From squarewave at elisanet.fi  Fri Mar 11 08:41:34 2022
From: squarewave at elisanet.fi (Matti Koskinen)
Date: Fri, 11 Mar 2022 18:41:34 +0200
Subject: [CM] Simple example
Message-ID: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>

Hi all,

I got Grace compiled and it works, but since I?ve used it many years ago, can?t remember anything at all. 
If someone could post a simple example how to read midi from keyboard and save it to a midi file, I?d be more than happy.

Thanks 

-m




From bil at ccrma.Stanford.EDU  Fri Mar 11 13:53:38 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Fri, 11 Mar 2022 13:53:38 -0800
Subject: [CM] Snd 22.2
Message-ID: <ae29ce74ba599c3c14074677cd857b3b@ccrma.stanford.edu>

Snd 22.2:

moved s7_apply_* from s7.h to xen.h if DISABLE_DEPRECATED.

checked: sbcl 2.2.1|2, notcurses 3.0.6|7

Thanks!: Mike Scholz



From treegestalt at gmail.com  Fri Mar 11 17:48:32 2022
From: treegestalt at gmail.com (Forrest Curo)
Date: Fri, 11 Mar 2022 17:48:32 -0800
Subject: [CM] Simple example
In-Reply-To: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
References: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
Message-ID: <CAAn-Ecz=EFSjen8jCXUW81vTFCeZupX-dGx=Qw+4Cht8yaMNFw@mail.gmail.com>

If you run Grace and check the Help menu, you will find several examples
listed, explaining how to do most of what a user is likely to want to know.

On Fri, Mar 11, 2022 at 8:42 AM Matti Koskinen <squarewave at elisanet.fi>
wrote:

> Hi all,
>
> I got Grace compiled and it works, but since I?ve used it many years ago,
> can?t remember anything at all.
> If someone could post a simple example how to read midi from keyboard and
> save it to a midi file, I?d be more than happy.
>
> Thanks
>
> -m
>
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220311/8c30d086/attachment.html>

From squarewave at elisanet.fi  Fri Mar 11 19:16:48 2022
From: squarewave at elisanet.fi (Matti Koskinen)
Date: Sat, 12 Mar 2022 05:16:48 +0200
Subject: [CM] Simple example
In-Reply-To: <CAAn-Ecz=EFSjen8jCXUW81vTFCeZupX-dGx=Qw+4Cht8yaMNFw@mail.gmail.com>
References: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
	<CAAn-Ecz=EFSjen8jCXUW81vTFCeZupX-dGx=Qw+4Cht8yaMNFw@mail.gmail.com>
Message-ID: <8E8CC349-F824-4008-B881-306EED451D66@elisanet.fi>

Hi,

Thanks, but none of the examples show what I want.
I thought that

(with-midi(?test.mid?)
  (mp:receive myplayer))

would do the trick. It doesn?t give any errors, nor does it write the midi file either.

Tnx

-m


> On 12. Mar 2022, at 3.48, Forrest Curo <treegestalt at gmail.com> wrote:
> 
> If you run Grace and check the Help menu, you will find several examples listed, explaining how to do most of what a user is likely to want to know.
> 
> On Fri, Mar 11, 2022 at 8:42 AM Matti Koskinen <squarewave at elisanet.fi <mailto:squarewave at elisanet.fi>> wrote:
> Hi all,
> 
> I got Grace compiled and it works, but since I?ve used it many years ago, can?t remember anything at all. 
> If someone could post a simple example how to read midi from keyboard and save it to a midi file, I?d be more than happy.
> 
> Thanks 
> 
> -m
> 
> 
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu <mailto:Cmdist at ccrma.stanford.edu>
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist <https://cm-mail.stanford.edu/mailman/listinfo/cmdist>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220312/b43937ac/attachment.html>

From squarewave at elisanet.fi  Sat Mar 12 13:30:48 2022
From: squarewave at elisanet.fi (Matti Koskinen)
Date: Sat, 12 Mar 2022 23:30:48 +0200
Subject: [CM] Simple example
In-Reply-To: <8E8CC349-F824-4008-B881-306EED451D66@elisanet.fi>
References: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
	<CAAn-Ecz=EFSjen8jCXUW81vTFCeZupX-dGx=Qw+4Cht8yaMNFw@mail.gmail.com>
	<8E8CC349-F824-4008-B881-306EED451D66@elisanet.fi>
Message-ID: <9BDE7780-D0AC-480B-9996-7141B1071597@elisanet.fi>

Hi,

Now I remember, how I created the midi file. The player wrote a text file, that could be converted to a midi file with txt2midi programme. The txt2midi and midi2txt can?t be compiled with newer c-compilers, but this link https://grfia.dlsi.ua.es/gen.php?id=resources <https://grfia.dlsi.ua.es/gen.php?id=resources>
has newer version, that compiled nicely one Bg Sur. Now if I?d only had any idea of what was done, except how hard it was getting the timing right. It?s so long a ago, that the sources have long been in the heaven of bits:-)


