer
> kernel too.  So perhaps I should fast-track this?
>
>
>
>
> [ahc at localhost bin]$ ./cm.sh
>
> ; Installation directory: "/home/ahc/cmcvs/cm/"
> ; Compiling "src/gui/gtkffi-cmusbcl.lisp"
>
> Error in function (FLET LIBPATH
>                     ):
>    Library "/usr/lib/libgtk-x11-2.0.so" not found. Either GTK is not  
> installed
> or else cl-user:*gtk-libdir* needs to be set to the directory  
> containing GTK
> on your machine.
>    [Condition of type SIMPLE-ERROR]
>
> Restarts:
>   0: [CONTINUE] Return NIL from load of
> "/home/ahc/cmcvs/cm/bin/./../src/cm.lisp".
>   1: [ABORT   ] Skip remaining initializations.
>
> Debug  (type H for help)
>
> ((FLET LIBPATH
>    )
>  "libgtk-x11-2.0")
> Source: (ERROR
>          "Library ~S not found. Either GTK is not installed or else
> cl-user:*gtk-libdir* needs to be set to the directory containing GTK  
> on your
> machine."
>          P)
> 0] 1
>
> ; Compilation unit aborted.
>
>
>
>
> [ahc at localhost ahc]$ find /usr/lib/gt
> gtk      gtk-2.0
> [ahc at localhost ahc]$ dir /usr/lib/gtk/
> themes
> [ahc at localhost ahc]$ dir /usr/lib/gtk/themes/
> engines
> [ahc at localhost ahc]$ dir /usr/lib/gtk/themes/engines/
> libeazel-engine.la  libmetal.so   libpixmap.so     libredmond95.so
> libwonderland.so
> libeazel-engine.so  libnotif.la   libraleigh.la    libthinice.la
> libgalaxy.so        libnotif.so   libraleigh.so    libthinice.so
> libmetal.la         libpixmap.la  libredmond95.la  libwonderland.la
> [ahc at localhost ahc]$ dir /usr/lib/gtk-2.0/
> 2.2.0  modules
> [ahc at localhost ahc]$ dir /usr/lib/gtk-2.0/2.2.0/
> engines  immodules  loaders
> [ahc at localhost ahc]$ dir /usr/lib/gtk-2.0/modules/
> libferret.so  libgail.so
> [ahc at localhost ahc]$
>
>
>
>
> [ find /usr/ -name "gtk*" ]
>
> [ ... ]
> /usr/share/gtkrc/gtkrc.zh_TW
> /usr/share/gimp/1.2/help/C/dialogs/color_selectors/gtk.html
> /usr/share/mcc/themes/default/gtkrc
> /usr/share/dia/python/gtkcons.py
> /usr/bin/gtk-query-immodules-2.0
> /usr/lib/perl5/vendor_perl/5.8.3/i386-linux-thread-multi/Gtk2/Install/ 
> gtk2perl-autogen.h
> /usr/lib/perl5/vendor_perl/5.8.3/i386-linux-thread-multi/Gtk2/Install/ 
> gtk.typemap
> /usr/lib/perl5/vendor_perl/5.8.3/i386-linux-thread-multi/Gtk2/Install/ 
> gtk2perl.typemap
> /usr/lib/perl5/vendor_perl/5.8.3/i386-linux-thread-multi/Gtk2/Install/ 
> gtk2perl.h
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-0.99.typemap
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-0.99.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.0.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.1.1.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.1.3.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.1.4.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.1.6.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.1.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.2.defs
> /usr/lib/perl5/vendor_perl/5.8.1/i386-linux-thread-multi/Gtk/Install/ 
> gtk-1.2.typemap
> /usr/lib/python2.3/site-packages/gtk-2.0
> /usr/lib/python2.3/site-packages/gtk-2.0/gtk
> /usr/lib/gtk-2.0
> /usr/lib/libDrakX/interactive/gtk.pm
> /usr/lib/gtk
>



From gogins at pipeline.com  Mon Oct 10 10:43:02 2005
From: gogins at pipeline.com (Michael Gogins)
Date: Mon, 10 Oct 2005 13:43:02 -0400 (GMT-04:00)
Subject: [CM] Common Music and Csound 5
Message-ID: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>

I will undertake a Lisp or Scheme binding for Csound 5 so that the Csound API can be called natively from Common Music.

Before I do this, I would like to know which version of Scheme or Lisp I should target. I will be doing the development work on Windows and Linux. Rick Taube's book comes with CLisp for Windows and this also can be used on OS X and Linux, so I would guess this is the best target.  CLisp runs Common Music just fine on my PC. 

I do not have time to maintain multiple bindings, so I need to know the best compromise.

Unfortunately, the most mature SWIG facilities for Lisp are actually for Guile and MzScheme (both Scheme). On my PC, Guile does not run and building it requires many dependencies, which I can obtain if consensus favors Guile over CLisp.

As part of this work, I will be moving the C++ form of the Csound API from the CsoundVST project into the Csound API library proper, and also moving the Python wrappers. 

This will NOT introduce any additional dependencies in the Csound API library build except for SWIG, and I can book in the generated wrapper code so that even SWIG is not required for the normal build.

Please advise me which is the best version of Lisp or Scheme to use for Common Music and the Csound API.

I would like to finish this work before Csound 5 achieves release status.

Thanks,
Mike





From arenato at genius.org.br  Mon Oct 10 11:41:41 2005
From: arenato at genius.org.br (arenato at genius.org.br)
Date: Mon, 10 Oct 2005 16:41:41 -0200
Subject: [CM] CLM: ATS and Motif
Message-ID: <OF3C51D4EB.056C631C-ON04257096.0065E677-83257096.0066D775@genius.org.br>

I have problems when I tried to compile ATS with CLM and CMUCL. The 
version of CMUCL is 19b. I tried with both CLM-3 and CLM-2.
First,, during the compilation it?s necesary the library libXm.so.2, when 
my system have libXm.so.3 (Lesstif). I recompiled CMUCL with Lesstif 
version 2.2.3  without succes. Second, when I compile ATS inside lisp and 
CLM, there is a function no defined ( I don?t remember well, GET_FUN....) 
when compile "do_partials", and the compilation fail.
Would you give me any suggestions about? Thanks in advance.

Alejandro Renato
Genius Insituto de Tecnolog?a
**************************************************************************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051010/f3820695/attachment.html>

From taube at uiuc.edu  Mon Oct 10 11:50:47 2005
From: taube at uiuc.edu (Rick Taube)
Date: Mon, 10 Oct 2005 13:50:47 -0500
Subject: [CM] Re: Common Music and Csound 5
In-Reply-To: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>
References: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>
Message-ID: <dccdd48f711d627a87977bb743d8543b@uiuc.edu>

On Oct 10, 2005, at 12:43 PM, Michael Gogins wrote:

> I will undertake a Lisp or Scheme binding for Csound 5 so that the 
> Csound API can be called natively from Common Music.
>
> Before I do this, I would like to know which version of Scheme or Lisp 
> I should target. I will be doing the development work on Windows and 
> Linux. Rick Taube's book comes with CLisp for Windows and this also 
> can be used on OS X and Linux, so I would guess this is the best 
> target.  CLisp runs Common Music just fine on my PC.
>
> I do not have time to maintain multiple bindings, so I need to know 
> the best compromise.

Hi, In my opinion the best compromise would be to write your code using 
CFFI for Common Lisp -- that way your code should run in every Lisp 
that CFFI supports, and it currently supports most CLs on all the major 
platforms. You can read about CFFI here:

	http://common-lisp.net/project/cffi/

Note that currently CM does not use CFFI yet but I will definately be 
moving to it (both CM and Lambda-GTK) in the fairly near future At any 
rate that would not have any bearing on your work, ehich if I 
understandd it, could be used by any common lisp (or scheme) program.

For Scheme there is no simple solution -- my understanding is that the 
next release of Gauche Scheme  (0.8.6) will include an "easy" FFI 
interface.

	http://www.shiro.dreamhost.com/scheme/gauche/

If you do a scheme binding I would perfer Gauche over Guile, mainly 
because we have CM's realtime scheduler working in Gauche (sort of...), 
but also because I have had too many problems trying to build Guile on 
OS X. Hoever CM 2.7.0 does run in Guile, Gauche and STklos Schemes.

> As part of this work, I will be moving the C++ form of the Csound API 
> from the CsoundVST project into the Csound API library proper, and 
> also moving the Python wrappers.

An FFI to C++ will entail additional headache, there is a project 
called Verragano that does this, I belileve that it uses CFFI to help 
do this:

	http://common-lisp.net/project/fetter/
>
> I would like to finish this work before Csound 5 achieves release 
> status.

Depending on what you want to do, another possibilty would be to 
connect Csound and CM using OSC via UDP -- I think that OSC/UDP is 
working in CM right now, or if it isnt, it will be very shortly (ie 
days not weeks)

Hope this helps!

--rick



From carl.boingie at verizon.net  Mon Oct 10 17:26:10 2005
From: carl.boingie at verizon.net (Carl Edwards)
Date: Mon, 10 Oct 2005 20:26:10 -0400
Subject: [CM] Common Music and Csound 5
In-Reply-To:  <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>
Message-ID: <BF707EE2.28B7%carl.boingie@verizon.net>



Hi Michael,

That's very exciting-- what great news!

My 2 cents:

I've always found Clisp the easiest way to go on Windows. I'd re-install
Cygwin if you went that way though not very eagerly.



Carl Edwards





> I will undertake a Lisp or Scheme binding for Csound 5 so that the Csound API
> can be called natively from Common Music.
> 
> Before I do this, I would like to know which version of Scheme or Lisp I
> should target. I will be doing the development work on Windows and Linux. Rick
> Taube's book comes with CLisp for Windows and this also can be used on OS X
> and Linux, so I would guess this is the best target.  CLisp runs Common Music
> just fine on my PC.
> 
> I do not have time to maintain multiple bindings, so I need to know the best
> compromise.
> 
> Unfortunately, the most mature SWIG facilities for Lisp are actually for Guile
> and MzScheme (both Scheme). On my PC, Guile does not run and building it
> requires many dependencies, which I can obtain if consensus favors Guile over
> CLisp.
> 
> As part of this work, I will be moving the C++ form of the Csound API from the
> CsoundVST project into the Csound API library proper, and also moving the
> Python wrappers. 
> 
> This will NOT introduce any additional dependencies in the Csound API library
> build except for SWIG, and I can book in the generated wrapper code so that
> even SWIG is not required for the normal build.
> 
> Please advise me which is the best version of Lisp or Scheme to use for Common
> Music and the Csound API.
> 
> I would like to finish this work before Csound 5 achieves release status.
> 
> Thanks,
> Mike



From znmeb at cesmail.net  Mon Oct 10 19:31:03 2005
From: znmeb at cesmail.net (M. Edward (Ed) Borasky)
Date: Mon, 10 Oct 2005 19:31:03 -0700
Subject: [CM] Common Music and Csound 5
In-Reply-To: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>
References: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>
Message-ID: <434B23E7.5070807@cesmail.net>

Speaking mostly as a Linux user, the best Lisp for Common Music in my 
environment has proved to be CMUCL. SBCL is "almost" there but CMUCL 
supports Rick's new GUI out of the box. Clisp doesn't support the GUI 
but will run CM. Guile doesn't support the GUI but will run CM. GCL 
won't run CM.

On the Windows side, I haven't used any Lisp but the CygWin port of 
Clisp. Most of the CygWin Lisp/Scheme packages use Clisp or Guile. In 
any event, both Clisp and Guile are ready on CygWin, so the choice is up 
to you.

So ... Rick ... how difficult would it be to put the CM GUI up on Clisp 
or Guile?

Michael Gogins wrote:

>I will undertake a Lisp or Scheme binding for Csound 5 so that the Csound API can be called natively from Common Music.
>
>Before I do this, I would like to know which version of Scheme or Lisp I should target. I will be doing the development work on Windows and Linux. Rick Taube's book comes with CLisp for Windows and this also can be used on OS X and Linux, so I would guess this is the best target.  CLisp runs Common Music just fine on my PC. 
>
>I do not have time to maintain multiple bindings, so I need to know the best compromise.
>
>Unfortunately, the most mature SWIG facilities for Lisp are actually for Guile and MzScheme (both Scheme). On my PC, Guile does not run and building it requires many dependencies, which I can obtain if consensus favors Guile over CLisp.
>
>As part of this work, I will be moving the C++ form of the Csound API from the CsoundVST project into the Csound API library proper, and also moving the Python wrappers. 
>
>This will NOT introduce any additional dependencies in the Csound API library build except for SWIG, and I can book in the generated wrapper code so that even SWIG is not required for the normal build.
>
>Please advise me which is the best version of Lisp or Scheme to use for Common Music and the Csound API.
>
>I would like to finish this work before Csound 5 achieves release status.
>
>Thanks,
>Mike
>
>
>
>_______________________________________________
>Cmdist mailing list
>Cmdist at ccrma.stanford.edu
>http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>
>  
>

-- 
M. Edward (Ed) Borasky

http://www.borasky-research.net/
http://borasky-research.blogspot.com/

http://pdxneurosemantics.com
http://pdx-sales-coach.com
http://algocompsynth.com



From taube at uiuc.edu  Tue Oct 11 04:03:24 2005
From: taube at uiuc.edu (Rick Taube)
Date: Tue, 11 Oct 2005 06:03:24 -0500
Subject: [CM] Re: Common Music and Csound 5
In-Reply-To: <000201c5ce1b$8d7c9c80$6b01a8c0@Generator>
References: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net> <dccdd48f711d627a87977bb743d8543b@uiuc.edu> <000201c5ce1b$8d7c9c80$6b01a8c0@Generator>
Message-ID: <f6279cff66f2773814232db729a5eb4c@uiuc.edu>

On Oct 10, 2005, at 2:38 PM, Michael Gogins wrote:

> Thanks for your prompt and helpful reply.
>
> I may leave the C++ binding out -- I was under the impression from 
> (only) looking at SWIG that C++ was a better way to bind. The 
> procedural interface in Python works fine now like this:
>
> cs = csound.csoundCreate()
> csound.csoundCompile(cs)
> csound.csoundPerform(cs)
>
> Presumably the Lisp version would be roughly as simple -- cs here is a 
> pointer to Csound.

As you will probably find out nothing is simple in Lisp ffi work since 
ffis are not covered in the language spec. but cffi will probably be 
the closest you can get to "simple"!

>
> It does sound like the best thing for me to do is provide the CFFI 
> interface. What do you need to do to make this interface accessible 
> from CM?

Nothing special needs to happen. You could modularize your Csound glue 
code as a Lisp package called Csound (or whatever) and then "export" 
your API for other programs to use. That way your code will be a 
generic Lisp API to Csound that any Lisp program could load, it 
shouldnt have to know anything about CM. Then I (or maybe you and I...) 
would add new CM input/output stream support based on your packge. This 
is how I treat systems like CLM, CMN, Fomus, Midishare etc and it seems 
to work well.

