From ko1 at atdot.net  Sun Oct  2 15:43:47 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 03 Oct 2005 04:43:47 +0900
Subject: [Yarv-devel] YARV Compile and Disassemble CGI
Message-ID: <43403873.1060407@atdot.net>

Hi,

I made a cgi script that compile and disassemble ruby script to YARV
instruction sequence.

http://www.atdot.net/yc/

If there is any question or any bug, please ask me.

Enjoy.
-- 
// SASADA Koichi at atdot dot net
//


From ko1 at atdot.net  Mon Oct  3 01:27:54 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 03 Oct 2005 14:27:54 +0900
Subject: [Yarv-devel] New massign tests
In-Reply-To: <dgvmk5$afm$1@sea.gmane.org>
References: <CDA50CAC-1659-4A18-B4CB-BF0DCFF5BD91@ntlworld.com>	<4329D179.7060302@atdot.net>
	<dgvmk5$afm$1@sea.gmane.org>
Message-ID: <4340C15A.6050906@atdot.net>

Hi,

? wrote:
> SASADA Koichi wrote:
> 
>> Should massign syntax return array?  This behaviour is high runtime 
>> cost and we use these syntax at only few cases.
> 
> 
> Hm, can you detect when it is done in a void context? (Similar to ruby 
> -we "1".) I would imagine it would keep the run-time cost low in most 
> cases -- it would only cause less performance if the return value is 
> potentially significant. (Like when a massign is the last statement of a 
> method.)
> 
> Sorry if it is not as simple as that.
> 
It's easy to detect void context or not.  Yes, your solution is best for
save compatibility.  But it is doubtful that this specification is good
language design (especially, introducing multi-value).

-- 
// SASADA Koichi at atdot dot net
//


From ko1 at atdot.net  Sun Oct  9 02:30:34 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Sun, 09 Oct 2005 15:30:34 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
Message-ID: <4348B90A.1060501@atdot.net>

Hi,

I'm making a poster for OOPSLA05 about YARV (A0 x 2).  It's my first
time to make posters for academic place :)

http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf
(now, optimization is empty)

Any comments are welcome.

Thanks,
-- 
SASADA Koichi at atdot dot net

From nohmad at gmail.com  Sun Oct  9 10:50:05 2005
From: nohmad at gmail.com (Gyoung-Yoon Noh)
Date: Sun, 9 Oct 2005 23:50:05 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <4348B90A.1060501@atdot.net>
References: <4348B90A.1060501@atdot.net>
Message-ID: <ff903fda0510090750q3d080506vb58a68943976ea69@mail.gmail.com>

Hi,

I found a typo 'SmapleModule' in 'Ruby Language Introduction' section.
Perhaps you can find it using pdf viewers's search menu.

Keep up the good work,

On 10/9/05, SASADA Koichi <ko1 at atdot.net> wrote:
> Hi,
>
> I'm making a poster for OOPSLA05 about YARV (A0 x 2).  It's my first
> time to make posters for academic place :)
>
> http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf
> (now, optimization is empty)
>
> Any comments are welcome.
>
> Thanks,
> --
> SASADA Koichi at atdot dot net
>


--
http://nohmad.sub-port.net


From ko1 at atdot.net  Sun Oct  9 13:21:16 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 10 Oct 2005 02:21:16 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <ff903fda0510090750q3d080506vb58a68943976ea69@mail.gmail.com>
References: <4348B90A.1060501@atdot.net>
	<ff903fda0510090750q3d080506vb58a68943976ea69@mail.gmail.com>
Message-ID: <4349518C.5030703@atdot.net>

Hi,

Gyoung-Yoon Noh wrote:
> Hi,
> 
> I found a typo 'SmapleModule' in 'Ruby Language Introduction' section.
> Perhaps you can find it using pdf viewers's search menu.
> 
> Keep up the good work,

Thanks.  I fixed it.

-- 
SASADA Koichi at atdot dot net

From daniel.amelang at gmail.com  Sun Oct  9 14:30:41 2005
From: daniel.amelang at gmail.com (Daniel Amelang)
Date: Sun, 9 Oct 2005 11:30:41 -0700
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <4348B90A.1060501@atdot.net>
References: <4348B90A.1060501@atdot.net>
Message-ID: <a65ae5c90510091130ueb90527md424455023f9790e@mail.gmail.com>

1)Looking at your ruby compared to scheme example in "Other Ruby Features":

I would recommend replacing some of the '...' with something more
descriptive. It's hard to tell from the example which '...' is
referring to method parameters, block/lambda parameters and
block/lambda body. I recommend something like this:

method_name(method_args) { |lambda_args|
  ...
}

(method_name method_args
   (lambda (lambda_args) ...))