-m


> On 12. Mar 2022, at 5.16, Matti Koskinen <squarewave at elisanet.fi> wrote:
> 
> Hi,
> 
> Thanks, but none of the examples show what I want.
> I thought that
> 
> (with-midi(?test.mid?)
>   (mp:receive myplayer))
> 
> would do the trick. It doesn?t give any errors, nor does it write the midi file either.
> 
> Tnx
> 
> -m
> 
> 
>> On 12. Mar 2022, at 3.48, Forrest Curo <treegestalt at gmail.com <mailto:treegestalt at gmail.com>> wrote:
>> 
>> If you run Grace and check the Help menu, you will find several examples listed, explaining how to do most of what a user is likely to want to know.
>> 
>> On Fri, Mar 11, 2022 at 8:42 AM Matti Koskinen <squarewave at elisanet.fi <mailto:squarewave at elisanet.fi>> wrote:
>> Hi all,
>> 
>> I got Grace compiled and it works, but since I?ve used it many years ago, can?t remember anything at all. 
>> If someone could post a simple example how to read midi from keyboard and save it to a midi file, I?d be more than happy.
>> 
>> Thanks 
>> 
>> -m
>> 
>> 
>> _______________________________________________
>> Cmdist mailing list
>> Cmdist at ccrma.stanford.edu <mailto:Cmdist at ccrma.stanford.edu>
>> https://cm-mail.stanford.edu/mailman/listinfo/cmdist <https://cm-mail.stanford.edu/mailman/listinfo/cmdist>
> 
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220312/6edfef12/attachment.html>

From squarewave at elisanet.fi  Sat Mar 12 22:05:19 2022
From: squarewave at elisanet.fi (Matti Koskinen)
Date: Sun, 13 Mar 2022 08:05:19 +0200
Subject: [CM] Simple example
In-Reply-To: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
References: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
Message-ID: <78ED4242-687D-4729-9F87-7E77FD2C9EED@elisanet.fi>

Found a solution, surely not the most elegant one, but seems to work:

;
;; Receiving MIDI Input
;


; Here is the definition of a MIDI receiver that routes midi on and
; off messages to the output port and as alist to *m* global variable
(define *m* (list (now)))

(define (myplayer data)
  (let* ((opcode (first data))
         (channel (second data))
         (keyn (third data))
         (velocity (fourth data)))
    (set! *m* (append *m* (list (list data  (now)))))
    (if (= opcode mm:on)
        (mp:on :time (now) :key keyn :vel velocity :chan channel)
    (if (= opcode mm:off)
        (mp:off :key keyn :chan channel)))))

; Set receiver, then play your midi keyboard

(mp:receive myplayer)


; When you are done clear it...

(mp:receive #f)

;save the *m* to a file
(define (play-it-again-sam file)
    
  (sprout (let* ((notes (rest *m*))
              (wasnow (first *m*)))
             
  (process for d in notes do
           (print  d)
           (let* ((data (first d))
                  (op (first data))
                  (ch (second data))
                  (keyn (third data))
                  (veloc(fourth data)))
             (if (= op mm:on)
               (mp:on :time (- wasnow (last d)) :chan ch
                      :key keyn :vel veloc))
             (if (= op mm:off)
               (mp:off :time (- wasnow (last d)) 
                       :chan ch :key keyn))))) file))
;test it
(play-it-again-sam "ts3.mid")


> On 11. Mar 2022, at 18.41, Matti Koskinen <squarewave at elisanet.fi> wrote:
> 
> Hi all,
> 
> I got Grace compiled and it works, but since I?ve used it many years ago, can?t remember anything at all. 
> If someone could post a simple example how to read midi from keyboard and save it to a midi file, I?d be more than happy.
> 
> Thanks 
> 
> -m
> 
> 
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist




From juanig at ccrma.Stanford.EDU  Mon Mar 14 14:31:04 2022
From: juanig at ccrma.Stanford.EDU (Juan Reyes)
Date: Mon, 14 Mar 2022 14:31:04 -0700
Subject: [CM] Simple example
In-Reply-To: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
References: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
Message-ID: <8c32f8f1-42e9-e216-16ba-ac9cc25f8fac@ccrma.stanford.edu>


  -On the subject-matter of compiling Grace-.

I take it Matti compiled Grace for Mac OSX using Juce.

Been trying to compile Grace for Linux Fedora 34 with no luck so far.

Even Juce doesn't seem to work fine, and thus gave up.

Don't know if anyone has a recent binary running on Linux or at least
pointers for compiling it.

Grace is S7!, making things still appealing.