> As for Gauche, it doesn't seem to run on Windows, so it's out.

I think there is an mingw version, but I dont knwo if it works:
	http://www.shiro.dreamhost.com/scheme/vault/Gauche-mingw-0.8.5.zip
its a binary, so if you have mingw you can install and try it out.



From taube at uiuc.edu  Tue Oct 11 04:29:21 2005
From: taube at uiuc.edu (Rick Taube)
Date: Tue, 11 Oct 2005 06:29:21 -0500
Subject: [CM] Common Music and Csound 5
In-Reply-To: <434B23E7.5070807@cesmail.net>
References: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net> <434B23E7.5070807@cesmail.net>
Message-ID: <939cdc0ab00295131e2e2aa39a187f5b@uiuc.edu>

> So ... Rick ... how difficult would it be to put the CM GUI up on 
> Clisp or Guile?

Over the past two years it seems CLisp development has really been 
cooking, they now have a good ffi, callback support and it looks like 
they are working on native os threads too. So theoretically the GUI 
would work right now if there were a Clisp ffi binding for it.
But rather than writing a clisp specific ffi, if Lambda-GTK generated a 
CFFI backend then CM's gui code could use that to run in all 4 lisps 
(openmcl, sbcl, cmucl and clisp). At that point I would convert CMs 
other ffi's (Portmidi and Midishare) to CFFI, and then require that 
CFFI be installed before CM is compiled.

So thats the way to go I think, provided that CFFI can actully do the 
things that GTK requires,  like refereneceing included structs, setting 
slots in included structs (widget.rectangle.x), slots with arrays of 
pointers. If CFFI is strong enough, then it probably wouldnt be too 
much work to take the existing lambda-gtk backend for cmu/sbcl and 
change its output statements to output CFFI definitions rather than 
SBCL definitions. Its just a small matter of someone rolling up their 
sleeves and doing it.

Scheme is another story. Stklos, Guile and Gauche all have built in GTK 
apis, but as far as I know there is no SFRI for a common scheme FFI. 
This means someone would have to simply choose one, then take the CM 
gui sources and convert them to Scheme using either Guile or Gauche orr 
STklos.

> Michael Gogins wrote:
>
>> I will undertake a Lisp or Scheme binding for Csound 5 so that the 
>> Csound API can be called natively from Common Music.
>>
>> Before I do this, I would like to know which version of Scheme or 
>> Lisp I should target. I will be doing the development work on Windows 
>> and Linux. Rick Taube's book comes with CLisp for Windows and this 
>> also can be used on OS X and Linux, so I would guess this is the best 
>> target.  CLisp runs Common Music just fine on my PC.
>> I do not have time to maintain multiple bindings, so I need to know 
>> the best compromise.
>>
>> Unfortunately, the most mature SWIG facilities for Lisp are actually 
>> for Guile and MzScheme (both Scheme). On my PC, Guile does not run 
>> and building it requires many dependencies, which I can obtain if 
>> consensus favors Guile over CLisp.
>>
>> As part of this work, I will be moving the C++ form of the Csound API 
>> from the CsoundVST project into the Csound API library proper, and 
>> also moving the Python wrappers.
>> This will NOT introduce any additional dependencies in the Csound API 
>> library build except for SWIG, and I can book in the generated 
>> wrapper code so that even SWIG is not required for the normal build.
>>
>> Please advise me which is the best version of Lisp or Scheme to use 
>> for Common Music and the Csound API.
>>
>> I would like to finish this work before Csound 5 achieves release 
>> status.
>>
>> Thanks,
>> Mike
>>
>>
>>
>> _______________________________________________
>> Cmdist mailing list
>> Cmdist at ccrma.stanford.edu
>> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>>
>>
>
> -- 
> M. Edward (Ed) Borasky
>
> http://www.borasky-research.net/
> http://borasky-research.blogspot.com/
>
> http://pdxneurosemantics.com
> http://pdx-sales-coach.com
> http://algocompsynth.com
>



From dakbari at gmail.com  Mon Oct 10 19:48:18 2005
From: dakbari at gmail.com (David Akbari)
Date: Mon, 10 Oct 2005 22:48:18 -0400
Subject: [CM] Re: [Cs-dev] Re: Common Music and Csound 5
In-Reply-To: <dccdd48f711d627a87977bb743d8543b@uiuc.edu>
References: <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net> <dccdd48f711d627a87977bb743d8543b@uiuc.edu>
Message-ID: <057573b77627ec70b3456706028e59bb@gmail.com>

On Oct 10, 2005, at 2:50 PM, Rick Taube wrote:
>
> another possibilty would be to connect Csound and CM using OSC via UDP 
> -- I think that OSC/UDP is working in CM right now, or if it isnt, it 
> will be very shortly (ie days not weeks)
>

IMHO, this produces the most efficient, least invasive solution.


-David



From arenato at genius.org.br  Tue Oct 11 09:41:47 2005
From: arenato at genius.org.br (arenato at genius.org.br)
Date: Tue, 11 Oct 2005 14:41:47 -0200
Subject: [CM] (no subject)
Message-ID: <OFC615F0AE.1D5D2B95-ON04257097.005B4D7D-83257097.005C85AA@genius.org.br>

Dear Juan Pamp?n,
I have tested in linux red hat 9 and solaris 9, with ATS-1.0-CMUCL-2 and 
CLM-2. The error is:

;; Loading #p"/opt/CLM/ATS-1.0-CLM-2/src/do-partials.sparcf".


Error in KERNEL:%COERCE-TO-FUNCTION:  the function DOUBLE-FLOAT is 
undefined.
   [Condition of type UNDEFINED-FUNCTION]

Restarts:
  0: [CONTINUE] Return NIL from load of "/opt/CLM/ATS-1.0-CLM-2/all.cl".
  1: [ABORT   ] Return to Top-Level.

Debug  (type H for help)

(KERNEL:%COERCE-TO-FUNCTION DOUBLE-FLOAT)
Source: Error finding source: 
Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no longer 
exists:
  target:code/fdefinition.lisp.

Thanks in advance

Alejandro


You should use the ATS sources for CMUCL (and CLM-2) that you can find at:
ftp://ccrma-ftp.stanford.edu/pub/Lisp/ATS/ATS-1.0-CMUCL.tar.gz
if you still have problems please send me a printout of the error and
I'll help you fix things.
Cheers,
JUAN

On 10/10/05, arenato at genius.org.br <arenato at genius.org.br> wrote:
>
> I have problems when I tried to compile ATS with CLM and CMUCL. The 
version
> of CMUCL is 19b. I tried with both CLM-3 and CLM-2.
> First,, during the compilation it?s necesary the library libXm.so.2, 
when my
> system have libXm.so.3 (Lesstif). I recompiled CMUCL with Lesstif 
version
> 2.2.3  without succes. Second, when I compile ATS inside lisp and CLM, 
there
> is a function no defined ( I don?t remember well, GET_FUN....) when 
compile
> "do_partials", and the compilation fail.
> Would you give me any suggestions about? Thanks in advance.
>
> Alejandro Renato
> Genius Insituto de Tecnolog?a
**************************************************************************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051011/18b48c7b/attachment.html>

From testcase at asu.edu  Tue Oct 11 10:04:06 2005
From: testcase at asu.edu (todd ingalls)
Date: Tue, 11 Oct 2005 10:04:06 -0700
Subject: [CM] Re: [Cs-dev] Re: Common Music and Csound 5
In-Reply-To: <057573b77627ec70b3456706028e59bb@gmail.com>
References:  <1550077.1128966183086.JavaMail.root@mswamui-billy.atl.sa.earthlink.net>
 <dccdd48f711d627a87977bb743d8543b@uiuc.edu>
 <057573b77627ec70b3456706028e59bb@gmail.com>
Message-ID: <d561989a59772c53b29a4eeb485a191a@asu.edu>

also,
would it not be possible to pipe the output to stdin of csound instead 
of a score file and get realtime that way? small modifications would 
have to be made as the time value of these events is different than in 
a score file


On Oct 10, 2005, at 7:48 PM, David Akbari wrote:

>
> On Oct 10, 2005, at 2:50 PM, Rick Taube wrote:
>>
>> another possibilty would be to connect Csound and CM using OSC via 
>> UDP -- I think that OSC/UDP is working in CM right now, or if it 
>> isnt, it will be very shortly (ie days not weeks)
>>
>
> IMHO, this produces the most efficient, least invasive solution.
>
>
> -David
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From heisters at 0x09.com  Wed Oct 12 06:56:58 2005
From: heisters at 0x09.com (Ian Smith-Heisters)
Date: Wed, 12 Oct 2005 09:56:58 -0400
Subject: [CM] SND in FLOSS audio article
Message-ID: <434D162A.7020305@0x09.com>

Hi all,

I did a brief overlook of 3 Linux audio apps at ArsTechnica.com. One of 
them is SND. Unfortunately, I don't have much experience with SND so 
far, so it's a bit light on meat, but it seems to have gotten some 
people interested.

Cheers,
Ian
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3170 bytes
Desc: S/MIME Cryptographic Signature
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051012/b1d921d1/attachment.bin>

From public at 0x09.com  Thu Oct 13 06:05:23 2005
From: public at 0x09.com (Ian Smith-Heisters)
Date: Thu, 13 Oct 2005 09:05:23 -0400
Subject: [CM] SND in FLOSS audio article
In-Reply-To: <434D162A.7020305@0x09.com>
References: <434D162A.7020305@0x09.com>
Message-ID: <434E5B93.20604@0x09.com>

I'm officially a TOTAL retard. here's the linkage 
http://arstechnica.com/guides/tweaks/linux-audio.ars

Sorry.

Ian Smith-Heisters wrote:
> Hi all,
> 
> I did a brief overlook of 3 Linux audio apps at ArsTechnica.com. One of 
> them is SND. Unfortunately, I don't have much experience with SND so 
> far, so it's a bit light on meat, but it seems to have gotten some 
> people interested.
> 
> Cheers,
> Ian



From dlphillips at woh.rr.com  Thu Oct 13 06:46:43 2005
From: dlphillips at woh.rr.com (Dave Phillips)
Date: Thu, 13 Oct 2005 09:46:43 -0400
Subject: [CM] SND in FLOSS audio article
In-Reply-To: <434E5B93.20604@0x09.com>
References: <434D162A.7020305@0x09.com> <434E5B93.20604@0x09.com>
Message-ID: <434E6543.2020109@woh.rr.com>

Hi Ian:

  Just a note to say thanks for the excellent article and for that 
stunning screenshot of Snd's OpenGL display. ;-)

  A very fine article, highly recommended.

Best regards,

Dave Phillips


Ian Smith-Heisters wrote:

> I'm officially a TOTAL retard. here's the linkage 
> http://arstechnica.com/guides/tweaks/linux-audio.ars
>
> Sorry.
>
> Ian Smith-Heisters wrote:
>
>> Hi all,
>>
>> I did a brief overlook of 3 Linux audio apps at ArsTechnica.com. One 
>> of them is SND. Unfortunately, I don't have much experience with SND 
>> so far, so it's a bit light on meat, but it seems to have gotten some 
>> people interested.
>>
>> Cheers,
>> Ian
>
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>



From bil at ccrma.Stanford.EDU  Thu Oct 13 07:08:25 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Thu, 13 Oct 2005 07:08:25 -0700
Subject: [CM] SND in FLOSS audio article
In-Reply-To: <434E5BEC.20207@0x09.com>
References: <434D162A.7020305@0x09.com> <20051012220025.M99521@ccrma.Stanford.EDU> <434E5BEC.20207@0x09.com>
Message-ID: <20051013140030.M90352@ccrma.Stanford.EDU>

I enjoyed your article!  On the interface: I've always thought a
tool should be invisible -- you work on something, and if the tool
is appropriate, you only think about the thing you're working on.

But... everyone and his brother has complained about Snd's interface, so I've
been working on it for several months, and have a "to-do" list
that will take several more to dig through.  I think the preferences
dialog will make it a lot easier to set up Snd.

The other comment I'd make is that I aim Snd/Clm at composers, not recording
engineers -- the emotional energy or direction comes from
way back in 1976 -- I was a grad student in composition, not
interested in electronic music, but Leland Smith (the composition
prof) insisted I take Chowning's computer music course; not only
was J.C. an inspiring person, but in the process I stumbled on
programming, and I can't describe the joy that brought me -- ever
since I've been coasting on that wave.  So, I really don't care
about other DAWs, as I guess is obvious.  I'd respond to the
complaints with: the damned thing is fully programmable; if
you want something different, and you've got anything at all 
on the ball, write an if-then statement, and you will be free.

(Also, the rt stuff was all the work of Kjetil Matheussen -- I wish he
could get the credit in your article).



From bil at ccrma.Stanford.EDU  Thu Oct 13 11:44:11 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Thu, 13 Oct 2005 11:44:11 -0700
Subject: [CM] snd 7.16
Message-ID: <20051013184026.M3814@ccrma.Stanford.EDU>

Snd 7.16

Marten Stromberg provided a PDF version of snd.html, but I can't decide
  where it should go.

new dialog: first half (Motif version) of a preferences dialog (Options menu item).

added cycle speed style via click, mix|track-speed-style
      analyze-ladspa as synonym for analyse-ladspa
      save/insert/mix buttons in region browser
      save-region-dialog
      --enable-threads configure switch, and first use thereof (FIR filtering)
      snd-g|xprefs.c, preferences-dialog function

snd-error is now (throw 'snd-error ...)

init file handling has become more complex: ~/.snd is still loaded if present,
  as is /etc/snd.conf, but Snd also now looks for ~/.snd_guile|ruby, and
  /etc/snd_guile|ruby.conf, and the preferences dialog and
  Options:save options function write to ~/.snd_prefs_guile|ruby
  (so that the user's init file is not changed by the preference mechanism).
  This may seem unmotivated, but watch this space!

changed recorder-in|out-format -> recorder-in|out-data-format,
        recorder-out-type -> recorder-out-header-type
        save-macros now requires a filename (it won't write to ~/.snd unless you ask it to)


Checked: gtk 2.8.3|4|5|6, FreeBSD 5.4, gsl 1.7, Ruby 1.8.3

With much help from: Anders Vinjar, Kjetil Matheussen, Mike Scholz, Marten Stromberg



From plutek at infinity.net  Mon Oct 17 09:59:46 2005
From: plutek at infinity.net (Peter Lutek)
Date: Mon, 17 Oct 2005 12:59:46 -0400
Subject: [CM] SND - cursor follow playback?
Message-ID: <4353D882.90101@infinity.net>

holding ctrl while clicking the play button makes the cursor follow 
playback, but from the beginning of the file.

am i missing something, or is there in fact no way to make the cursor 
follow playback when playing from somewhere other than the beginning 
(i.e. play from mark or play selection)?

thanks!
-p



From bil at ccrma.Stanford.EDU  Mon Oct 17 10:13:45 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Mon, 17 Oct 2005 10:13:45 -0700
Subject: [CM] SND - cursor follow playback?
In-Reply-To: <4353D882.90101@infinity.net>
References: <4353D882.90101@infinity.net>
Message-ID: <20051017171232.M5309@ccrma.Stanford.EDU>

see cursor-follows-play in extsnd.html, and also play ("pfc"). 