Just an idea. You can see that I'm trying to make it as easy as
possible to see the similarities (even naming the block_args
'lambda_args', which is not totally correct, but I think appropriate
for the context).

2) Same idea with the Java/Ruby comparison. In the Java example, you
have "// ..." as the bodies, while the ruby example has one or two
lines of descriptive comments. I would just keep it simple and use '#
... ' in the ruby also. That way you emphasize the similarity and
don't introduce unnecessary differences that detract from your
purpose.

On the same note, consider putting the Java example on the left and
the ruby on on the right (same with the scheme example), because
people won't understand the ruby examples until they see the
java/scheme ones *first*. So make it easier and put the familiar ones
first (on the left, like English is read).

So you'd have this:

// Java exception handling
try {
    // ...
}
catch (SomeError e) {
    // ...
}
catch (Exception e) {
    // ...
}
finally {
    // ...
}

# Ruby exception handling
begin
  # ...
rescue SomeError => e
  # ...
rescue => e
  # ...
ensure
  # ...
end

Again, just to make the similarities stand out.

3) Is userlevel threads really a feature? :)

4) You might want to mention other ruby libraries that are popular
like distributed computing, parallel computing (MPI), SOAP, XMLRPC.
These names get peoples attention more than just 'Network' or 'System
Management', in my opinion.

5) 'Evil eval' should just be 'eval' I think. The issue of eval being
evil is not a introductory topic, in my opinion.

6) 'grand gaol' should be 'grand goal'

7) 'these bars shows' should be 'these bars show'

8) I don't understand the sentence: 'The results show that the
applications which
have VM depends bottleneck is speed up obviously.' Try to rephrase it.
It seems to be an important point, but it's not comprehensible.

Great work! See you at RubyConf and OOPSLA.

Dan Amelang


From wilsonb at gmail.com  Sun Oct  9 21:14:12 2005
From: wilsonb at gmail.com (Wilson Bilkovich)
Date: Sun, 9 Oct 2005 21:14:12 -0400
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <4348B90A.1060501@atdot.net>
References: <4348B90A.1060501@atdot.net>
Message-ID: <d4e4955b0510091814n19b1151emcadc912b5884cc09@mail.gmail.com>

Looks good. I found a couple more issues with the wording, though:

Page 1:
>># if exception was occur or not
Should be "if an exception has occured" or similar.

>>Smalltalk, Self and other dynamic OO language
Should be "languages"

Page 2:
>>The results show that the applications which
>>have VM depends bottleneck is speed up
>>obviously. But other application (GC, Object
>>Allocation, Regexp, Bignum depends) doesn't
>>speedup on YARV.

This is fairly hard to parse. I'd vote for something like:
The results show that the speed of VM-dependent applications is
greatly increased.
However, other applications (those depending on GC, Object allocation,
Regexp, and Bignum) are not significantly affected by YARV.

Thanks; looking forward to the presentation at RubyConf.

--Wilson.

On 10/9/05, SASADA Koichi <ko1 at atdot.net> wrote:
> Hi,
>
> I'm making a poster for OOPSLA05 about YARV (A0 x 2).  It's my first
> time to make posters for academic place :)
>
> http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf
> (now, optimization is empty)
>
> Any comments are welcome.
>
> Thanks,
> --
> SASADA Koichi at atdot dot net
>


From ko1 at atdot.net  Sun Oct  9 23:37:04 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 10 Oct 2005 12:37:04 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <a65ae5c90510091130ueb90527md424455023f9790e@mail.gmail.com>
References: <4348B90A.1060501@atdot.net>
	<a65ae5c90510091130ueb90527md424455023f9790e@mail.gmail.com>
Message-ID: <4349E1E0.8050001@atdot.net>

Hi,

I reflected your comments.