   Regards,

   --* Juan Reyes



> 
> I got Grace compiled and it works, but since I?ve used it many years 
> ago, can?t remember anything at all. If someone could post a simple 
> example how to read midi from keyboard and save it to a midi file, 
> I?d be more than happy.
> 


From juanig at ccrma.Stanford.EDU  Tue Mar 15 15:31:28 2022
From: juanig at ccrma.Stanford.EDU (Juan Reyes)
Date: Tue, 15 Mar 2022 15:31:28 -0700
Subject: [CM] Grace on Linux Fedora 34
In-Reply-To: <8c32f8f1-42e9-e216-16ba-ac9cc25f8fac@ccrma.stanford.edu>
References: <02A3271C-7900-496A-9878-94A5D1524D67@elisanet.fi>
	<8c32f8f1-42e9-e216-16ba-ac9cc25f8fac@ccrma.stanford.edu>
Message-ID: <96a87590-fb17-e35c-db3d-1d856a47edbc@ccrma.stanford.edu>


Hi,

Was able to compile Grace 3.10.3 with Juce on Linux Fedora 34.

Lot's of warnings and a tweak(*).

Perhaps some of the Juce code needs to be updated to newer Juce versions
to avoid warnings.

   --* Juan


(*) Tweak is that
   'JUCE_MODAL_LOOPS_PERMITTED=1'

    Should be added to Projucer's Preprocessor Definitions

Steps below:

#  mkdir parent
#  cd parent
#  git clone https://github.com/juce-framework/JUCE.git
#  curl ftp://ccrma-ftp.stanford.edu/pub/Lisp/sndlib.tar.gz | tar -zx
#  git clone https://github.com/ricktaube/grace.git grace

#  cd sndlib
#  premake4 --with-g++
#  make

Now for Juce:

Need to compile (make) Projucer App (binary)

(make sure you have dependencies on
"doc/Linux_dependencies.md Linux_dependencies.md")

libasound2-dev libjack-jackd2-dev \
ladspa-sdk \
libcurl4-openssl-dev  \
libfreetype6-dev \
libx11-dev libxcomposite-dev libxcursor-dev \
libxcursor-dev libxext-dev libxinerama-dev \
libxrandr-dev libxrender-dev \
libwebkit2gtk-4.0-dev \
libglu1-mesa-dev mesa-common-dev


# cd  ../
# cd  JUCE/extras/Projucer/Builds/LinuxMakefile
# make

Once we have a binary and on the Builds directory:

#  cd build/

Run Projucer:

#  ./Projucer

Now on  Projucer's Window on the File Menu,

OPEN ../parent/grace/Grace.jucer

On the Exporters tab make sure Linux Makefile is yellowed
or outlined.

On Projucer's right side list scroll down on to
"Preprocessors Definitions"

Add the following definition: JUCE_MODAL_LOOPS_PERMITTED=1

Then Preprocessor Definitions should look like:

     JUCE_MODAL_LOOPS_PERMITTED=1
     HAVE_SCHEME=1

On Projucer's File menu: select 'SAVE'

This will create a new Makefile on

  '../parent/grace/Builds/LinuxMakefile'

Now,

#  cd ../parent/grace/Builds/LinuxMakefile
and,
#  make

Grace binary should be on the Build directory

# cd Build

Run Grace

./Grace &



> 
>  ?-On the subject-matter of compiling Grace-.
> 
> I take it Matti compiled Grace for Mac OSX using Juce.
> 
> Been trying to compile Grace for Linux Fedora 34 with no luck so far.
> 
> Even Juce doesn't seem to work fine, and thus gave up.
> 
> Don't know if anyone has a recent binary running on Linux or at least
> pointers for compiling it.
> 
> Grace is S7!, making things still appealing.
> 


From wdouglass at carnegierobotics.com  Wed Mar 16 06:18:12 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Wed, 16 Mar 2022 13:18:12 +0000
Subject: [CM] Lint suggests do
Message-ID: <f89f9adca11c0494b638372bd8942d1417b82a06.camel@carnegierobotics.com>

Hi all,

I've started running lint.scm against my (working) code, and i've
understood and implemented a lot of the suggestions that it provides. 

One that I don't understand is the suggestion that i replace a lot of
named lets with `do` loops

Are do loops more efficient, or is this a stylistic choice? i've always
found them to be sort of unlispy, and i'd love to learn more here.

Thanks,
Woody Douglass




From bil at ccrma.Stanford.EDU  Wed Mar 16 06:39:31 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Wed, 16 Mar 2022 06:39:31 -0700
Subject: [CM] Lint suggests do
In-Reply-To: <f89f9adca11c0494b638372bd8942d1417b82a06.camel@carnegierobotics.com>
References: <f89f9adca11c0494b638372bd8942d1417b82a06.camel@carnegierobotics.com>
Message-ID: <ed8d54d530acea30ff538ef46fb7be89@ccrma.stanford.edu>

It's mostly stylistic -- I hate named let.  But
I think a do-loop is nearly always faster in s7
due to its roots in Snd.  I can add a flag to control
that report if you like.  Please let me know of any
stupidities in lint -- I use it a lot, but
I am set in my ways.