From taube at uiuc.edu  Mon Oct 17 11:26:06 2005
From: taube at uiuc.edu (Rick Taube)
Date: Mon, 17 Oct 2005 13:26:06 -0500
Subject: [CM] sndplay on tiger?
In-Reply-To: <20051017171232.M5309@ccrma.Stanford.EDU>
References: <4353D882.90101@infinity.net> <20051017171232.M5309@ccrma.Stanford.EDU>
Message-ID: <701483eea365841792b57facb2a81728@uiuc.edu>

I just installed clm-3 on all 15 new g5s in my lab, works great except 
for sndplay. (mus-audio-describe) is returning NIL without printing 
anything. The only thing I can think of is that the Lab is running 
Tiger (10.4) while my machines at home and in my office run 10.3.9 
where sndplay works, even out my motu.
we are using m-audio firewire interfaces in the lab and these are 
selected in the usual way in core audio. Ive tried  switching to the 
internal speakers but im not getting sound out there either from 
sndplay.  so im wondering if its an issue with sndplay and Tiger rather 
than sndplay and the m-audio.  Is anyone running clm-3 on tiger with 
sound output working?

  
  



From bil at ccrma.Stanford.EDU  Mon Oct 17 11:59:39 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Mon, 17 Oct 2005 11:59:39 -0700
Subject: [CM] sndplay on tiger?
In-Reply-To: <701483eea365841792b57facb2a81728@uiuc.edu>
References: <4353D882.90101@infinity.net> <20051017171232.M5309@ccrma.Stanford.EDU> <701483eea365841792b57facb2a81728@uiuc.edu>
Message-ID: <20051017185245.M7145@ccrma.Stanford.EDU>

I just tried clm + ccl on an Imac (G5 I think -- can't keep these numbers straight),
OSX 10.4, built-in audio, and sndplay seems ok (it stutters sometimes).  I have
an emagic 2|6 (or is it 6|2?) in a box somewhere -- I know it worked earlier, but
I never connected it to this machine.  On this setup, (mus-audio-describe) prints
out the usual junk.  I don't think I changed anything in audio.c in the OSX section.

Now that I look back at it, that code was written in "type-as-fast-as-possible"
mode -- I'll add some error messages so there's some chance of finding out
what has gone wrong. 



From taube at uiuc.edu  Mon Oct 17 12:42:14 2005
From: taube at uiuc.edu (Rick Taube)
Date: Mon, 17 Oct 2005 14:42:14 -0500
Subject: [CM] sndplay on tiger?
In-Reply-To: <20051017185245.M7145@ccrma.Stanford.EDU>
References: <4353D882.90101@infinity.net> <20051017171232.M5309@ccrma.Stanford.EDU> <701483eea365841792b57facb2a81728@uiuc.edu> <20051017185245.M7145@ccrma.Stanford.EDU>
Message-ID: <8dd91c758ed12f5fb2b6697e97ea6569@uiuc.edu>

ok i see maybe part of the problem is we compiiled clm on an "install 
machine" and then distributed that file system out to the clients so 
mybe the install machine wasnt good for the ./configure step.

  so i downloaded  a new tarball, built on the local macine and sndplay 
partly works now

the remaining problem is that when sndplay outputs to the M-AUDIO it 
somehow flips the srate and channel settings on the device as it starts 
playing the file so we hear things at srate 96000 and only in 1 
channel. wierd!!

here is the mus-audio state before and then after playing a sound file 
--  second listing has the octave up:


? (mus-audio-describe)
found 2 audio devices
   Apple: Built-in Audio (default output)
     srate: 44100, chans: 2, bits/sample: 32, format: lpcm, buf: 4096
     flags: float big-endian packed
     vols: master: 0.712
     This device supports 6 formats:
       srate: 48000, chans: 2, bits/sample: 32, format: lpcm
       srate: 44100, chans: 2, bits/sample: 32, format: lpcm
       srate: 32000, chans: 2, bits/sample: 32, format: lpcm
       srate: 48000, chans: 2, bits/sample: 16, format: cac3
       srate: 44100, chans: 2, bits/sample: 16, format: cac3
       srate: 32000, chans: 2, bits/sample: 16, format: cac3

   M-Audio: FireWire Solo Multichannel (output)
     srate: 44100, chans: 4, bits/sample: 32, format: lpcm, buf: 16384
     flags: float big-endian packed
     vols: 1.000 1.000 1.000 1.000
     This device supports 5 formats:
       srate: 96000, chans: 4, bits/sample: 32, format: lpcm
       srate: 88200, chans: 4, bits/sample: 32, format: lpcm
       srate: 48000, chans: 4, bits/sample: 32, format: lpcm
       srate: 44100, chans: 4, bits/sample: 32, format: lpcm
       srate: 48000, chans: 2, bits/sample: 16, format: cac3

   Apple: Built-in Audio (default input)
     srate: 44100, chans: 2, bits/sample: 32, format: lpcm, buf: 4096
     flags: float big-endian packed
     vols: 0.000 0.000
     This device supports 3 formats:
       srate: 48000, chans: 2, bits/sample: 32, format: lpcm
       srate: 44100, chans: 2, bits/sample: 32, format: lpcm
       srate: 32000, chans: 2, bits/sample: 32, format: lpcm

   M-Audio: FireWire Solo Multichannel (input)
     srate: 44100, chans: 4, bits/sample: 32, format: lpcm, buf: 16384
     flags: float big-endian packed
     vols: 1.000 1.000 1.000 1.000
     This device supports 4 formats:
       srate: 96000, chans: 4, bits/sample: 32, format: lpcm
       srate: 88200, chans: 4, bits/sample: 32, format: lpcm
       srate: 48000, chans: 4, bits/sample: 32, format: lpcm
       srate: 44100, chans: 4, bits/sample: 32, format: lpcm
NIL
? (dac)
"test.aiff"
? (mus-audio-describe)
found 2 audio devices
   Apple: Built-in Audio (default output)
     srate: 44100, chans: 2, bits/sample: 32, format: lpcm, buf: 4096
     flags: float big-endian packed
     vols: master: 0.712
     This device supports 6 formats:
       srate: 48000, chans: 2, bits/sample: 32, format: lpcm
       srate: 44100, chans: 2, bits/sample: 32, format: lpcm
       srate: 32000, chans: 2, bits/sample: 32, format: lpcm
       srate: 48000, chans: 2, bits/sample: 16, format: cac3
       srate: 44100, chans: 2, bits/sample: 16, format: cac3
       srate: 32000, chans: 2, bits/sample: 16, format: cac3

   M-Audio: FireWire Solo Multichannel (output)
     srate: 96000, chans: 4, bits/sample: 32, format: lpcm, buf: 16384
     flags: float big-endian packed
     vols: 1.000 1.000 1.000 1.000
     This device supports 5 formats:
       srate: 96000, chans: 4, bits/sample: 32, format: lpcm
       srate: 88200, chans: 4, bits/sample: 32, format: lpcm
       srate: 48000, chans: 4, bits/sample: 32, format: lpcm
       srate: 44100, chans: 4, bits/sample: 32, format: lpcm
       srate: 48000, chans: 2, bits/sample: 16, format: cac3

   Apple: Built-in Audio (default input)
     srate: 44100, chans: 2, bits/sample: 32, format: lpcm, buf: 4096
     flags: float big-endian packed
     vols: 0.000 0.000
     This device supports 3 formats:
       srate: 48000, chans: 2, bits/sample: 32, format: lpcm
       srate: 44100, chans: 2, bits/sample: 32, format: lpcm
       srate: 32000, chans: 2, bits/sample: 32, format: lpcm

   M-Audio: FireWire Solo Multichannel (input)
     srate: 96000, chans: 4, bits/sample: 32, format: lpcm, buf: 16384
     flags: float big-endian packed
     vols: 1.000 1.000 1.000 1.000
     This device supports 4 formats:
       srate: 96000, chans: 4, bits/sample: 32, format: lpcm
       srate: 88200, chans: 4, bits/sample: 32, format: lpcm
       srate: 48000, chans: 4, bits/sample: 32, format: lpcm
       srate: 44100, chans: 4, bits/sample: 32, format: lpcm
NIL
?




From k.gebbert at gmail.com  Tue Oct 18 07:43:51 2005
From: k.gebbert at gmail.com (karlos)
Date: Tue, 18 Oct 2005 15:43:51 +0100
Subject: [CM] snd-ls does not get build
Message-ID: <4fdd16f70510180743i1aa5b639ta3df0c5605a0e569@mail.gmail.com>

Hi,
 I am fairly, er, very new to snd and I have the follwoing problems with it.
I downloaded the snd-ls package and build it and, after looking at the
scheme-script, and the output in the Terminal, it should have been build
properly. Even the install script runs fine, at least when one believes
whats posted in the terminal. Yet, there is no snd-ls binary I can locate,
neither does seem the build to be any different from the ones with unpatched
source I did before.
Furthermore, with the normal source, not snd-ls, there is a problem with the
xg.so module, which should have been build but isn't. Do I get it right? Is
it a dependancy I was not aware of? Anyway, it is not there, and most, if
not all of the stuff I tried to load was not working. Can anyone sense
something here ) )) ))) ?
 I am on Gentoo, and I think I have met all the build-requirements (using
Guile btw.), otherwise it would not have been build at all, I suppose.
 Thanks,
 Karsten
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051018/86f7847d/attachment.html>

From plessas at mur.at  Tue Oct 18 06:11:31 2005
From: plessas at mur.at (Peter Plessas)
Date: Tue, 18 Oct 2005 15:11:31 +0200
Subject: [CM] snd compilation error
Message-ID: <4354F483.3060502@mur.at>

Hi List,

I am here on a debian testing powerpc box, trying to compile snd-7 from 
the daily tarball, getting the following error:

gcc -c -DLOCALEDIR=\"/usr/local/share/locale\" -DHAVE_CONFIG_H   -I. -g 
-O2       snd-xen.c
snd-xen.c: In function 'snd_catch_scm_error':
snd-xen.c:368: error: 'scm_the_last_stack_fluid' undeclared (first use 
in this function)
snd-xen.c:368: error: (Each undeclared identifier is reported only once
snd-xen.c:368: error: for each function it appears in.)

If you have any idea what this could be, i'd be happy to hear!

Looking forward to using a recent snd soon!

regards,
Peter Plessas

PS: please CC me, as i am not yet subscribed to this mailing list.



From kjetil at ccrma.Stanford.EDU  Tue Oct 18 12:31:32 2005
From: kjetil at ccrma.Stanford.EDU (Kjetil S. Matheussen)
Date: Tue, 18 Oct 2005 12:31:32 -0700 (PDT)
Subject: [CM] snd-ls does not get build
In-Reply-To: <20051018190003.18805.86649.Mailman@cm-mail.stanford.edu>
References: <20051018190003.18805.86649.Mailman@cm-mail.stanford.edu>
Message-ID: <Pine.LNX.4.61.0510181224590.1063@ttleush.stanford.edu>

karlos:
>
> Hi,
> I am fairly, er, very new to snd and I have the follwoing problems with it=
> .
> I downloaded the snd-ls package and build it and, after looking at the
> scheme-script, and the output in the Terminal, it should have been build
> properly. Even the install script runs fine, at least when one believes
> whats posted in the terminal. Yet, there is no snd-ls binary I can locate,
> neither does seem the build to be any different from the ones with unpatche=
> d
> source I did before.

It should be located here: /usr/local/bin/snd-ls .
And in case it is, perhaps you are missing /usr/local/bin in your path?


> Furthermore, with the normal source, not snd-ls, there is a problem with th=
> e
> xg.so module, which should have been build but isn't. Do I get it right? Is
> it a dependancy I was not aware of? Anyway, it is not there, and most, if
> not all of the stuff I tried to load was not working. Can anyone sense
> something here ) )) ))) ?

The easiest thing is probably to specify --with-static-xg when 
configuring. That should solve it.

You can also write "make xg.so", but then you have to link xg.so at 
run-time.



-- 



From bil at ccrma.Stanford.EDU  Tue Oct 18 14:15:40 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Tue, 18 Oct 2005 14:15:40 -0700
Subject: [CM] snd compilation error
In-Reply-To: <4354F483.3060502@mur.at>
References: <4354F483.3060502@mur.at>
Message-ID: <20051018211248.M87673@ccrma.Stanford.EDU>

scm_the_last_stack_fluid is in guile 1.3.4 and guile 1.4; after that its name
changed.  Apparently the configure script thinks you have an old guile,
but the compiler is finding a newer one.  If it's not obvious from config.log,
send me config.h and config.log and maybe I'll see what's wrong. 




From rundall at uiuc.edu  Tue Oct 18 17:31:05 2005
From: rundall at uiuc.edu (Jake Rundall)
Date: Tue, 18 Oct 2005 19:31:05 -0500
Subject: [CM] 4-channel playback in clm-3 (OS X)
Message-ID: <6a01cc3c.4d6cccd8.1802b700@expms3.cites.uiuc.edu>

Hi,

I'm trying to create a 4-channel piece in clm-3.  I'm running CLM and using 
openmcl in OS X.  I've used Rick Taube's vkey instrument to create 4-channel 
.aiff files, but I'm not sure how to play them back, because dac (sndplay I guess) 
doesn't seem to handle this type of file.  Here are a few possible options that I'm 
looking for comments and suggestions on:

1) Processing the file to create four mono sound files, one for each channel, and 
playing in another program.  Pro Tools would do it, but I'm looking for 
something less expensive, or perhaps freeware....

2) Working with ambisonics format... I've heard that there is a supercollider 
patch that can play this type of format.  How does one write an ambisonics file?  
Is there a specific ambisonics file format?  Would you use dlocsig to write the 
file?  Does this work with clm-3?

3) Any other options that I haven't mentioned?  If you have other suggestions 
about how to work with 4-channel sound files in clm-3 (on OS X), I'd be really 
happy to hear about them.