Thanks,
-- 
SASADA Koichi at atdot dot net

From ko1 at atdot.net  Sun Oct  9 23:38:17 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 10 Oct 2005 12:38:17 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <d4e4955b0510091814n19b1151emcadc912b5884cc09@mail.gmail.com>
References: <4348B90A.1060501@atdot.net>
	<d4e4955b0510091814n19b1151emcadc912b5884cc09@mail.gmail.com>
Message-ID: <4349E229.3020104@atdot.net>

Hi,

I reflected your comments.


reviesed version:
http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf

Regards,
-- 
SASADA Koichi at atdot dot net

From surrender_it at yahoo.it  Mon Oct 10 03:19:05 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Mon, 10 Oct 2005 00:19:05 -0700 (PDT)
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <4348B90A.1060501@atdot.net>
Message-ID: <20051010071906.13370.qmail@web26207.mail.ukl.yahoo.com>



--- SASADA Koichi <ko1 at atdot.net> wrote:

> Hi,
> 
> I'm making a poster for OOPSLA05 about YARV (A0 x
> 2).  It's my first
> time to make posters for academic place :)
> 
> http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf
> (now, optimization is empty)
> 
> Any comments are welcome.
> 
> Thanks,


I am not a native english speaker but I think the
lines - in "future work":
  "Now YARV can almost Ruby syntax, native thread
support"  

- in "optimization for vm speedup"
  "Create new instructions which are embedded specific
operands or combined some instructions" 

sounds a little bit wrong. 

I hope a real english speaking person can check this
:) 

icq #69488917


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

From sdate at everestkc.net  Mon Oct 10 06:49:41 2005
From: sdate at everestkc.net (Shashank Date)
Date: Mon, 10 Oct 2005 05:49:41 -0500
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <4348B90A.1060501@atdot.net>
References: <4348B90A.1060501@atdot.net>
Message-ID: <434A4745.4000500@everestkc.net>

Hi,

SASADA Koichi wrote:

>http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf
>(now, optimization is empty)
>

I have not followed all the changes suggested by others, so these may be
a repetitions.

On Page 2:

1. In "Project Objective", the statement reads:

To solve this problem, several virtual _machines has_ been developed,
but none has achieved adequate performance...

It should be:

To solve this problem, several virtual _machines have_ been developed,
but none has achieved adequate performance...

2. In "Optimization of VM Speedup", under the bullet point: Operands,
Instructions Unification (a.k.a. Super Instruction),
the point is:

Create new instructions which are embedded specific operands or combined
some instructions.

It is not clear to me what you want to say about the new instructions.
But by looking at the context, my guess is:

Create new instructions which embed specific operands or combine some
instructions.

3. Down below under "Evaluation", the statement reads:

Right two graphs show results of YARV speedup ratio compared to current
Ruby interpreter.

I would put it as:

The two graphs on the right show results of YARV speedup ratio compared
to current Ruby interpreter.

4. In the same box on the left, under "Status and Future work", the
Issues list:

Native Thread support and run Ruby program parallel

It should be:

Native Thread support and run Ruby programs in parallel

5. Under the Issues list again:

AOT (Ahead-of-Time) Compilation completely

It should be just:

AOT (Ahead-of-Time) Compilation

Or:

Complete AOT (Ahead-of-Time) Compilation

6. Finally,

Apply other optimizations using example from Smalltalk, Self and other
dynamic OO languages.

Should be:

Apply other optimizations using examples from Smalltalk, Self and other
dynamic OO languages.

The word example should be plural. Unless, of course, there is only one
example. :-)

Hope this helps .... wish you the very best at RubyConf 05 and OOPSLA'05.

Please consider writting for "Ruby Code and Style", the online Enlgish
zine we launched today (10/10/2005).

Thanks,
-- Shashank
















From sdate at everestkc.net  Mon Oct 10 07:02:18 2005
From: sdate at everestkc.net (Shashank Date)
Date: Mon, 10 Oct 2005 06:02:18 -0500
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <434A4745.4000500@everestkc.net>
References: <4348B90A.1060501@atdot.net> <434A4745.4000500@everestkc.net>
Message-ID: <434A4A3A.2040003@everestkc.net>

Hi,

Shashank Date wrote:

>I have not followed all the changes suggested by others, so these may be
>a repetitions.
>  
>

You won't feel confident about my recommendations if I'm making mistakes
in my own email ;-)
Should have been: "... so these may be repetitions."