From wdouglass at carnegierobotics.com  Wed Mar 16 06:57:09 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Wed, 16 Mar 2022 13:57:09 +0000
Subject: [CM] Lint suggests do
In-Reply-To: <ed8d54d530acea30ff538ef46fb7be89@ccrma.stanford.edu>
References: <f89f9adca11c0494b638372bd8942d1417b82a06.camel@carnegierobotics.com>
	<ed8d54d530acea30ff538ef46fb7be89@ccrma.stanford.edu>
Message-ID: <d0cb1fbe9639477cf974fc2b8c91ac9e8bac0f99.camel@carnegierobotics.com>

Oh no problem, i don't mind lint being a bit opinionated! Honestly,
since i started linting my code i've learned a lot of stylistic things
that I wouldn't have thought of. My scheme code still has a "this was
written by a C programmer" smell...

Thanks a lot,
Woody

On Wed, 2022-03-16 at 06:39 -0700, bil at ccrma.Stanford.EDU wrote:
> It's mostly stylistic -- I hate named let.  But
> I think a do-loop is nearly always faster in s7
> due to its roots in Snd.  I can add a flag to control
> that report if you like.  Please let me know of any
> stupidities in lint -- I use it a lot, but
> I am set in my ways.
> 



From wdouglass at carnegierobotics.com  Mon Mar 21 09:21:26 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Mon, 21 Mar 2022 16:21:26 +0000
Subject: [CM] pre-parsing scheme code
Message-ID: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>

All,

I have a habit of creating a lot of s7 lists in C. this manifested in
my C code being littered with nested `s7_list_nl` calls, and became
unwieldy pretty quickly. So i wrote the attached script to "pre-parse"
scheme s-expressions that i use in my C code, and generate C code -- it
has done wonders for the maintainability of my codebase; also, it makes
running the linter against the embedded scheme less finicky.

Have any of you done something similar to this? is there a reason it's
a bad idea? is there a better way?

I guess this is just a request for comment; thanks for any insight you
can provide. also, feel free to use/improve this script!
-Woody Douglass
-------------- next part --------------
A non-text attachment was scrubbed...
Name: scheme_to_c.scm
Type: text/x-scheme
Size: 3451 bytes
Desc: scheme_to_c.scm
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220321/80f9f1e4/attachment.bin>

From k.s.matheussen at gmail.com  Mon Mar 21 10:29:11 2022
From: k.s.matheussen at gmail.com (Kjetil Matheussen)
Date: Mon, 21 Mar 2022 18:29:11 +0100
Subject: [CM] pre-parsing scheme code
In-Reply-To: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
Message-ID: <CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>

Nice code. But isn't the result of doing:

  s7_eval_c_string("(list 1 2 3)");

the same as the result of doing:

  s7_list_nl(s, 3, s7_make_integer(s, 1),\
    s7_make_integer(s, 2),\
    s7_make_integer(s, 3),\
    NULL);

?

Except that the second version probably runs much faster...


On Mon, Mar 21, 2022 at 5:25 PM Woody Douglass
<wdouglass at carnegierobotics.com> wrote:
>
> All,
>
> I have a habit of creating a lot of s7 lists in C. this manifested in
> my C code being littered with nested `s7_list_nl` calls, and became
> unwieldy pretty quickly. So i wrote the attached script to "pre-parse"
> scheme s-expressions that i use in my C code, and generate C code -- it
> has done wonders for the maintainability of my codebase; also, it makes
> running the linter against the embedded scheme less finicky.
>
> Have any of you done something similar to this? is there a reason it's
> a bad idea? is there a better way?
>
> I guess this is just a request for comment; thanks for any insight you
> can provide. also, feel free to use/improve this script!
> -Woody Douglass
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist


From wdouglass at carnegierobotics.com  Mon Mar 21 10:31:46 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Mon, 21 Mar 2022 17:31:46 +0000
Subject: [CM] pre-parsing scheme code
In-Reply-To: <CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
Message-ID: <16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>

The second version running much faster is the goal; so i'm mechanically
transforming the first version to the second