Thanks,

Jake Rundall



From rm at seid-online.de  Wed Oct 19 05:15:10 2005
From: rm at seid-online.de (R. Mattes)
Date: Wed, 19 Oct 2005 14:15:10 +0200
Subject: [CM] Recent SBCL problems
Message-ID: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>

Just a note for SBCL users that don't follow the sbcl-devel mailing
list: Christophe Rhodes posted a "bandaid" fix (unfortunately not as a
follow-up to Rick Taube's bugreport) that seems to cure the type
inference problems with recent sbcls. Here's a copy of it:


(in-package "SB-KERNEL")
(defun round-numeric-bound (x class format up-p)
  (if x
      (let ((cx (if (consp x) (car x) x)))
        (ecase class
          ((nil rational) x)
          (integer
           (if (and (consp x) (integerp cx))
               (if up-p (1+ cx) (1- cx))
               (if up-p (ceiling cx) (floor cx))))
          (float
           (let ((res
                  (cond
                    ((and format (subtypep format 'double-float))
                     (if (<= most-negative-double-float cx
most-positive-double-float)
                         (coerce cx format)
                         nil))
                    (t
                     (if (<= most-negative-single-float cx
most-positive-single-float)
                         ;; FIXME
                         (coerce cx (or format 'single-float))
                         nil)))))
             (if (consp x) (list res) res)))))
      nil))

This needs to be loaded _before_ compilation of cm/cmn.

 HTH Ralf Mattes 



From bil at ccrma.Stanford.EDU  Wed Oct 19 07:52:29 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Wed, 19 Oct 2005 07:52:29 -0700
Subject: [CM] Recent SBCL problems
In-Reply-To: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
Message-ID: <20051019145149.M94480@ccrma.Stanford.EDU>

That bugfix also makes it possible to use sbcl with clm -- thanks!




From rundall at uiuc.edu  Wed Oct 19 10:13:58 2005
From: rundall at uiuc.edu (Jake Rundall)
Date: Wed, 19 Oct 2005 12:13:58 -0500
Subject: [CM] Re: 4-channel playback in clm-3 (OS X)
Message-ID: <6c7824fd.4dc89d15.19f15800@expms3.cites.uiuc.edu>

Hi,

I just wanted to say that I figured out what I was doing wrong with the 4-
channel playback in clm-3.  It turns out that dac/sndplay does indeed work with 
4-channel .aiff -- I simply didn't have my MOTU hooked up as I was testing it.  I 
was operating on the incorrect assumption that dac would play the file even if 
my computer output was not set up to do it.   I also didn' realize that dac was 
trying to let me know that this was the problem.  Anyway, everything's working 
perfectly.  

Incidentally, I'm still interested in learning about the option of working with 
ambisonics.  So, if anyone can let me know where to go to start learning about 
this, please let me know.  Specifically, What kind of file format does 
ambisonics use? and How does one write an ambisonics file?  How does one play 
back an ambisonics file?

Thanks, and sorry for my confusion.

Jake Rundall

---- Original message ----
>Date: Tue, 18 Oct 2005 19:31:05 -0500
>From: Jake Rundall <rundall at uiuc.edu>  
>Subject: 4-channel playback in clm-3 (OS X)  
>To: cmdist at ccrma.stanford.edu
>
>Hi,
>
>I'm trying to create a 4-channel piece in clm-3.  I'm running CLM and using 
>openmcl in OS X.  I've used Rick Taube's vkey instrument to create 4-channel 
>.aiff files, but I'm not sure how to play them back, because dac (sndplay I 
guess) 
>doesn't seem to handle this type of file.  Here are a few possible options that 
I'm 
>looking for comments and suggestions on:
>
>1) Processing the file to create four mono sound files, one for each channel, 
and 
>playing in another program.  Pro Tools would do it, but I'm looking for 
>something less expensive, or perhaps freeware....
>
>2) Working with ambisonics format... I've heard that there is a supercollider 
>patch that can play this type of format.  How does one write an ambisonics 
file?  
>Is there a specific ambisonics file format?  Would you use dlocsig to write the 
>file?  Does this work with clm-3?
>
>3) Any other options that I haven't mentioned?  If you have other suggestions 
>about how to work with 4-channel sound files in clm-3 (on OS X), I'd be really 
>happy to hear about them.
>
>Thanks,
>
>Jake Rundall



From taube at uiuc.edu  Wed Oct 19 11:25:08 2005
From: taube at uiuc.edu (Rick Taube)
Date: Wed, 19 Oct 2005 13:25:08 -0500
Subject: [CM] Recent SBCL problems
In-Reply-To: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
Message-ID: <17526feedaed505722e7ea0be1dbaf14@uiuc.edu>

yes thanks! that seems to fix cm compiling now too. unfortunately 
callbacks still dont seem to work, even when i do (setq 
sb-alien::*values-are-okay* t) so i am still unable to get gtk to run 
in sbcl. sigh. this was all working beautifully a year ago...

On Oct 19, 2005, at 7:15 AM, R. Mattes wrote:

> Just a note for SBCL users that don't follow the sbcl-devel mailing
> list: Christophe Rhodes posted a "bandaid" fix (unfortunately not as a
> follow-up to Rick Taube's bugreport) that seems to cure the type
> inference problems with recent sbcls. Here's a copy of it:



From HerbM at learnquick.com  Wed Oct 19 12:07:32 2005
From: HerbM at learnquick.com (Herb Martin)
Date: Wed, 19 Oct 2005 14:07:32 -0500
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
Message-ID: <EIOMFSM-000218-LU@mail2.learnquick.com>

On attempting to (load "all.lisp") the errors included below are displayed.


My suspicion is that something is not yet defined for CLisp
but I see nothing CLisp or GNU specific (that I am supposed
to do) so far.

Is there are more extensive guide than the included readme.clm?

I am currently reading the all.lisp in an attempt to discover more info.

VCVars32.bat has been run to set the Lib, include, and path environment for
MSC (VS Studio.NET),
and I copied the windows specific .h file to the config.h.

I would NOT be sending an error report or request for help, quite yet,
except the readme does indicate that such are requested if the compile does
not just work....

If this is not the appropriate forum please direct me to the correct
location (if you know);

RTFM is welcome if it is accompanied by a title or source and preferably a
page number or 
section. <grin>

=======================
[1]> (load "all.lisp")
;; Loading file all.lisp ...
;   using existing configuration file config.h
; Compiling "E:\\dev\\CLMusic\\clm-3\\io.c"
; Compiling "E:\\dev\\CLMusic\\clm-3\\headers.c"
; Compiling "E:\\dev\\CLMusic\\clm-3\\audio.c"
; Compiling "E:\\dev\\CLMusic\\clm-3\\sound.c"
; Compiling "E:\\dev\\CLMusic\\clm-3\\clm.c"
; Compiling "E:\\dev\\CLMusic\\clm-3\\cmus.c"

*** - EVAL: undefined function COMPILE-AND-LOAD
The following restarts are available:
USE-VALUE      :R1      You may input a value to be used instead of
(FDEFINITION
 'COMPILE-AND-LOAD).
RETRY          :R2      Retry
STORE-VALUE    :R3      You may input a new value for (FDEFINITION
'COMPILE-AND-
LOAD).
SKIP           :R4      skip (IF # #)
STOP           :R5      stop loading file E:\dev\CLMusic\clm-3\all.lisp
ABORT          :R6      ABORT
Break 1 [2]>

--
Herb Martin



From bil at ccrma.Stanford.EDU  Wed Oct 19 12:13:12 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Wed, 19 Oct 2005 12:13:12 -0700
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <EIOMFSM-000218-LU@mail2.learnquick.com>
References: <EIOMFSM-000218-LU@mail2.learnquick.com>
Message-ID: <20051019191137.M60295@ccrma.Stanford.EDU>

Clisp is no longer supported in CLM.  It's a long unhappy story... The Clisp
developers seem to have finally taken an interest in foreign code lately,
so maybe I'll find time, or someone else will, to write the necessary FFI
links.  Sorry!



From HerbM at learnquick.com  Wed Oct 19 12:33:27 2005
From: HerbM at learnquick.com (Herb Martin)
Date: Wed, 19 Oct 2005 14:33:27 -0500
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <20051019191137.M60295@ccrma.Stanford.EDU>
Message-ID: <EIOMH00-0002YO-JP@mail2.learnquick.com>

> From: Bill Schottstaedt [mailto:bil at ccrma.Stanford.EDU] 
> Clisp is no longer supported in CLM.  It's a long unhappy 
> story... The Clisp developers seem to have finally taken an 
> interest in foreign code lately, so maybe I'll find time, or 
> someone else will, to write the necessary FFI links.  Sorry!
> 

Is there are free/open Lisp for Windows that is supported?

No apology necessary -- it's kind of you to provide
free software and free help.

My intention is to create some pre-school learning and
play (is there a difference) software and your music
libraries looked like a great addition.

I am also looking at DRScheme (since having trouble
with CLM loading) but my original intention was to
stay in Common Lisp.

--
Herb Martin



From geb_a at yahoo.com  Wed Oct 19 12:50:26 2005
From: geb_a at yahoo.com (geb a)
Date: Wed, 19 Oct 2005 12:50:26 -0700 (PDT)
Subject: [CM] snd and mzscheme
Message-ID: <20051019195026.32849.qmail@web50901.mail.yahoo.com>

Dear Sir,

I am a high school teacher using mzscheme for teaching
purposes.  Does SND work with mzscheme and are there
installation instructions for mzscheme available?

Thank you for your time.

Sincerely,

Dan Anderson
Viewpoint


		
__________________________________ 
Start your day with Yahoo! - Make it your home page! 
http://www.yahoo.com/r/hs



From bil at ccrma.Stanford.EDU  Wed Oct 19 13:46:41 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Wed, 19 Oct 2005 13:46:41 -0700
Subject: [CM] snd and mzscheme
In-Reply-To: <20051019195026.32849.qmail@web50901.mail.yahoo.com>
References: <20051019195026.32849.qmail@web50901.mail.yahoo.com>
Message-ID: <20051019203755.M40765@ccrma.Stanford.EDU>

Currently, the only Scheme supported by Snd is Guile.  There's no technical reason
that mzscheme is not supported; I even have the macros to do it somewhere,
but people sometimes rub each other the wrong way... (I felt the mzscheme
people "dissed" my work for no reason at all, so I tore out all the mzscheme
code -- if someone else wants to add it, I won't stand in the way).

(You could probably still find the exchange in their mailing list archive,
complete with shocked "how can you use such language, there are
children here", which was true, though not the way they meant.)



From k.gebbert at gmail.com  Wed Oct 19 14:11:08 2005
From: k.gebbert at gmail.com (karlos)
Date: Wed, 19 Oct 2005 21:11:08 +0000
Subject: [CM] Re: Cmdist digest, Vol 1 #808 - 7 msgs
In-Reply-To: <20051019190002.6021.51392.Mailman@cm-mail.stanford.edu>
References: <20051019190002.6021.51392.Mailman@cm-mail.stanford.edu>
Message-ID: <4fdd16f70510191411p39325d48oa4f5ed71f9693df4@mail.gmail.com>

yes, I had a look, and there is ./snd, but without the dash-ls. it also
seems that, when running this one, nothing in particular changes to the
other, non-ls, packages I compiled. I also tried to do: make xg.so

the output after ./configure with various option (including
--with-static-xg)

make: *** No rule to make target `xg.so'. Stop.

hmm..
is there an newer version than this available -->

snd-ls-0.9.5.3.tar.gz
<http://www.notam02.no/arkiv/src/snd/snd-ls-0.9.5.3.tar.gz>

maybe is this the problem?

thanks,

karsten
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051019/45faf3f5/attachment.html>

From taube at uiuc.edu  Wed Oct 19 14:24:06 2005
From: taube at uiuc.edu (Rick Taube)
Date: Wed, 19 Oct 2005 16:24:06 -0500
Subject: [CM] Re: CLisp 2.35
In-Reply-To: <Pine.LNX.4.63.0510191555110.13181@chinstrap.tjhsst.edu>
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de> <17526feedaed505722e7ea0be1dbaf14@uiuc.edu> <Pine.LNX.4.63.0510191555110.13181@chinstrap.tjhsst.edu>
Message-ID: <e3a995157a92aad0bf4ea111dc77f394@uiuc.edu>

Hi, you might try forcing the version using cm.sh's -V arg:

	cm.sh -V 2.35 -l clisp

for more tricks see: cm.sh -h

note that clisp-2.35 works fine on osx:

galen:/Lisp/cm hkt$ cm -l clisp
[1]>
; Installation directory: "/Lisp/cm/"
; Loading "bin/clisp_2.35_darwin-powerpc/clm-stubs.fas"
; Loading "bin/clisp_2.35_darwin-powerpc/cmn-stubs.fas"
[...]



On Oct 19, 2005, at 2:59 PM, RDLatimer wrote:

> I'm on Debian, have just installed Clisp current version 2.35,
> but Common Music isn't able to see this version number.
>
> $ ./bin/cm.sh
> [Fatal  ]  /usr/bin/clisp: version '()' unsupported.
> [Notice ]  Need clisp with -repl option (version 2.31 or higher).
> Aborting.
>
>
> I think this most current version of CLisp isn't reporting a version 
> number.  I've emailed them about this:
>
>  (lisp-implementation-version)
> "() (built 3335644808) (memory 3338735517)"
>
>
> Is there a way to let cm.sh know it's okay to continue?  I think this 
> version of Clisp is okay, it's just not returning a version number.
>
> Thanks, Randy Latimer



From taube at uiuc.edu  Wed Oct 19 14:26:36 2005
From: taube at uiuc.edu (Rick Taube)
Date: Wed, 19 Oct 2005 16:26:36 -0500
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <20051019191137.M60295@ccrma.Stanford.EDU>
References: <EIOMFSM-000218-LU@mail2.learnquick.com> <20051019191137.M60295@ccrma.Stanford.EDU>
Message-ID: <90a82f0649880c946836e6f6ec5e30b1@uiuc.edu>

if anyone decides to tackle this this should try using CFFI - that way 
clm would run pretty much everywhere and bill could get out of the ffi 
business altoghter.

http://common-lisp.net/project/cffi/



On Oct 19, 2005, at 2:13 PM, Bill Schottstaedt wrote:

>
> Clisp is no longer supported in CLM.  It's a long unhappy story... The 
> Clisp
> developers seem to have finally taken an interest in foreign code 
> lately,
> so maybe I'll find time, or someone else will, to write the necessary 
> FFI
> links.  Sorry!
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From testcase at asu.edu  Wed Oct 19 16:49:01 2005
From: testcase at asu.edu (todd ingalls)
Date: Wed, 19 Oct 2005 16:49:01 -0700
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <90a82f0649880c946836e6f6ec5e30b1@uiuc.edu>
References: <EIOMFSM-000218-LU@mail2.learnquick.com>
 <20051019191137.M60295@ccrma.Stanford.EDU>
 <90a82f0649880c946836e6f6ec5e30b1@uiuc.edu>
Message-ID: <2231f15f2337af8cb2b108acc397aa29@asu.edu>

another thing of interest as well is verrazano 
http://common-lisp.net/project/fetter/ (or fetter, i am not straight on 
the correct name) for auto-generating the CFFI bindings. I have been 
looking at this for generating bindings myself.

in just a few minutes i was able to generate CFFI bindings for sndlib.h 
and clm.h that look mostly reasonable. However, I can't speak to the 
complete accuracy and reliability, but I would be willing to help out 
anyone who wanted to look into this more.