>Please consider writting for "Ruby Code and Style", the online Enlgish
>zine we launched today (10/10/2005).
>  
>
Forgot to mention the link:

http://www.artima.com/rubycs/index.html

Thanks for listening,

-- Shashank




From kashia at vfemail.net  Mon Oct 10 08:21:08 2005
From: kashia at vfemail.net (Kashia Buch)
Date: Mon, 10 Oct 2005 14:21:08 +0200
Subject: [Yarv-devel] evalc.patch
Message-ID: <op.sye93j0wg3glhz@jo>

Hi,

this happened after applying the evalc.patch to current ruby CVS HEAD  
"2005-10-09":

jo:~/build/ruby john$ patch -p0 < evalc.patch
patching file eval.c
Hunk #2 succeeded at 905 with fuzz 2 (offset -61 lines).
Hunk #3 succeeded at 921 (offset -61 lines).
Hunk #4 succeeded at 956 (offset -60 lines).
Hunk #5 succeeded at 4557 with fuzz 2 (offset -167 lines).
Hunk #6 succeeded at 4946 with fuzz 2 (offset -174 lines).
Hunk #7 succeeded at 4961 (offset -180 lines).
Hunk #8 succeeded at 5175 (offset -191 lines).
Hunk #9 succeeded at 5651 with fuzz 1 (offset -238 lines).
Hunk #10 succeeded at 5743 (offset -236 lines).
Hunk #11 succeeded at 6007 with fuzz 2 (offset -224 lines).
Hunk #12 succeeded at 6026 (offset -224 lines).
Hunk #13 succeeded at 6036 (offset -225 lines).
Hunk #14 succeeded at 6600 (offset -254 lines).
Hunk #15 succeeded at 7713 (offset -324 lines).

jo:~/build/ruby john$ make

...

gcc -g -O2  -fno-common -pipe -fno-common   -I. -I.  -c eval.c
eval.c: In function `rb_call0':
eval.c:5659: error: `nosuper' undeclared (first use in this function)
eval.c:5659: error: (Each undeclared identifier is reported only once
eval.c:5659: error: for each function it appears in.)
eval.c:5659: warning: passing arg 6 of pointer to function discards  
qualifiers from pointer target type
make: *** [eval.o] Error 1

the lines in which the error occures:

if(IS_YARV_WORKING()){
     if(nd_type(body) == NODE_SCOPE){
       rb_warning("Ruby method is invoked under YARV environment");
     }
     else{
       return YARV_CALL0(klass, recv, id, oid, argc, argv, body, nosuper);
     }
   }

nosuper isn't defined in that function, I just added a "int nosuper = 0"  
before that call, it compiled flawlessly, but I somehow doubt that this is  
really correct.


Kash
-- 
Feel the love
http://pinkjuice.com/pics/ruby.png

From kashia at vfemail.net  Mon Oct 10 08:21:24 2005
From: kashia at vfemail.net (Kashia Buch)
Date: Mon, 10 Oct 2005 14:21:24 +0200
Subject: [Yarv-devel] evalc.patch
Message-ID: <op.syfi9yq5g3glhz@jo>

Hi, me again, next error:

gcc main.o  libruby1.9-static.a -lpthread -ldl -lobjc   -o miniruby -g  
-O2  -fno-common -pipe -fno-common
ld: Undefined symbols:
_va_init_list
make: *** [miniruby] Error 1

which should be defined somehere..

jo:~/build/ruby john$ grep -r va_init_list *
eval.c:    va_init_list(args, n);
Binary file eval.o matches
evalc.patch:+    va_init_list(args, n);
...
Binary file libruby1.9-static.a matches

so, it seems to be the patches fault again.

jo:~/build/ruby john$ nm eval.o | grep init_list
          U _va_init_list

so, Undefined symbol in eval.o....

in ext/tk/tcltklib.c:#define va_init_list(a,b) va_start(a,b)

so, adding this #define to eval.c did the trick, but again, I'm not sure  
weither this is the intended behaviour...

Kash

-- 
Feel the love
http://pinkjuice.com/pics/ruby.png

From ko1 at atdot.net  Mon Oct 10 09:28:11 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 10 Oct 2005 22:28:11 +0900
Subject: [Yarv-devel] evalc.patch
In-Reply-To: <op.sye93j0wg3glhz@jo>
References: <op.sye93j0wg3glhz@jo>
Message-ID: <434A6C6B.8080909@atdot.net>

Hi,

I'm Sorry, but I have not maintained evalc.patch.  Please use non-patch 
version (Ruby full set).

I'll remove eval.c patch next release (0.3.2).