-Woody

On Mon, 2022-03-21 at 18:29 +0100, Kjetil Matheussen wrote:
> CAUTION: This email originated from outside of the organization. Do
> not click links or open attachments unless you recognize the sender.
> 
> 
> Nice code. But isn't the result of doing:
> 
>   s7_eval_c_string("(list 1 2 3)");
> 
> the same as the result of doing:
> 
>   s7_list_nl(s, 3, s7_make_integer(s, 1),\
>     s7_make_integer(s, 2),\
>     s7_make_integer(s, 3),\
>     NULL);
> 
> ?
> 
> Except that the second version probably runs much faster...
> 
> 
> On Mon, Mar 21, 2022 at 5:25 PM Woody Douglass
> <wdouglass at carnegierobotics.com> wrote:
> > All,
> > 
> > I have a habit of creating a lot of s7 lists in C. this manifested
> > in
> > my C code being littered with nested `s7_list_nl` calls, and became
> > unwieldy pretty quickly. So i wrote the attached script to "pre-
> > parse"
> > scheme s-expressions that i use in my C code, and generate C code
> > -- it
> > has done wonders for the maintainability of my codebase; also, it
> > makes
> > running the linter against the embedded scheme less finicky.
> > 
> > Have any of you done something similar to this? is there a reason
> > it's
> > a bad idea? is there a better way?
> > 
> > I guess this is just a request for comment; thanks for any insight
> > you
> > can provide. also, feel free to use/improve this script!
> > -Woody Douglass
> > _______________________________________________
> > Cmdist mailing list
> > Cmdist at ccrma.stanford.edu
> > https://cm-mail.stanford.edu/mailman/listinfo/cmdist



From bil at ccrma.Stanford.EDU  Mon Mar 21 12:17:20 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 21 Mar 2022 12:17:20 -0700
Subject: [CM] pre-parsing scheme code
In-Reply-To: <16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
Message-ID: <2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>

I didn't realize there was such a difference in speed:

#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include "s7.h"

int main(int argc, char **argv)
{
   s7_scheme *s;
   int i;
   s = s7_init();
#if 1
   for (i = 0; i < 2000000; i++)
     s7_eval_c_string(s, "(list 1 2 3)");
#else
   for (i = 0; i < 2000000; i++)
     s7_list_nl(s, 4, s7_make_symbol(s, "list"),	\
	       s7_make_integer(s, 1),		\
	       s7_make_integer(s, 2),		\
	       s7_make_integer(s, 3),		\
	       NULL);
#endif
}

/* gcc -o ex1 ex1.c s7.o -O2 -lm -I. -ldl */
/* 3045 vs 703 callgrind */

A factor of 4!




From bil at ccrma.Stanford.EDU  Mon Mar 21 12:25:52 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 21 Mar 2022 12:25:52 -0700
Subject: [CM] pre-parsing scheme code
In-Reply-To: <2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
Message-ID: <704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>

Gah -- I forgot to eval the expression in the second
case -- then it's 1600, which is more what I expected.

   for (i = 0; i < 2000000; i++)
     s7_eval(s, s7_list_nl(s, 4, s7_make_symbol(s, "list"),	\
	       s7_make_integer(s, 1),		\
	       s7_make_integer(s, 2),		\
	       s7_make_integer(s, 3),		\
	       NULL), s7_rootlet(s));



From bil at ccrma.Stanford.EDU  Mon Mar 21 12:27:35 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 21 Mar 2022 12:27:35 -0700
Subject: [CM] pre-parsing scheme code
In-Reply-To: <704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
	<704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
Message-ID: <b5a1959c68fe58478418da7d18acaac5@ccrma.stanford.edu>

Sorry about that repetition -- I seem to have a flakey
net connection today -- can't see any squirrels chewing
on the cable.



From wdouglass at carnegierobotics.com  Mon Mar 21 12:31:20 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Mon, 21 Mar 2022 19:31:20 +0000
Subject: [CM] pre-parsing scheme code
In-Reply-To: <704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
	<704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
Message-ID: <12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>

Thanks for testing this! A factor of 2 is still nothing to sneeze at;
I'm running s7 on a resource-constrained embeddded machine, so I'll
take every cycle i can get at runtime!

On Mon, 2022-03-21 at 12:25 -0700, bil at ccrma.Stanford.EDU wrote:
> Gah -- I forgot to eval the expression in the second
> case -- then it's 1600, which is more what I expected.
> 
>    for (i = 0; i < 2000000; i++)
>      s7_eval(s, s7_list_nl(s, 4, s7_make_symbol(s, "list"),	\
> 	       s7_make_integer(s, 1),		\
> 	       s7_make_integer(s, 2),		\
> 	       s7_make_integer(s, 3),		\
> 	       NULL), s7_rootlet(s));
> 



From bil at ccrma.Stanford.EDU  Mon Mar 21 13:04:06 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 21 Mar 2022 13:04:06 -0700
Subject: [CM] pre-parsing scheme code
In-Reply-To: <12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
	<704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
	<12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>
Message-ID: <5ee7f7478dae575edac4ec4a0ecbaeb6@ccrma.stanford.edu>

If you're running the same code many times, it
might be worth it to call s7_optimize, save
the function it returns, then call that function
rather than eval+expression-writing code.  There's
an example in snd-sig.c.  The idea there is that
you're running a function over every sample of
some sound, so you first optimize it, then apply
that to each sample. snd-sig.c uses s7_set_slot
and so on to change the (user-level) function's
argument to be each new sample.  I should write a
simpler example...  (It's faster because you don't
rebuild the code each time you use it,
and the optimized evaluation is usually at least
twice as fast.)