On Oct 19, 2005, at 2:26 PM, Rick Taube wrote:

> if anyone decides to tackle this this should try using CFFI - that way 
> clm would run pretty much everywhere and bill could get out of the ffi 
> business altoghter.
>
> http://common-lisp.net/project/cffi/
>
>
>
> On Oct 19, 2005, at 2:13 PM, Bill Schottstaedt wrote:
>
>>
>> Clisp is no longer supported in CLM.  It's a long unhappy story... 
>> The Clisp
>> developers seem to have finally taken an interest in foreign code 
>> lately,
>> so maybe I'll find time, or someone else will, to write the necessary 
>> FFI
>> links.  Sorry!
>>
>> _______________________________________________
>> Cmdist mailing list
>> Cmdist at ccrma.stanford.edu
>> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From testcase at asu.edu  Wed Oct 19 17:20:04 2005
From: testcase at asu.edu (todd ingalls)
Date: Wed, 19 Oct 2005 17:20:04 -0700
Subject: [CM] SuperCollider realtime support in cm
Message-ID: <409b2623ee1300d4f745cba8262d8978@asu.edu>

Hello,
People may have seen this already, but I just wanted to officially 
announce realtime supercollider support in realtime. There are many 
additional in the cm dictionary for the supercollider topic and there 
are also some examples in etc/example/rt-sc.cm . Things seem to work 
basically as I expect, however, I would really love feedback from 
people about how things are working and also ideas for improvement and 
making things clearer.


todd



From HerbM at learnquick.com  Wed Oct 19 18:44:09 2005
From: HerbM at learnquick.com (Herb Martin)
Date: Wed, 19 Oct 2005 20:44:09 -0500
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <2231f15f2337af8cb2b108acc397aa29@asu.edu>
Message-ID: <EIOMY5P-00024C-1E@mail2.learnquick.com>

todd ingalls
> Sent: Wednesday, October 19, 2005 6:49 PM
> To: cm list
> Subject: Re: [CM] CLM-3 Compile-load on Windows (XP) Clisp 
> 2.35 C++ from VS.Net
> 
> another thing of interest as well is verrazano 
> http://common-lisp.net/project/fetter/ (or fetter, i am not 
> straight on the correct name) for auto-generating the CFFI 
> bindings. I have been looking at this for generating bindings myself.
> 
> in just a few minutes i was able to generate CFFI bindings 
> for sndlib.h and clm.h that look mostly reasonable. However, 
> I can't speak to the complete accuracy and reliability, but I 
> would be willing to help out anyone who wanted to look into this more.

It does look very interesting.  Thanks to you and
everyone who posts such leads/references.

This one does have a lot of pre-requisites and that
is usually a concern on Windows when trying to arrange
such ports:

<http://common-lisp.net/project/fetter/getting-started.html> 

ASDF: http://www.cliki.net/asdf/ 
S-XML: http://common-lisp.net/project/s-xml/ 
SPLIT-SEQENCE: http://www.cliki.net/SPLIT-SEQUENCE/ 
CFFI: http://common-lisp.net/project/cffi/ 
GCC-XML: http://www.gccxml.org/ 


--
Herb Martin
	



From gogins at pipeline.com  Wed Oct 19 18:58:20 2005
From: gogins at pipeline.com (Michael Gogins)
Date: Wed, 19 Oct 2005 21:58:20 -0400
Subject: [CM] FFI for Csound 5
References: <EIOMFSM-000218-LU@mail2.learnquick.com> <20051019191137.M60295@ccrma.Stanford.EDU> <90a82f0649880c946836e6f6ec5e30b1@uiuc.edu> <2231f15f2337af8cb2b108acc397aa29@asu.edu>
Message-ID: <000d01c5d519$bf15e8f0$6701a8c0@Generator>

I have created an experimental CLisp FFI file for Csound 5, which is in 
SourceForge CVS for csound5. You can exercise this by installing the current 
version of Csound 5. In the csound5 directory, execute "clisp 
examples/test.lisp".

I am just a beginner at this, and I plan to replace the CLisp FFI with CFFI. 
Still, I would appreciate it if some of you would look over the test.lisp 
and csound.lisp files and give me some feedback.

Csound obviously works well enough with CM as it is. The direct interface 
will not at first add a lot, though it should in the long run.

The main improvement in the short run will be the ability to run Csound in a 
separate thread, stop it in mid-rendering, and restart it. It should also be 
possible to start Csound performing at any point in the score. It will also 
be possible to embed Csound instrument definitions in the LISP code and to 
otherwise build up CSD files.

I suppose the next step would be a csd.lisp along the lines of sco.lisp, 
with support for the direct interface, stopping and restarting, and seeking 
in the score. I'm not a good enough LISP programmer right now to do this, 
though.

It is a problem for me that my main platform is Windows (that might change, 
but not right now), because CLisp does not appear to support threads, so I 
can't test the stopping and re-starting, etc. Is there another Windows LISP 
that runs CM that supports threads?

Thanks for any help you can provide,
Mike 




From HerbM at learnquick.com  Wed Oct 19 21:09:59 2005
From: HerbM at learnquick.com (Herb Martin)
Date: Wed, 19 Oct 2005 23:09:59 -0500
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <93d818a442d94b2cd376be45fb1c550a@asu.edu>
Message-ID: <EION4WR-0001TG-MG@mail2.learnquick.com>

> and i found it also needs parse-number-1.0
> 
> but again, this is only to produce the bindings, once the 
> initial bindings are created nothing but CFFI is needed.

So if I had to run it on a Linux (maybe another
Posix system like CygWin) the bindings could
be produced and copied over to the Win32 system
where they would likely work as well?

That's good to know too.  And I added the parse-number
to my notes as another likely requirement.

Thanks.

--
Herb Martin

> -----Original Message-----
> From: todd ingalls [mailto:testcase at asu.edu] 
> Sent: Wednesday, October 19, 2005 9:31 PM
> To: Herb Martin
> Cc: Cmdist at ccrma.Stanford.EDU
> Subject: Re: [CM] CLM-3 Compile-load on Windows (XP) Clisp 
> 2.35 C++ from VS.Net
> 
> 
> 
> On Oct 19, 2005, at 6:44 PM, Herb Martin wrote:
> 
> > todd ingalls
> >> Sent: Wednesday, October 19, 2005 6:49 PM
> >> To: cm list
> >> Subject: Re: [CM] CLM-3 Compile-load on Windows (XP) Clisp
> >> 2.35 C++ from VS.Net
> >>
> >> another thing of interest as well is verrazano
> >> http://common-lisp.net/project/fetter/ (or fetter, i am not
> >> straight on the correct name) for auto-generating the CFFI
> >> bindings. I have been looking at this for generating 
> bindings myself.
> >>
> >> in just a few minutes i was able to generate CFFI bindings
> >> for sndlib.h and clm.h that look mostly reasonable. However,
> >> I can't speak to the complete accuracy and reliability, but I
> >> would be willing to help out anyone who wanted to look 
> into this more.
> >
> > It does look very interesting.  Thanks to you and
> > everyone who posts such leads/references.
> >
> > This one does have a lot of pre-requisites and that
> > is usually a concern on Windows when trying to arrange
> > such ports:
> >
> > <http://common-lisp.net/project/fetter/getting-started.html>
> >
> > ASDF: http://www.cliki.net/asdf/
> > S-XML: http://common-lisp.net/project/s-xml/
> > SPLIT-SEQENCE: http://www.cliki.net/SPLIT-SEQUENCE/
> > CFFI: http://common-lisp.net/project/cffi/
> > GCC-XML: http://www.gccxml.org/
> >
> >
> > --
> > Herb Martin
> > 	
> >
> > _______________________________________________
> > Cmdist mailing list
> > Cmdist at ccrma.stanford.edu
> > http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
> 



From immanuell at enfocus.be  Thu Oct 20 01:16:27 2005
From: immanuell at enfocus.be (Immanuel Litzroth)
Date: Thu, 20 Oct 2005 09:16:27 +0100
Subject: [CM] Recent SBCL problems
In-Reply-To: <17526feedaed505722e7ea0be1dbaf14@uiuc.edu> (Rick Taube's
 message of "Wed, 19 Oct 2005 13:25:08 -0500")
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
	<17526feedaed505722e7ea0be1dbaf14@uiuc.edu>
Message-ID: <ek6gbo2s.fsf@immanuelinux.site>

"Rick Taube" <taube at uiuc.edu> writes:

> yes thanks! that seems to fix cm compiling now too. unfortunately
> callbacks still dont seem to work, even when i do (setq
> sb-alien::*values-are-okay* t) so i am still unable to get gtk to run
> in sbcl. sigh. this was all working beautifully a year ago...

Are you using a recent cvs? Callbacks are supposed to be working (since
the beginning of this week I think)! I have not checked this myself but
I ran the cffi testsuite and all tests passed.
Cheers
Immanuel



From rm at seid-online.de  Thu Oct 20 01:23:20 2005
From: rm at seid-online.de (R. Mattes)
Date: Thu, 20 Oct 2005 10:23:20 +0200
Subject: [CM] Recent SBCL problems
In-Reply-To: <17526feedaed505722e7ea0be1dbaf14@uiuc.edu>
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
	 <17526feedaed505722e7ea0be1dbaf14@uiuc.edu>
Message-ID: <1129796600.22740.1.camel@hobbes.mh-freiburg.de>

On Wed, 2005-10-19 at 13:25 -0500, Rick Taube wrote:
> yes thanks! that seems to fix cm compiling now too.

Yes, my local setup finally works again.

>  unfortunately 
> callbacks still dont seem to work, even when i do (setq 
> sb-alien::*values-are-okay* t) so i am still unable to get gtk to run 
> in sbcl. sigh. this was all working beautifully a year ago...

Sigh, i know. Callbacks never worked on Linux/PPC unfortunately.
I guess that's the price one has to pay far a fast developing Lisp.
 

Cheers Ralf Mattes

> On Oct 19, 2005, at 7:15 AM, R. Mattes wrote:
> 
> > Just a note for SBCL users that don't follow the sbcl-devel mailing
> > list: Christophe Rhodes posted a "bandaid" fix (unfortunately not as a
> > follow-up to Rick Taube's bugreport) that seems to cure the type
> > inference problems with recent sbcls. Here's a copy of it:
> 
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From andersvi at extern.uio.no  Thu Oct 20 01:53:53 2005
From: andersvi at extern.uio.no (andersvi at extern.uio.no)
Date: Thu, 20 Oct 2005 10:53:53 +0200
Subject: [CM] portmidi reset
In-Reply-To: <90a82f0649880c946836e6f6ec5e30b1@uiuc.edu> (Rick Taube's
 message of "Wed, 19 Oct 2005 16:26:36 -0500")
References: <EIOMFSM-000218-LU@mail2.learnquick.com>
	<20051019191137.M60295@ccrma.Stanford.EDU>
	<90a82f0649880c946836e6f6ec5e30b1@uiuc.edu>
Message-ID: <cop4q7c4li6.fsf_-_@luciano.localdomain>

Ive searched abit in portmidi.lisp without feeling lucky.  Is
there a way to have the CM-portmidi interface update with newly
added devices after cm is booted?

-anders



From mlang at delysid.org  Thu Oct 20 03:26:45 2005
From: mlang at delysid.org (Mario Lang)
Date: Thu, 20 Oct 2005 12:26:45 +0200
Subject: [CM] portmidi reset
In-Reply-To: <cop4q7c4li6.fsf_-_@luciano.localdomain>
	(andersvi@extern.uio.no's message of "Thu, 20 Oct 2005 10:53:53
	+0200")
References: <EIOMFSM-000218-LU@mail2.learnquick.com>
	<20051019191137.M60295@ccrma.Stanford.EDU>
	<90a82f0649880c946836e6f6ec5e30b1@uiuc.edu>
	<cop4q7c4li6.fsf_-_@luciano.localdomain>
Message-ID: <87wtk8iivu.fsf@lexx.delysid.org>

andersvi at extern.uio.no writes:

> Ive searched abit in portmidi.lisp without feeling lucky.  Is
> there a way to have the CM-portmidi interface update with newly
> added devices after cm is booted?

I remembert reading somewhere in the docs that it is not
possible (i.e., documented deficiency).  However, it would
be useful, I've felt a need for it several times as well.

-- 
CYa,
  Mario



From taube at uiuc.edu  Thu Oct 20 04:14:29 2005
From: taube at uiuc.edu (Rick Taube)
Date: Thu, 20 Oct 2005 06:14:29 -0500
Subject: [CM] portmidi reset
In-Reply-To: <cop4q7c4li6.fsf_-_@luciano.localdomain>
References: <EIOMFSM-000218-LU@mail2.learnquick.com> <20051019191137.M60295@ccrma.Stanford.EDU> <90a82f0649880c946836e6f6ec5e30b1@uiuc.edu> <cop4q7c4li6.fsf_-_@luciano.localdomain>
Message-ID: <8446ed7665004c0a4a62a4488ff9a09a@uiuc.edu>

Hi Anders, it seems that Portmidi reads descriptors only when pm_init 
is called. their docs say:
     The returned structure is owned by the PortMidi implementation and 
must
     not be manipulated or freed. The pointer is guaranteed to be valid
     between calls to Pm_Initialize() and Pm_Terminate().

reading that, it might be possible in lisp to call pm:terminate and 
then pm:initialize again:

	(portmidi-close)
	(pm:terminate)

	(pm:initialize)

you may have to set the global pm:*portmidi* to T as well.



On Oct 20, 2005, at 3:53 AM, andersvi at extern.uio.no wrote:

> Ive searched abit in portmidi.lisp without feeling lucky.  Is
> there a way to have the CM-portmidi interface update with newly
> added devices after cm is booted?
>
> -anders
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From taube at uiuc.edu  Thu Oct 20 04:18:33 2005
From: taube at uiuc.edu (Rick Taube)
Date: Thu, 20 Oct 2005 06:18:33 -0500
Subject: [CM] Recent SBCL problems
In-Reply-To: <ek6gbo2s.fsf@immanuelinux.site>
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de> <17526feedaed505722e7ea0be1dbaf14@uiuc.edu> <ek6gbo2s.fsf@immanuelinux.site>
Message-ID: <73b8e90dc777c2e7725daa0e2f6daff4@uiuc.edu>

No, ive been using the binary they released for 0.9.5. can i send you 
my editied gtkffi-cmusbcl.lisp to see if gtk will compile and load in 
your version? you would simply drop the file into src/gui/ , then build 
the system and then try to open a window...