Kashia Buch wrote:
> Hi,
> 
> this happened after applying the evalc.patch to current ruby CVS HEAD  
> "2005-10-09":
> 
> jo:~/build/ruby john$ patch -p0 < evalc.patch
> patching file eval.c
> Hunk #2 succeeded at 905 with fuzz 2 (offset -61 lines).
> Hunk #3 succeeded at 921 (offset -61 lines).
> Hunk #4 succeeded at 956 (offset -60 lines).
> Hunk #5 succeeded at 4557 with fuzz 2 (offset -167 lines).
> Hunk #6 succeeded at 4946 with fuzz 2 (offset -174 lines).
> Hunk #7 succeeded at 4961 (offset -180 lines).
> Hunk #8 succeeded at 5175 (offset -191 lines).
> Hunk #9 succeeded at 5651 with fuzz 1 (offset -238 lines).
> Hunk #10 succeeded at 5743 (offset -236 lines).
> Hunk #11 succeeded at 6007 with fuzz 2 (offset -224 lines).
> Hunk #12 succeeded at 6026 (offset -224 lines).
> Hunk #13 succeeded at 6036 (offset -225 lines).
> Hunk #14 succeeded at 6600 (offset -254 lines).
> Hunk #15 succeeded at 7713 (offset -324 lines).
> 
> jo:~/build/ruby john$ make
> 
> ...
> 
> gcc -g -O2  -fno-common -pipe -fno-common   -I. -I.  -c eval.c
> eval.c: In function `rb_call0':
> eval.c:5659: error: `nosuper' undeclared (first use in this function)
> eval.c:5659: error: (Each undeclared identifier is reported only once
> eval.c:5659: error: for each function it appears in.)
> eval.c:5659: warning: passing arg 6 of pointer to function discards  
> qualifiers from pointer target type
> make: *** [eval.o] Error 1
> 
> the lines in which the error occures:
> 
> if(IS_YARV_WORKING()){
>     if(nd_type(body) == NODE_SCOPE){
>       rb_warning("Ruby method is invoked under YARV environment");
>     }
>     else{
>       return YARV_CALL0(klass, recv, id, oid, argc, argv, body, nosuper);
>     }
>   }
> 
> nosuper isn't defined in that function, I just added a "int nosuper = 
> 0"  before that call, it compiled flawlessly, but I somehow doubt that 
> this is  really correct.
> 
> 
> Kash


-- 
SASADA Koichi at atdot dot net

From ko1 at atdot.net  Mon Oct 10 09:31:55 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Mon, 10 Oct 2005 22:31:55 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <434A4745.4000500@everestkc.net>
References: <4348B90A.1060501@atdot.net> <434A4745.4000500@everestkc.net>
Message-ID: <434A6D4B.5020001@atdot.net>

Hi,

Thanks.
I reflected your comments.

Shashank Date wrote:
> Hi,
> 
> SASADA Koichi wrote:
> 
> 
>>http://www.atdot.net/~ko1/tmp/OOPSLA05_poster.pdf
>>(now, optimization is empty)
>>
> 
> 
> I have not followed all the changes suggested by others, so these may be
> a repetitions.
> 
> On Page 2:
> 
> 1. In "Project Objective", the statement reads:
> 
> To solve this problem, several virtual _machines has_ been developed,
> but none has achieved adequate performance...
> 
> It should be:
> 
> To solve this problem, several virtual _machines have_ been developed,
> but none has achieved adequate performance...
> 
> 2. In "Optimization of VM Speedup", under the bullet point: Operands,
> Instructions Unification (a.k.a. Super Instruction),
> the point is:
> 
> Create new instructions which are embedded specific operands or combined
> some instructions.
> 
> It is not clear to me what you want to say about the new instructions.
> But by looking at the context, my guess is:
> 
> Create new instructions which embed specific operands or combine some
> instructions.
> 
> 3. Down below under "Evaluation", the statement reads:
> 
> Right two graphs show results of YARV speedup ratio compared to current
> Ruby interpreter.
> 
> I would put it as:
> 
> The two graphs on the right show results of YARV speedup ratio compared
> to current Ruby interpreter.
> 
> 4. In the same box on the left, under "Status and Future work", the
> Issues list:
> 
> Native Thread support and run Ruby program parallel
> 
> It should be:
> 
> Native Thread support and run Ruby programs in parallel
> 
> 5. Under the Issues list again:
> 
> AOT (Ahead-of-Time) Compilation completely
> 
> It should be just:
> 
> AOT (Ahead-of-Time) Compilation
> 
> Or:
> 
> Complete AOT (Ahead-of-Time) Compilation
> 
> 6. Finally,
> 
> Apply other optimizations using example from Smalltalk, Self and other
> dynamic OO languages.
> 
> Should be:
> 
> Apply other optimizations using examples from Smalltalk, Self and other
> dynamic OO languages.
> 
> The word example should be plural. Unless, of course, there is only one
> example. :-)
> 
> Hope this helps .... wish you the very best at RubyConf 05 and OOPSLA'05.