From bil at ccrma.Stanford.EDU  Mon Mar 21 13:32:19 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 21 Mar 2022 13:32:19 -0700
Subject: [CM] pre-parsing scheme code
In-Reply-To: <12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
	<704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
	<12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>
Message-ID: <9afbd5da5c6c813aa5d33ea27d2ccbbb@ccrma.stanford.edu>

Here's an example:

#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include "s7.h"

int main(int argc, char **argv)
{
   s7_scheme *s;
   s7_pfunc f;
   s7_pointer body;
   int i;
   s = s7_init();
   body = s7_list_nl(s, 4, s7_make_symbol(s, "list"),	\
	       s7_make_integer(s, 1),		\
	       s7_make_integer(s, 2),		\
	       s7_make_integer(s, 3),		\
	       NULL);
   f = s7_optimize(s, s7_list(s, 1, body));
   if (!f) fprintf(stderr, "oops\n");
   else
     for (i = 0; i < 2000000; i++)
       f(s);
}

/* 130 in callgrind */



From wdouglass at carnegierobotics.com  Tue Mar 22 04:39:36 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Tue, 22 Mar 2022 11:39:36 +0000
Subject: [CM] pre-parsing scheme code
In-Reply-To: <9afbd5da5c6c813aa5d33ea27d2ccbbb@ccrma.stanford.edu>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
	<704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
	<12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>
	<9afbd5da5c6c813aa5d33ea27d2ccbbb@ccrma.stanford.edu>
Message-ID: <93818b3cebfbfae882d8c192ec3ae7b4d87bfdd0.camel@carnegierobotics.com>

I didn't know about s7_optimize -- super cool! i'll start using it,
thanks!

-Woody Douglass

On Mon, 2022-03-21 at 13:32 -0700, bil at ccrma.Stanford.EDU wrote:
> Here's an example:
> 
> #include <stdlib.h>
> #include <stdio.h>
> #include <string.h>
> #include "s7.h"
> 
> int main(int argc, char **argv)
> {
>    s7_scheme *s;
>    s7_pfunc f;
>    s7_pointer body;
>    int i;
>    s = s7_init();
>    body = s7_list_nl(s, 4, s7_make_symbol(s, "list"),	\
> 	       s7_make_integer(s, 1),		\
> 	       s7_make_integer(s, 2),		\
> 	       s7_make_integer(s, 3),		\
> 	       NULL);
>    f = s7_optimize(s, s7_list(s, 1, body));
>    if (!f) fprintf(stderr, "oops\n");
>    else
>      for (i = 0; i < 2000000; i++)
>        f(s);
> }
> 
> /* 130 in callgrind */
> 



From bil at ccrma.Stanford.EDU  Tue Mar 22 05:50:26 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Tue, 22 Mar 2022 05:50:26 -0700
Subject: [CM] pre-parsing scheme code
In-Reply-To: <93818b3cebfbfae882d8c192ec3ae7b4d87bfdd0.camel@carnegierobotics.com>
References: <c38196bdbcb32ba83ea4cf100df42b83a8f1e9e9.camel@carnegierobotics.com>
	<CAC6niEKL1crnLyOZSiq0E6Vo5529Fnq92L19FrBjy2KW46Ue7Q@mail.gmail.com>
	<16c3a2e1ef634b6044297aa159a92a1c5ad45cf1.camel@carnegierobotics.com>
	<2a4255f5abc17d7550c0d3c3345305a5@ccrma.stanford.edu>
	<704b1ab9f2adf072aaee76426b37155e@ccrma.stanford.edu>
	<12fbf06c317728698c0150f51799bfa0f09a00c2.camel@carnegierobotics.com>
	<9afbd5da5c6c813aa5d33ea27d2ccbbb@ccrma.stanford.edu>
	<93818b3cebfbfae882d8c192ec3ae7b4d87bfdd0.camel@carnegierobotics.com>
Message-ID: <5a3eef5d94a3cef8ff507eeab92447b3@ccrma.stanford.edu>

I added that example to s7.html, and added GC protection
for "body".  I didn't advertise s7_optimize in the past
because I was fiddling with it for a long time.  There
is also s7_float_optimize which is probably not
documented, and others like it in s7.c.  If s7_optimize
returns NULL, you can fall back on s7_eval (NULL just
means the optimizer gave up for some reason).




From wdouglass at carnegierobotics.com  Tue Mar 22 11:38:38 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Tue, 22 Mar 2022 18:38:38 +0000
Subject: [CM] s7_call_with_location
Message-ID: <ceb7c4f72b324d2ed6c33008ba334403f28911a4.camel@carnegierobotics.com>

All,

What are the rules RE: s7_call_with_location? does it work with any
applicable object? conversely, how difficult would it be to implement
s7_eval_with_location?