On Oct 20, 2005, at 3:16 AM, Immanuel Litzroth wrote:

> "Rick Taube" <taube at uiuc.edu> writes:
>
>> yes thanks! that seems to fix cm compiling now too. unfortunately
>> callbacks still dont seem to work, even when i do (setq
>> sb-alien::*values-are-okay* t) so i am still unable to get gtk to run
>> in sbcl. sigh. this was all working beautifully a year ago...
>
> Are you using a recent cvs? Callbacks are supposed to be working (since
> the beginning of this week I think)! I have not checked this myself but
> I ran the cffi testsuite and all tests passed.
> Cheers
> Immanuel
>



From testcase at asu.edu  Wed Oct 19 19:30:34 2005
From: testcase at asu.edu (todd ingalls)
Date: Wed, 19 Oct 2005 19:30:34 -0700
Subject: [CM] CLM-3 Compile-load on Windows (XP) Clisp 2.35 C++ from VS.Net
In-Reply-To: <EIOMY5P-00024C-1E@mail2.learnquick.com>
References: <EIOMY5P-00024C-1E@mail2.learnquick.com>
Message-ID: <93d818a442d94b2cd376be45fb1c550a@asu.edu>

and i found it also needs parse-number-1.0

but again, this is only to produce the bindings, once the initial 
bindings are created nothing but CFFI is needed.


On Oct 19, 2005, at 6:44 PM, Herb Martin wrote:

> todd ingalls
>> Sent: Wednesday, October 19, 2005 6:49 PM
>> To: cm list
>> Subject: Re: [CM] CLM-3 Compile-load on Windows (XP) Clisp
>> 2.35 C++ from VS.Net
>>
>> another thing of interest as well is verrazano
>> http://common-lisp.net/project/fetter/ (or fetter, i am not
>> straight on the correct name) for auto-generating the CFFI
>> bindings. I have been looking at this for generating bindings myself.
>>
>> in just a few minutes i was able to generate CFFI bindings
>> for sndlib.h and clm.h that look mostly reasonable. However,
>> I can't speak to the complete accuracy and reliability, but I
>> would be willing to help out anyone who wanted to look into this more.
>
> It does look very interesting.  Thanks to you and
> everyone who posts such leads/references.
>
> This one does have a lot of pre-requisites and that
> is usually a concern on Windows when trying to arrange
> such ports:
>
> <http://common-lisp.net/project/fetter/getting-started.html>
>
> ASDF: http://www.cliki.net/asdf/
> S-XML: http://common-lisp.net/project/s-xml/
> SPLIT-SEQENCE: http://www.cliki.net/SPLIT-SEQUENCE/
> CFFI: http://common-lisp.net/project/cffi/
> GCC-XML: http://www.gccxml.org/
>
>
> --
> Herb Martin
> 	
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From immanuell at enfocus.be  Thu Oct 20 05:58:20 2005
From: immanuell at enfocus.be (Immanuel Litzroth)
Date: Thu, 20 Oct 2005 13:58:20 +0100
Subject: [CM] Recent SBCL problems
In-Reply-To: <73b8e90dc777c2e7725daa0e2f6daff4@uiuc.edu> (Rick Taube's
 message of "Thu, 20 Oct 2005 06:18:33 -0500")
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
	<17526feedaed505722e7ea0be1dbaf14@uiuc.edu>
	<ek6gbo2s.fsf@immanuelinux.site>
	<73b8e90dc777c2e7725daa0e2f6daff4@uiuc.edu>
Message-ID: <3bmwny4z.fsf@immanuelinux.site>

"Rick Taube" <taube at uiuc.edu> writes:

> No, ive been using the binary they released for 0.9.5. can i send you
> my editied gtkffi-cmusbcl.lisp to see if gtk will compile and load in
> your version? you would simply drop the file into src/gui/ , then
> build the system and then try to open a window...

No problem, just send it and i'll try to do it today.
GRTZ
Immanuel



From juanig at ccrma.Stanford.EDU  Thu Oct 20 09:23:29 2005
From: juanig at ccrma.Stanford.EDU (Juan Reyes)
Date: Thu, 20 Oct 2005 09:23:29 -0700
Subject: [CM] Configure options for Snd on osx-10.4
Message-ID: <1129825409.13164.6.camel@strawberri.maginvent.org>

 I was able to build Snd with Motif and Gtk on OS-X-10.4 
(fink) with the following configure options:

MOTIF VERSION:
=============

  $ ./configure CFLAGS=-I/sw/include LDFLAGS=-L/sw/lib 
GUILE_CONFIG_path=/sw/bin/
--disable-nls --with-guile --with-motif --with-static-xm 
--with-motif-prefix=/sw


GTK Version
===========

  $ ./configure CFLAGS=-I/sw/include LDFLAGS=-L/sw/lib 
GUILE_CONFIG_path=/sw/bin --disable-nls --with-guile --with-gtk 
--with-static-xm --with-builtin-gtkrc


Perhaps the only caveat (AFAIK) is the fftw compilation which takes a
really long time.


The web page at Users at PlanetCCRMA has been updated with this
information.

  --* Juan Reyes



From kjetil at ccrma.Stanford.EDU  Thu Oct 20 15:02:45 2005
From: kjetil at ccrma.Stanford.EDU (Kjetil Svalastog Matheussen)
Date: Thu, 20 Oct 2005 15:02:45 -0700 (PDT)
Subject: [CM] Problems installing snd-ls / xg.so
In-Reply-To: <20051020082401.5880.76545.Mailman@cm-mail.stanford.edu>
References: <20051020082401.5880.76545.Mailman@cm-mail.stanford.edu>
Message-ID: <Pine.LNX.4.63.0510201458280.11843@cmn27.stanford.edu>

karlos:
>
> yes, I had a look, and there is ./snd, but without the dash-ls. it also
> seems that, when running this one, nothing in particular changes to the

I don't understand whats wrong... Is there a directory called 
/usr/local/snd-ls ?

And in case there is, the following command should start snd-ls:

/usr/local/snd-ls/snd-7/snd -noglob -noinit -l /usr/local/snd-ls/init.scm


> other, non-ls, packages I compiled. I also tried to do: make xg.so
>
> the output after ./configure with various option (including
> --with-static-xg)
>
> make: *** No rule to make target `xg.so'. Stop.
>

I don't thing this should happen...



> hmm..
> is there an newer version than this available -->
>
> snd-ls-0.9.5.3.tar.gz
> <http://www.notam02.no/arkiv/src/snd/snd-ls-0.9.5.3.tar.gz>
>

No, thats the latest version. I've installed on many machines.


-- 



From csr21 at cam.ac.uk  Fri Oct 21 06:23:03 2005
From: csr21 at cam.ac.uk (Christophe Rhodes)
Date: Fri, 21 Oct 2005 14:23:03 +0100
Subject: [CM] Recent SBCL problems
In-Reply-To: <17526feedaed505722e7ea0be1dbaf14@uiuc.edu> (Rick Taube's message
	of "Wed, 19 Oct 2005 13:25:08 -0500")
References: <1129724111.24589.8.camel@hobbes.mh-freiburg.de>
	<17526feedaed505722e7ea0be1dbaf14@uiuc.edu>
Message-ID: <sq64rrngw8.fsf@cam.ac.uk>

Rick Taube <taube at uiuc.edu> writes:

> yes thanks! that seems to fix cm compiling now too.

In the light of this, I committed my band-aid fix to sbcl CVS (so that
sbcl-0.9.6 will contain it when it is released).

> unfortunately callbacks still dont seem to work, even when i do
> (setq sb-alien::*values-are-okay* t) so i am still unable to get gtk
> to run in sbcl. sigh. this was all working beautifully a year ago...

Could you post a problem report about this to the sbcl-devel mailing
list, if you haven't already?

Thanks,

Christophe



From howell at ucsc.edu  Fri Oct 21 13:27:07 2005
From: howell at ucsc.edu (David Cope)
Date: Fri, 21 Oct 2005 13:27:07 -0700
Subject: [CM] WACM
Message-ID: <p04320403bf7eff02dff0@[128.114.11.162]>



Dear Colleague,

This summer, for the fourth year, we'll be offering the Workshop in 
Algorithmic Computer Music (WACM) at UC Santa Cruz. Student 
evaluations from past years have been uniformly positive and we're 
excited about doing it again.

One of the things the students benefit most from is limiting the 
workshop to fifteen students, which allows for a very high 
student/teacher ratio. But it also means that people applying later 
in the year may not get in.

So I'm sending you this email to remind you about the workshop and to 
give you an opportunity to apply before our main publicity goes out. 
I won't barrage you with details here (a general blurb appears 
below), but if you're interested you can visit the website, which is 
at:

http://summer.ucsc.edu/wacm

... and please don't hesitate to contact me if you have any 
questions. If you know anyone else who might benefit from the 
workshop, please feel free to forward this message.

All best,

Dave Cope


The fourth Workshop in Algorithmic Computer Music (WACM) will take 
place from June 26 through July 9, 2006 at the University of 
California, Santa Cruz (UCSC). The workshop will be held at UCSC's 
Music Center, which features state of the art facilities as well as 
sweeping vistas of the Monterey Bay. David Cope, Paul Nauert, Peter 
Elsea, and Soren Goodman will be among those on hand to teach and 
advise workshop attendees. Participants will take classes on the 
basic techniques of algorithmic composition and algorithmic music 
analysis, learning and using the computer programming language Lisp. 
Participants will create three significant software projects: a 
Markov-based rules program, a genetic algorithm, and software modeled 
on the Experiments in Musical Intelligence program. Music analysis 
software and techniques will also be covered in depth. Many 
compositional approaches will be discussed in detail including 
rules-based techniques, data-driven models, genetic algorithms, 
neural networks, fuzzy logic, mathematical modeling, and 
sonification. Software programs such as Common Music, Max, Open 
Music, and so on will also be presented and used.

The workshop is limited to 15 participants to ensure an excellent 
student/teacher ratio..

The first two days of the workshop are optional and will cover basic 
Lisp programming. Those individuals with extensive programming 
experience and/or prior experience using Lisp may not need to attend 
these first two days.

Each participant will receive a class reference manual, class 
software in source code, and other printed and software tools and 
guides relevant to the workshop's goals. Access to the world class 
computer music facilities at UCSC will allow performances and special 
lectures by invited guests. The workshop includes five daily hours of 
individual computer time (with an advisor close at hand). Each week 
will culminate in performances of participant works and presentations 
of analysis and composition software. By the workshop's end, each 
participant will have gained facility with Lisp, have written 
compositional and analytical software programs, and be presented with 
a CD of their compositions.

Prospective students must have an ability to read music and 
understand basic music theory as well as have at least a basic 
facility with computers (text editing, etc.). Students need not be 
active musicians or programmers. Room and board is available on 
campus if desired. 5 quarter units of credit is available.
-- 

To obtain Experiments in Musical Intelligence and other music by 
David Cope go to http://www.spectrumpress.com.

For information on the fourth Workshop on Algorithmic Computer Music 
(WACM) go to http://summer.ucsc.edu/wacm/.

David Cope
   Professor, Music Department
     Music Center, University of California, Santa Cruz
     Santa Cruz, CA 95064
     831 459-3417
     msgs: 831 459 2804
   fax: 831 459-5584
   website: http://arts.ucsc.edu/faculty/cope
   email:  howell at ucsc.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051021/1e617ff7/attachment.html>

From nando at ccrma.Stanford.EDU  Fri Oct 21 17:00:18 2005
From: nando at ccrma.Stanford.EDU (Fernando Lopez-Lezcano)
Date: Fri, 21 Oct 2005 17:00:18 -0700
Subject: [CM] Re: 4-channel playback in clm-3 (OS X)
In-Reply-To: <6c7824fd.4dc89d15.19f15800@expms3.cites.uiuc.edu>
References: <6c7824fd.4dc89d15.19f15800@expms3.cites.uiuc.edu>
Message-ID: <1129939218.5001.20.camel@cmn3.stanford.edu>

On Wed, 2005-10-19 at 12:13 -0500, Jake Rundall wrote:
> Incidentally, I'm still interested in learning about the option of working with 
> ambisonics. 

The "dlocsig" ug can output B format Ambisonics and decoded Ambisonics. 

B Format is going to be a 4 channel soundfile (this is for first order
Ambisonics) that contains an omnidirection component and x y and z
components of the wave front. It cannot be played directly. It can be
decoded to play back in n-channel setups, both 2d and 3d. 

> So, if anyone can let me know where to go to start learning about 
> this, please let me know. 

  http://www.york.ac.uk/inst/mustech/3d_audio/welcome.html

> Specifically, What kind of file format does 
> ambisonics use? and How does one write an ambisonics file?  How does one play 
> back an ambisonics file?

CLM3 changed internally (from CLM2) and dlocsig could not longer be
implemented as a unit generator using a single structure to contain all
the internals of the ug (as the built in ug's are). For my 220b class I
did a more compact version that only needs passing around three, it is
here:

http://ccrma.stanford.edu/courses/220b/topics/spatialization/examples/dlocsig.lisp

The matching example move-sound macro shows some usage:
http://ccrma.stanford.edu/courses/220b/topics/spatialization/examples/move-sound.ins

A custom fm-violin that uses dlocsig:
http://ccrma.stanford.edu/courses/220b/topics/spatialization/examples/vd.ins

This is the default mode for dlocsig:
  (defparameter dlocsig-render-using amplitude-panning)
[just amplitude panning using vbap]

You can render to 4 channel Ambisonics format with:

  (defparameter dlocsig-render-using b-format-ambisonics)

Then you have to decode the file separately to play it. For 2d only
Ambisonics you need a minimum of 4 speakers in a square arrangement.

Or you can generate and decode Ambisonics on the fly with:

  (defparameter dlocsig-render-using decoded-ambisonics)

What it does depends on the speaker arrangement that is used (which in
turn depends on the number of output channels being used in with-sound
and the value of dlocsig-3d, "t" or "nil"). 

I know, complicated... :-)
-- Fernando




From emres at bilgi.edu.tr  Sun Oct 23 16:04:28 2005
From: emres at bilgi.edu.tr (Emre Sevinc)
Date: Mon, 24 Oct 2005 02:04:28 +0300
Subject: [CM] Common Music, Taube's book, etc.
Message-ID: <9940BA60F8E6414F9F0307DD411BCF6906BF9356@Ex2k.bilgi.networks>


I've just had my copy of Notes from the Metalevel
and started reading it. 

I also got the 2.7 version of Common Music
from CVS a few days ago. It compiled fine except
lgtk part. (Yes I'm using SBCL, Debian GNU/Linux). 

Do I miss a lot if I continue using Common Music
without lgtk related graphical interface? What
does it provide?