Thanks,
Woody Douglass



From bil at ccrma.Stanford.EDU  Tue Mar 22 11:53:00 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Tue, 22 Mar 2022 11:53:00 -0700
Subject: [CM] =?utf-8?q?s7=5Fcall=5Fwith=5Flocation?=
In-Reply-To: <ceb7c4f72b324d2ed6c33008ba334403f28911a4.camel@carnegierobotics.com>
References: <ceb7c4f72b324d2ed6c33008ba334403f28911a4.camel@carnegierobotics.com>
Message-ID: <0ba72caeb2604fa7c901475c86860c5f@ccrma.stanford.edu>

> What are the rules RE: s7_call_with_location? does it work with any
> applicable object? conversely, how difficult would it be to implement
> s7_eval_with_location?

Yes, it should work with any applicable object.  To
implement s7_eval_with_location, use the location wrapper
code in s7_call_with_location, substituting s7_eval for
s7_call (and making the function arguments match s7_eval).




From wdouglass at carnegierobotics.com  Tue Mar 22 13:02:55 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Tue, 22 Mar 2022 20:02:55 +0000
Subject: [CM] s7_call_with_location
In-Reply-To: <0ba72caeb2604fa7c901475c86860c5f@ccrma.stanford.edu>
References: <ceb7c4f72b324d2ed6c33008ba334403f28911a4.camel@carnegierobotics.com>
	<0ba72caeb2604fa7c901475c86860c5f@ccrma.stanford.edu>
Message-ID: <f1a7c43e0ce051ee5468c49e05affd30ee8472a1.camel@carnegierobotics.com>

Thanks -- I'm applying the attached patch to my local s7 copy

seems to work so far

Woody Douglass


On Tue, 2022-03-22 at 11:53 -0700, bil at ccrma.Stanford.EDU wrote:
> > What are the rules RE: s7_call_with_location? does it work with any
> > applicable object? conversely, how difficult would it be to
> > implement
> > s7_eval_with_location?
> 
> Yes, it should work with any applicable object.  To
> implement s7_eval_with_location, use the location wrapper
> code in s7_call_with_location, substituting s7_eval for
> s7_call (and making the function arguments match s7_eval).
> 
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0003-eval-with-location.patch
Type: text/x-patch
Size: 1434 bytes
Desc: 0003-eval-with-location.patch
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220322/eb14e7b8/attachment.bin>

From bil at ccrma.Stanford.EDU  Tue Mar 22 13:19:22 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Tue, 22 Mar 2022 13:19:22 -0700
Subject: [CM] =?utf-8?q?s7=5Fcall=5Fwith=5Flocation?=
In-Reply-To: <f1a7c43e0ce051ee5468c49e05affd30ee8472a1.camel@carnegierobotics.com>
References: <ceb7c4f72b324d2ed6c33008ba334403f28911a4.camel@carnegierobotics.com>
	<0ba72caeb2604fa7c901475c86860c5f@ccrma.stanford.edu>
	<f1a7c43e0ce051ee5468c49e05affd30ee8472a1.camel@carnegierobotics.com>
Message-ID: <13983c421b32ce11ae330a5ff19d50cd@ccrma.stanford.edu>

> seems to work so far

good -- I'll add it to my version, but I think I'll
include the environment argument.  Thanks!




From wdouglass at carnegierobotics.com  Wed Mar 23 12:09:43 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Wed, 23 Mar 2022 19:09:43 +0000
Subject: [CM] Identify binary plugins
Message-ID: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>

All,

S7 identifies plugins by file extension -- ".so" files are considered
binary, and anything else is assumed to be scheme

The attached patch identifies by ELF header, so there doesn't need to
be any rules about plugin naming. Just thought i'd share

Thanks,
Woody Douglass
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0004-identify-elf.patch
Type: text/x-patch
Size: 1522 bytes
Desc: 0004-identify-elf.patch
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220323/5afb82a6/attachment.bin>

From bil at ccrma.Stanford.EDU  Wed Mar 23 14:57:33 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Wed, 23 Mar 2022 14:57:33 -0700
Subject: [CM] Identify binary plugins
In-Reply-To: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
References: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
Message-ID: <3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>

Thanks, but isn't ELF format specific to Linux?
My *.so files on a Mac appear to be something else.
Also, can't an ELF file be an executable?



From elronnd at elronnd.net  Wed Mar 23 15:00:14 2022
From: elronnd at elronnd.net (Elijah Stone)
Date: Wed, 23 Mar 2022 15:00:14 -0700 (PDT)
Subject: [CM] Identify binary plugins
In-Reply-To: <3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>
References: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
	<3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>
Message-ID: <f292e57-3ebb-10de-9df8-8927f4eab0b8@elronnd.net>

Afaik, by convention, the file extension for shared libraries is .dylib on 
macos and .dll on windows.