It's a strange feeling being able to use Scheme
syntax on SBCL :)

I still didn't come to the point of actually producing
sound, .midi files, .mp3 files, etc. I don't know
which of the software is more important or in what order
I should download and install them (what kind of dependency
exist, a dependency graph talkin' about which part is needed
for which functionality would be look good on Common Music's
or ccrma's web page).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051024/37f697b1/attachment.html>

From taube at uiuc.edu  Wed Oct 26 12:23:08 2005
From: taube at uiuc.edu (Rick Taube)
Date: Wed, 26 Oct 2005 14:23:08 -0500
Subject: [CM] clm permission issue?
Message-ID: <5fb6e12f7f5accf3a863e7060f4029ea@uiuc.edu>

if I install clm-3 logged in as "admin"  then when a different user 
trys to load all.lisp there is a strage error tha happens -- i cant 
figure how how or why rename-file is geting called!  I think I rememebr 
that installng/running as different users worked in the past.

camil5:/Local/lisp/clm-3 hkt$ su admin
Password:
camil5:/Local/lisp/clm-3 admin$ openmcl
Welcome to OpenMCL Version (Beta: Darwin) 0.14.3!
? (load "all")
;   using existing configuration file config.h
;Compiler warnings for "/Local/lisp/clm-3/ffi.lisp" :
;   Unused lexical variable DEV, in SL-DAC-1.
;   Unused lexical variable NAME, in SL-DAC-1.
;Compiler warnings :
;   Unused lexical variable DEV, in SL-DAC-1.
;   Unused lexical variable NAME, in SL-DAC-1.
;Compiler warnings for "/Local/lisp/clm-3/sound.lisp" :
;   Unused lexical variable CALLS, in SOUND-LET.
;Compiler warnings :
;   Unused lexical variable CALLS, in SOUND-LET.
;Compiler warnings for "/Local/lisp/clm-3/env.lisp" :
;   Unused lexical variable XX, in ENVELOPE-REPEAT.
;Compiler warnings :
;   Unused lexical variable XX, in ENVELOPE-REPEAT.
#P"/Local/lisp/clm-3/all.lisp"

? (quit)
camil5:/Local/lisp/clm-3 admin$ exit
camil5:/Local/lisp/clm-3 hkt$ whoami
hkt
camil5:/Local/lisp/clm-3 hkt$ openmcl
Welcome to OpenMCL Version (Beta: Darwin) 0.14.3!
? (load "all")
;   using existing configuration file config.h
 > Error in process listener(1): Failed to rename 
/Local/lisp/clm-3/walk.dfsl to /Local/lisp/clm-3/3339328444.tem: -13
 > While executing: RENAME-FILE
 > Type :POP to abort.
Type :? for other options.
1 > (quit)
camil5:/Local/lisp/clm-3 hkt$



From renueden at earthlink.net  Thu Oct 27 22:55:09 2005
From: renueden at earthlink.net (Ken)
Date: Thu, 27 Oct 2005 22:55:09 -0700
Subject: [CM] which lisp
Message-ID: <4361BD3D.9000307@earthlink.net>

so which lisp is the focus for realtime wok right now.  after a brief 
absence, i'm ready to push forward.  i'm on linux BTW, fedora core 3 at 
the moment if it matters.



From taube at uiuc.edu  Fri Oct 28 04:50:07 2005
From: taube at uiuc.edu (Rick Taube)
Date: Fri, 28 Oct 2005 06:50:07 -0500
Subject: [CM] which lisp
In-Reply-To: <4361BD3D.9000307@earthlink.net>
References: <4361BD3D.9000307@earthlink.net>
Message-ID: <88516a5fb05c988f97f579246842333c@uiuc.edu>

on linux sbcl is your best bet. use the latest version and let us know 
what problems you encounter.

On Oct 28, 2005, at 12:55 AM, Ken wrote:

> so which lisp is the focus for realtime wok right now.  after a brief 
> absence, i'm ready to push forward.  i'm on linux BTW, fedora core 3 
> at the moment if it matters.
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From emres at bilgi.edu.tr  Fri Oct 28 12:41:13 2005
From: emres at bilgi.edu.tr (Emre Sevinc)
Date: Fri, 28 Oct 2005 22:41:13 +0300
Subject: [CM] About the implementation of sturms function in Notes from the Metalevel
Message-ID: <9940BA60F8E6414F9F0307DD411BCF6906BF935F@Ex2k.bilgi.networks>

I came to the 12. Chapter of Notes from the Metalevel
book and while I was trying to compile the sturms function
I had some warning from SBCL:

The details are at:

http://paste.lisp.org/display/12980

;; the code below (which is taken from Common Music
;; application produces an error:
;;
;;-+  Warnings (1)
;; `-- ==>
;;       -1
;;     This is not a (OR (SINGLE-FLOAT (0.0)) (DOUBLE-FLOAT (0.0d0)) (RATIONAL (0))):
;;       -1
;;     See also:
;;       SBCL Manual, Handling of Types [node]


(define (strums key1 key2 rate dur amp)
  (let ((step (if (< key2 key1) -1 1)) ;; e.g. 1 2 gives no warn.
        (diff (abs (- key1 key2))))
    (loop repeat (+ diff 1)
          for key from key1 by step
          for beg from 0 by rate
          collect (new midi
                    :time beg
                    :duration dur
                    :amplitude amp
                    :keynum key))))



One of the possible solutions was to use a slightly different for in
the loop:


; this seems to solve the problem of
;; a potential *negative* stepping value

(define (strums key1 key2 rate dur amp)
  (let ((step (if (< key2 key1) -1 1))
        (diff (abs (- key1 key2))))
    (loop repeat (+ diff 1)
          for key = key1 then (+ key step)
          for beg from 0 by rate
          collect (new midi
                    :time beg
                    :duration dur
                    :amplitude amp
                    :keynum key))))

Another proposed solution is to to do a conditional rotatef of key1 and key2 
and then do a for key from key1 to key2.

PS: Maybe this is already corrected in the CD that came with the
book (I couldn't find scores.cm in CVS version of CM) but I'm just the type of 
guy who likes to type code from the book and I just wanted to share this
just in case somebody else is stuck in the same place.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20051028/6d383dee/attachment.html>

From bil at ccrma.Stanford.EDU  Fri Oct 28 13:11:29 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Fri, 28 Oct 2005 13:11:29 -0700
Subject: [CM] About the implementation of sturms function in Notes from the Metalevel
In-Reply-To: <9940BA60F8E6414F9F0307DD411BCF6906BF935F@Ex2k.bilgi.networks>
References: <9940BA60F8E6414F9F0307DD411BCF6906BF935F@Ex2k.bilgi.networks>
Message-ID: <20051028200453.M90606@ccrma.Stanford.EDU>

sbcl is picky about the step variable -- I complained about
it a long time ago, and was told that it was a feature.  Here's an even better sbcl-ism:

The value 0.0
             is not of type
               (OR (SINGLE-FLOAT (0.0)) (DOUBLE-FLOAT (0.0d0)) (RATIONAL (0)))

which happens in cmn! 



From csr21 at cam.ac.uk  Fri Oct 28 14:52:10 2005
From: csr21 at cam.ac.uk (Christophe Rhodes)
Date: Fri, 28 Oct 2005 22:52:10 +0100
Subject: [CM] About the implementation of sturms function in Notes from the Metalevel
In-Reply-To: <20051028200453.M90606@ccrma.Stanford.EDU> (Bill Schottstaedt's
	message of "Fri, 28 Oct 2005 13:11:29 -0700")
References: <9940BA60F8E6414F9F0307DD411BCF6906BF935F@Ex2k.bilgi.networks>
	<20051028200453.M90606@ccrma.Stanford.EDU>
Message-ID: <sqirvhb985.fsf@cam.ac.uk>

"Bill Schottstaedt" <bil at ccrma.Stanford.EDU> writes:

> sbcl is picky about the step variable -- I complained about
> it a long time ago, and was told that it was a feature.  Here's an even better sbcl-ism:
>
> The value 0.0
>              is not of type
>                (OR (SINGLE-FLOAT (0.0)) (DOUBLE-FLOAT (0.0d0)) (RATIONAL (0)))
>
> which happens in cmn! 

The step variable is defined to be a positive real, in the ANSI
standard, section 6.1.2.1.1:

  by

    The loop keyword by marks the increment or decrement supplied by
    form3. The value of form3 can be any positive number. The default
    value is 1.

Neither 0.0 nor -1 is a positive number.  The type
  (or (single-float (0.0)) (double-float (0.0d0)) (rational (0)))
precisely means "positive number".  I'm sorry that you don't find this
helpful.

Cheers,

Christophe



From taube at uiuc.edu  Sat Oct 29 05:04:20 2005
From: taube at uiuc.edu (Rick Taube)
Date: Sat, 29 Oct 2005 07:04:20 -0500
Subject: [CM] About the implementation of sturms function in Notes from the Metalevel
In-Reply-To: <9940BA60F8E6414F9F0307DD411BCF6906BF935F@Ex2k.bilgi.networks>
References: <9940BA60F8E6414F9F0307DD411BCF6906BF935F@Ex2k.bilgi.networks>
Message-ID: <71ec1848d3127d7b70cf42ff24f5e3fe@uiuc.edu>

> I came to the 12. Chapter of Notes from the Metalevel
>  book and while I was trying to compile the sturms function
>  I had some warning from SBCL:

Emre - thank you for taking the effort to note the problem and provide 
a solution. that was not a very robust example! i debated about adding 
error checks to code but decided against it to keep things as simple as 
possible.

it would be great if you would keep track of any/all problems you note 
as you work your way through the book, then we could then post the 
collection for others to take advantage of. There are other issue (like 
playback) that have arisen due to changes in more recent releases of 
the software.



>  The details are at:
>
> http://paste.lisp.org/display/12980
>
>  ;; the code below (which is taken from Common Music
>  ;; application produces an error:
>  ;;
>  ;;-+? Warnings (1)
>  ;; `-- ==>
>  ;;?????? -1
>  ;;???? This is not a (OR (SINGLE-FLOAT (0.0)) (DOUBLE-FLOAT (0.0d0)) 
> (RATIONAL (0))):
>  ;;?????? -1
>  ;;???? See also:
>  ;;?????? SBCL Manual, Handling of Types [node]
>
>
>  (define (strums key1 key2 rate dur amp)
>  ? (let ((step (if (< key2 key1) -1 1)) ;; e.g. 1 2 gives no warn.
>  ??????? (diff (abs (- key1 key2))))
>  ??? (loop repeat (+ diff 1)
>  ????????? for key from key1 by step
>  ????????? for beg from 0 by rate
>  ????????? collect (new midi
>  ??????????????????? :time beg
>  ??????????????????? :duration dur
>  ??????????????????? :amplitude amp
>  ??????????????????? :keynum key))))
>
>
>
>  One of the possible solutions was to use a slightly different for in
>  the loop:
>
>
>  ; this seems to solve the problem of
>  ;; a potential *negative* stepping value
>
>  (define (strums key1 key2 rate dur amp)
>  ? (let ((step (if (< key2 key1) -1 1))
>  ??????? (diff (abs (- key1 key2))))
>  ??? (loop repeat (+ diff 1)
>  ????????? for key = key1 then (+ key step)
>  ????????? for beg from 0 by rate
>  ????????? collect (new midi
>  ??????????????????? :time beg
>  ??????????????????? :duration dur
>  ??????????????????? :amplitude amp
>  ??????????????????? :keynum key))))
>
>  Another proposed solution is to to do a conditional rotatef of key1 
> and key2
>  and then do a for key from key1 to key2.
>
>  PS: Maybe this is already corrected in the CD that came with the
>  book (I couldn't find scores.cm in CVS version of CM) but I'm just 
> the type of
>  guy who likes to type code from the book and I just wanted to share 
> this
>  just in case somebody else is stuck in the same place.
>
>



From kjetil at ccrma.Stanford.EDU  Sat Oct 29 22:56:21 2005
From: kjetil at ccrma.Stanford.EDU (Kjetil S. Matheussen)
Date: Sat, 29 Oct 2005 22:56:21 -0700 (PDT)
Subject: [CM] which lisp
In-Reply-To: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu>
Message-ID: <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu>

Depends what kind of realtime work.
http://www.notam02.no/arkiv/doc/snd-rt/


Rick Taube:
>
> on linux sbcl is your best bet. use the latest version and let us know
> what problems you encounter.
>
> On Oct 28, 2005, at 12:55 AM, Ken wrote:
>
>> so which lisp is the focus for realtime wok right now.  after a brief
>> absence, i'm ready to push forward.  i'm on linux BTW, fedora core 3
>> at the moment if it matters.
>>




From taube at uiuc.edu  Sun Oct 30 03:58:40 2005
From: taube at uiuc.edu (Rick Taube)
Date: Sun, 30 Oct 2005 05:58:40 -0600
Subject: [CM] which lisp
In-Reply-To: <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu> <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu>
Message-ID: <c7c5ffa07aeaaec33bf63c726e8637c0@uiuc.edu>

Indeed! and my apologies for the incomplete answer.
--rick

On Oct 30, 2005, at 12:56 AM, Kjetil S. Matheussen wrote:

>
> Depends what kind of realtime work.
> http://www.notam02.no/arkiv/doc/snd-rt/
>



From michael at klingbeil.com  Sun Oct 30 07:10:33 2005
From: michael at klingbeil.com (Michael Klingbeil)
Date: Sun, 30 Oct 2005 11:10:33 -0400
Subject: [CM] which lisp
In-Reply-To: <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu>
 <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu>
Message-ID: <p0611040ebf8a90be11b5@[10.255.254.15]>

Isn't snd-rt for Scheme (and Guile) only? If so I guess the answer to 
"which Lisp" is... um... Guile??!