On Wed, 23 Mar 2022, bil at ccrma.Stanford.EDU wrote:

> Thanks, but isn't ELF format specific to Linux?
> My *.so files on a Mac appear to be something else.
> Also, can't an ELF file be an executable?
>
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist
>


From bil at ccrma.Stanford.EDU  Wed Mar 23 15:16:16 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Wed, 23 Mar 2022 15:16:16 -0700
Subject: [CM] Identify binary plugins
In-Reply-To: <f292e57-3ebb-10de-9df8-8927f4eab0b8@elronnd.net>
References: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
	<3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>
	<f292e57-3ebb-10de-9df8-8927f4eab0b8@elronnd.net>
Message-ID: <6ab9f8a2fd849647f2a3323dfe716425@ccrma.stanford.edu>

I'll add dylib as an acceptable extension.



From wdouglass at carnegierobotics.com  Thu Mar 24 05:05:35 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Thu, 24 Mar 2022 12:05:35 +0000
Subject: [CM] Identify binary plugins
In-Reply-To: <6ab9f8a2fd849647f2a3323dfe716425@ccrma.stanford.edu>
References: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
	<3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>
	<f292e57-3ebb-10de-9df8-8927f4eab0b8@elronnd.net>
	<6ab9f8a2fd849647f2a3323dfe716425@ccrma.stanford.edu>
Message-ID: <8968e7ff586a05f8300909404489c311401c0ec8.camel@carnegierobotics.com>

one more pass at this; dlopen uses the "magic numbers" at the beginning
of a file to identify the executable format; i think it's safe for s7
to do the same. I've modified the patch to include MacOS and Windows
exectuable file formats.

as an aside, ELF files can be executable! and dlopen can load them just
like a dll (as long as they have a dynamic symbol table). I can't see a
reason for s7 to take advantage of this, but I also can't see a reason
to prevent someone from loading an executable as an s7 extension either

Take a look at the revised patch, attached.

Thanks,
Woody Douglass

On Wed, 2022-03-23 at 15:16 -0700, bil at ccrma.Stanford.EDU wrote:
> I'll add dylib as an acceptable extension.
> 
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0004-identify-elf.patch
Type: text/x-patch
Size: 2080 bytes
Desc: 0004-identify-elf.patch
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20220324/4abc218c/attachment.bin>

From bil at ccrma.Stanford.EDU  Thu Mar 24 06:21:16 2022
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Thu, 24 Mar 2022 06:21:16 -0700
Subject: [CM] Identify binary plugins
In-Reply-To: <8968e7ff586a05f8300909404489c311401c0ec8.camel@carnegierobotics.com>
References: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
	<3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>
	<f292e57-3ebb-10de-9df8-8927f4eab0b8@elronnd.net>
	<6ab9f8a2fd849647f2a3323dfe716425@ccrma.stanford.edu>
	<8968e7ff586a05f8300909404489c311401c0ec8.camel@carnegierobotics.com>
Message-ID: <fd56a20cb51857fdf70c04b58b76e01e@ccrma.stanford.edu>

Thanks again, but I am not comfortable with this.
It means building in magic numbers and assuming
that everyone uses them, but I'd like s7 to work
anywhere there's a C compiler and (in this context)
dlopen, and be able to handle changes in the future
without maintaining an ever-growing list of magic
numbers.  (And it means any file starting with "MZ"
will be treated as an object file).



From wdouglass at carnegierobotics.com  Thu Mar 24 06:28:26 2022
From: wdouglass at carnegierobotics.com (Woody Douglass)
Date: Thu, 24 Mar 2022 13:28:26 +0000
Subject: [CM] Identify binary plugins
In-Reply-To: <fd56a20cb51857fdf70c04b58b76e01e@ccrma.stanford.edu>
References: <cdde56ff41f45c19c48780f3a87abdeaf3bd7bca.camel@carnegierobotics.com>
	<3470ba06a7a017ef23611b03355f8b72@ccrma.stanford.edu>
	<f292e57-3ebb-10de-9df8-8927f4eab0b8@elronnd.net>
	<6ab9f8a2fd849647f2a3323dfe716425@ccrma.stanford.edu>
	<8968e7ff586a05f8300909404489c311401c0ec8.camel@carnegierobotics.com>
	<fd56a20cb51857fdf70c04b58b76e01e@ccrma.stanford.edu>
Message-ID: <65664ddf6a93304d026f2e7addfa7acb153e9da1.camel@carnegierobotics.com>

That's fair, thanks for looking!

-Woody Douglass

On Thu, 2022-03-24 at 06:21 -0700, bil at ccrma.Stanford.EDU wrote:
> Thanks again, but I am not comfortable with this.
> It means building in magic numbers and assuming
> that everyone uses them, but I'd like s7 to work
> anywhere there's a C compiler and (in this context)
> dlopen, and be able to handle changes in the future
> without maintaining an ever-growing list of magic
> numbers.  (And it means any file starting with "MZ"
> will be treated as an object file).
>