Now that I think about it, maybe Scheme is the best Lisp after all!  ;)


Michael

>Depends what kind of realtime work.
>http://www.notam02.no/arkiv/doc/snd-rt/
>
>
>Rick Taube:
>>
>>on linux sbcl is your best bet. use the latest version and let us know
>>what problems you encounter.
>>
>>On Oct 28, 2005, at 12:55 AM, Ken wrote:
>>
>>>so which lisp is the focus for realtime wok right now.  after a brief
>>>absence, i'm ready to push forward.  i'm on linux BTW, fedora core 3
>>>at the moment if it matters.
>>>
>
>
>_______________________________________________
>Cmdist mailing list
>Cmdist at ccrma.stanford.edu
>http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist



From taube at uiuc.edu  Sun Oct 30 09:00:11 2005
From: taube at uiuc.edu (Rick Taube)
Date: Sun, 30 Oct 2005 11:00:11 -0600
Subject: [CM] which lisp
In-Reply-To: <p0611040ebf8a90be11b5@[10.255.254.15]>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu> <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu> <p0611040ebf8a90be11b5@[10.255.254.15]>
Message-ID: <29e4a4427fbb14ea9d89a78b94fba433@uiuc.edu>

just to beat a dead horse, i think these are the currently working 
options for "real time" in "lisp" :

                   SND SC/OSC PortMidi Midishare

Guile Scheme       x
Gauche Scheme            x
SBCL                     x      x        x
OPENMCL                  x      x        x
CMUCL                    -      -        -



From taube at uiuc.edu  Sun Oct 30 09:21:23 2005
From: taube at uiuc.edu (Rick Taube)
Date: Sun, 30 Oct 2005 11:21:23 -0600
Subject: [CM] which lisp
In-Reply-To: <4364FF42.6070904@earthlink.net>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu> <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu> <p0611040ebf8a90be11b5@[10.255.254.15]> <29e4a4427fbb14ea9d89a78b94fba433@uiuc.edu> <4364FF42.6070904@earthlink.net>
Message-ID: <0d51841ea5327e6a817fe509dc506b6e@uiuc.edu>

You should be able to load cm into any recent Guile. So if you have 
snd-rt then i would think you should be able to load cm into it:

galen:~ hkt$ cm -l guile
; loading guile.scm
; loading loop.scm
; loading level1.scm
; loading utils.scm
; loading mop.scm
; loading objects.scm
; loading io.scm
; loading scheduler.scm
; loading sco.scm
; loading clm.scm
; loading clm2.scm
; loading midi1.scm
; loading midi2.scm
; loading midi3.scm
; loading data.scm
; loading scales.scm
; loading spectral.scm
; loading patterns.scm
; loading osc.scm
; loading sc.scm

  /\\\
---\\\---------
----\\\--------
----/\\\------- Common Music 2.7.0
---/--\\\------
--/----\\\-----
  /      \\\/

guile> (lisp-version)
"Guile 1.6.7"
guile>



From taube at uiuc.edu  Sun Oct 30 09:26:40 2005
From: taube at uiuc.edu (Rick Taube)
Date: Sun, 30 Oct 2005 11:26:40 -0600
Subject: [CM] which lisp
In-Reply-To: <4364FF42.6070904@earthlink.net>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu> <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu> <p0611040ebf8a90be11b5@[10.255.254.15]> <29e4a4427fbb14ea9d89a78b94fba433@uiuc.edu> <4364FF42.6070904@earthlink.net>
Message-ID: <a75facdf947583aa3895ce72a2517f74@uiuc.edu>

> Sorry I'm still confused.  Does this mean that I can drive snd-rt from 
> CM in guile?

you would be driving cm from snd-rt. that is, you would be using 
snd-rt's real time capabilities (cm's rt isnt ported to guile now) but 
you could use cm as a "toolkit" or whatever. but i should proabably 
package cm as a scheme module or something if you want to try this, let 
me know.






From taube at uiuc.edu  Sun Oct 30 09:44:32 2005
From: taube at uiuc.edu (Rick Taube)
Date: Sun, 30 Oct 2005 11:44:32 -0600
Subject: [CM] which lisp
In-Reply-To: <436502DE.3040903@earthlink.net>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu> <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu> <p0611040ebf8a90be11b5@[10.255.254.15]> <29e4a4427fbb14ea9d89a78b94fba433@uiuc.edu> <4364FF42.6070904@earthlink.net> <0d51841ea5327e6a817fe509dc506b6e@uiuc.edu> <436502DE.3040903@earthlink.net>
Message-ID: <ccfff047c79b10c267ad068498641c03@uiuc.edu>

with respect to cm, right now the rts scheduler works in SBCL, OPENMCL, 
Gauche Scheme (i think) and CMUCL (kind of). It should be possible to 
implement rts in a Guile with native threads, or any scheme that 
supports SRFI-21. But snd-rt already has rt going so presumably you 
dont need it. CM's Midi IO doesnt work in ANY scheme but only because I 
havent have time to port it. But neither Midishare's nor Portmidi's API 
is hard to implement in any reasonably competent FFI, so someone could 
do it -- I just havent had a chance to and would welcome the help.

btw i ported midishare and portmidi apis to CFFI, so they now work in 
Clisp, Lispworks and any other Lisp that CFFI supports. This means it 
will be possible to do interacitve midi from cm on Windows in a few 
weeks...


On Oct 30, 2005, at 11:29 AM, Ken wrote:

> Sorry for my vagueness.
> But this dosen't support the realtime (rts) stuff, correct?
> I definitely need midi controller support, midi manipuation/generation 
> etc.
> I'm just looking for a lisp synthesis system that can be midi 
> controlled.
> Michael Gogins was/is going to add Lisp bindings to Csound, but he was 
> going to target Guile, which excludes (rts) right?
>
> Thanks for your time,
> Ken
>
> Rick Taube wrote:
>
>> You should be able to load cm into any recent Guile. So if you have 
>> snd-rt then i would think you should be able to load cm into it:
>>
>> galen:~ hkt$ cm -l guile
>> ; loading guile.scm
>> ; loading loop.scm
>> ; loading level1.scm
>> ; loading utils.scm
>> ; loading mop.scm
>> ; loading objects.scm
>> ; loading io.scm
>> ; loading scheduler.scm
>> ; loading sco.scm
>> ; loading clm.scm
>> ; loading clm2.scm
>> ; loading midi1.scm
>> ; loading midi2.scm
>> ; loading midi3.scm
>> ; loading data.scm
>> ; loading scales.scm
>> ; loading spectral.scm
>> ; loading patterns.scm
>> ; loading osc.scm
>> ; loading sc.scm
>>
>>  /\\\
>> ---\\\---------
>> ----\\\--------
>> ----/\\\------- Common Music 2.7.0
>> ---/--\\\------
>> --/----\\\-----
>>  /      \\\/
>>
>> guile> (lisp-version)
>> "Guile 1.6.7"
>> guile>
>>
>>
>



From kjetil at ccrma.Stanford.EDU  Sun Oct 30 11:33:17 2005
From: kjetil at ccrma.Stanford.EDU (Kjetil S. Matheussen)
Date: Sun, 30 Oct 2005 11:33:17 -0800 (PST)
Subject: [CM] which lisp
In-Reply-To: <4364E918.2080303@earthlink.net>
References: <20051028190004.24477.87991.Mailman@cm-mail.stanford.edu>
 <Pine.LNX.4.61.0510292254210.1056@ttleush.stanford.edu> <4364E918.2080303@earthlink.net>
Message-ID: <Pine.LNX.4.61.0510301120530.1056@ttleush.stanford.edu>

The reason for scheme is that snd's extension language is guile/scheme.
Snd-rt is heavely tied to SND, of course.

Yes, it provides realtime CLM. Its a little bit like supercollider 
(supercollider.sf.net), but with more direct control of the samples since 
its not processing the sound on a separate server, or in blocks (ie. there 
is no difference between sample-rate and control-rate).


Hmmm, I think you can use the compositional facilities of CM. One
way is to use CM inside SND (thats just loading the cm.scm file in snd),
and use its functions to schedule realtime snd-rt events.


Second option (which is a lot better regarding realtimeness) is
to run CM using SBCL and sending midi-events to SND-RT, since SND-RT
can receive midi inside the realtime thread:
http://www.notam02.no/arkiv/doc/snd-rt/html/rt-Z-H-6.html#node_sec_5.9
(And adding support for OSC isn't much work to implement, in case it 
should be needed.)




On Sun, 30 Oct 2005, Ken wrote:

> I've glanced over this, but don't really understand it all yet.
> Why scheme?
> All these lisps, make my head dizzy sometimes.
> So this is realtime CLM?
> Is, or can the compositional facilities of CM, be integrated into this, 
> (midi)?
>
> I come from the Windoze world, having used Reaktor, Kontakt, Ableton Live.  I 
> like the live midi-controlled compositional way of working, and am hoping to 
> use CM in this manner, with a gui of some kind.  I also like the style of 
> creating synths, and fx, complete with presets and gui's, so I'm looking 
> around at whats available in the open source gui'less world to see where I 
> could possibly put my development efforts towards creating something that 
> combines the best of the ideas that I have seen.
>
> I REALLY seem to like Lisp though, and would love to work in Lisp 
> exclusively.  Any thoughts appreciated.
>
> Ken
>
> Kjetil S. Matheussen wrote:
>
>> 
>> Depends what kind of realtime work.
>> http://www.notam02.no/arkiv/doc/snd-rt/
>> 
>> 
>> Rick Taube:
>> 
>>> 
>>> on linux sbcl is your best bet. use the latest version and let us know
>>> what problems you encounter.
>>> 
>>> On Oct 28, 2005, at 12:55 AM, Ken wrote:
>>> 
>>>> so which lisp is the focus for realtime wok right now.  after a brief
>>>> absence, i'm ready to push forward.  i'm on linux BTW, fedora core 3
>>>> at the moment if it matters.
>>>> 
>> 
>> 
>> _______________________________________________
>> Cmdist mailing list
>> Cmdist at ccrma.stanford.edu
>> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>> 
>
>



From rundall at uiuc.edu  Sun Oct 30 11:40:10 2005
From: rundall at uiuc.edu (Jake Rundall)
Date: Sun, 30 Oct 2005 13:40:10 -0600
Subject: [CM] MOTU srate problem; .aifc file type; pops
 and clicks using sndplay in OS X
Message-ID: <bd9335cd.53802489.ccb2500@expms3.cites.uiuc.edu>

Hi,

Three questions/problems today:

1. srate problem when playing files in clm-3 through MOTU

When I try to play files (in OS X) using dac/sndplay in clm-3 and do it through 
my MOTU (mk I), the MOTU only plays back stuff at 96 kHz as the srate.  I've got 
the latest driver for the MOTU.  Also, the MOTU cooperates with QuickTime and 
iTunes and all of that stuff, so it seems that it's not a problem with the MOTU.  
But, no matter what srate I write the file at, when I play it back in clm, the MOTU 
switches it's displayed srate to 48 kHz and then plays it back as if it were 
playing at twice that (96 kHz).  It's strange, because the MOTU mk I is not even 
supposed to play things at 96 kHz, just at 48 and 44.1, so I don't know what's 
happening.  I think Rick Taube is having a similar problem with his M-Audio 
boxes, so I don't think it's an issue only with MOTUs, but with firewire audio 
interfaces in general.  So, I'm curious what the problem is and why clm doesn't 
seem to cooperate with my firewire interface.  Luckily, I've found a solution, but 
this leads to another question.  I'll describe this all next.

2. paplay, and .aifc header-type

I've often used Apple's QuickTime Player to play back files written in clm, but I 
couldn't get it to handle 4-channel .aiff files (even though QuickTime is 
supposed to play back up to 24 channels -- I just don't know what types of files 
it is actually capable of handling...).  So, I looked for a while and found a 
command-line audio file player called paplay that is in a set of programs called 
the Multi-Channel ToolKit (mctools), created by Richard Dobson in the UK.  It's 
available at http://www.bath.ac.uk/~masrwd/mctools.html.  paplay cooperates 
well with my MOTU (no sampling rate miscommunication) and opens multi-
channel .aiff files and even plays them correctly.  The catch is, when I run with-
sound to write the files, I have to specify "header-type: mus-aiff".  It seems that 
the default file header for clm is .aifc (compressed).  I was wondering why this is 
-- does clm actually write compressed files as a default?  Would it be better to 
use an .aiff header as the default?  Luckily, as long as I tell clm to write a .aiff 
header, the paplay program is happy.

3. pops and clicks when using sndplay

This whole exploration reminded me of another general issue I've had with 
sndplay.  Ever since I've been using clm (a little over a year only), I've found that 
I get pops and clicks when I use dac/sndplay to play back my files.  Because of 
this, I almost always ended up playing back my files using QuickTime and there 
were no problems with pops and clicks.  Now that I've set paplay (mentioned 
above in #2) as *clm-player*, I can play back my files in clm using dac and I 
don't have the problem either, so it would seem that the problem is specifically 
with sndplay (and probably with the way it works in OS X).  Anyway, I'm very 
glad to have found a solution, but I'd like to be able to use sndplay to play back 
my files at some point.  Also, I may not be the only person out there 
experiencing pops and clicks while using sndplay in OS X, so I wonder if 
something could be improved on (maybe by looking at how paplay plays files or 
something).

I'd love to hear what other folks have to say and if there are any explanations for 
what I've been experiencing.

Thanks,

Jake Rundall



From bil at ccrma.Stanford.EDU  Mon Oct 31 04:25:40 2005
From: bil at ccrma.Stanford.EDU (Bill Schottstaedt)
Date: Mon, 31 Oct 2005 04:25:40 -0800
Subject: [CM] MOTU srate problem; .aifc file type; pops and clicks using
 sndplay in OS X
In-Reply-To: <bd9335cd.53802489.ccb2500@expms3.cites.uiuc.edu>
References: <bd9335cd.53802489.ccb2500@expms3.cites.uiuc.edu>
Message-ID: <43660D44.1070005@ccrma>

I'll look at paplay -- thanks for the info.  On aifc/aiff, aifc replaced
aiff about 125 years ago -- it does not mean the data is compressed.



From kjetil at ccrma.Stanford.EDU  Mon Oct 31 16:58:50 2005
From: kjetil at ccrma.Stanford.EDU (Kjetil S. Matheussen)
Date: Mon, 31 Oct 2005 16:58:50 -0800 (PST)
Subject: [CM] Re: clicks using sndplay in OS X
In-Reply-To: <20051030200002.22562.86998.Mailman@cm-mail.stanford.edu>
References: <20051030200002.22562.86998.Mailman@cm-mail.stanford.edu>
Message-ID: <Pine.LNX.4.61.0510311651200.1063@ttleush.stanford.edu>

Jake Rundall:
>
> 3. pops and clicks when using sndplay
>
> This whole exploration reminded me of another general issue I've had with
> sndplay.  Ever since I've been using clm (a little over a year only), I've found that
> I get pops and clicks when I use dac/sndplay to play back my files.  Because of
> this, I almost always ended up playing back my files using QuickTime and there
> were no problems with pops and clicks.  Now that I've set paplay (mentioned
> above in #2) as *clm-player*, I can play back my files in clm using dac and I
> don't have the problem either, so it would seem that the problem is specifically
> with sndplay (and probably with the way it works in OS X).  Anyway, I'm very
> glad to have found a solution, but I'd like to be able to use sndplay to play back
> my files at some point.  Also, I may not be the only person out there
> experiencing pops and clicks while using sndplay in OS X, so I wonder if
> something could be improved on (maybe by looking at how paplay plays files or
> something).
>

Coincidentally just looking at the sndplay.c file right now...
Can this be the cause of the pops and clicks on macoscx?

#if MAC_OSX
   #define BUFFER_SIZE 256
#else
#if WIN32
   /* this setting from Scott Middleton (actually used 8096) */
   #define BUFFER_SIZE 8192
#else
   #define BUFFER_SIZE 4096
#endif
#endif

(Why so small default buffer on macosx?)

Perhaps using the -bufsize option for sndplay helps?



--