> Please consider writting for "Ruby Code and Style", the online Enlgish
> zine we launched today (10/10/2005).
> 

Congratulation for your web-zine startup.
Sure :)  And we can collaborate about Rubyist Magazine and Ruby-CS


> Thanks,
> -- Shashank




Regards,
-- 
SASADA Koichi at atdot dot net

From kashia at vfemail.net  Mon Oct 10 10:02:31 2005
From: kashia at vfemail.net (Kashia Buch)
Date: Mon, 10 Oct 2005 16:02:31 +0200
Subject: [Yarv-devel] evalc.patch
In-Reply-To: <434A6C6B.8080909@atdot.net>
References: <op.sye93j0wg3glhz@jo> <434A6C6B.8080909@atdot.net>
Message-ID: <op.syfnyhwmg3glhz@jo>

Hi,

so I guess the patch isn't needed anymore, better anyway.

thanks

Kash

On Mon, 10 Oct 2005 15:28:11 +0200, SASADA Koichi <ko1 at atdot.net> wrote:
> Hi,
>
> I'm Sorry, but I have not maintained evalc.patch.  Please use non-patch  
> version (Ruby full set).
>
> I'll remove eval.c patch next release (0.3.2).
>
>


-- 
Feel the love
http://pinkjuice.com/pics/ruby.png

From ko1 at atdot.net  Wed Oct 12 04:39:16 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Wed, 12 Oct 2005 17:39:16 +0900
Subject: [Yarv-devel] making OOPSLA05 Poster
In-Reply-To: <4348B90A.1060501@atdot.net>
References: <4348B90A.1060501@atdot.net>
Message-ID: <434CCBB4.40209@atdot.net>

Hi,

I Thank everyone who give me comments about my poster.
I printed out my posters (A0 is too big!!) yesterday.

Now I'm practicing my rubyconf presentation.

See you at RubyConf2005!

Regards,
-- 
// SASADA Koichi at atdot dot net
//

From daniel at brockman.se  Sun Oct  9 18:31:41 2005
From: daniel at brockman.se (Daniel Brockman)
Date: Mon, 10 Oct 2005 00:31:41 +0200
Subject: [Yarv-devel] making OOPSLA05 Poster
References: <4348B90A.1060501@atdot.net>
Message-ID: <87slvajp9u.fsf@wigwam.deepwood.net>

In the sample code regarding classes and methods, there is a
comment that says "overload".  The correct term is "override".

-- 
Daniel Brockman <daniel at brockman.se>


From ko1 at atdot.net  Sat Oct 15 11:02:20 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Sun, 16 Oct 2005 00:02:20 +0900
Subject: [Yarv-devel] RubyConf 2005 PDF
Message-ID: <435119FC.7090803@atdot.net>

Hi,

Here is RubyConf 2005 my presentation PDF:

YARV Progress Report
http://www.atdot.net/yarv/RubyConf2005_ko1_YARV.pdf

Thank you very much for your (audiences) attention.


Regards,
--
SASADA Koichi

From aurelianocalvo at yahoo.com.ar  Tue Oct 18 18:09:11 2005
From: aurelianocalvo at yahoo.com.ar (Aureliano Calvo)
Date: Tue, 18 Oct 2005 19:09:11 -0300 (ART)
Subject: [Yarv-devel] Prevent operations in yarv
Message-ID: <20051018220911.23091.qmail@web50405.mail.yahoo.com>

Hi all,
I'm trying to implement a security manager for ruby
like the java security manager. In order to achieve
it, I need to be able to selectively prevent some
things to ocurr:

* Method definition
* Method undefinition
* Method redefinition
* Constant definition
* Constant modification
* Constant undefinition
* Module/Class definition
* Module/Class modification
* Module/Class undefinition

I've tried to implement a "pure ruby" solution, but
I've found that it's impossible to prevent the
definition, redefinition and undefinition of a
constant and it's very very tricky to prevent the
other operations (for instance method redefinition can
be prevented by aliasing every method and overloading
the method_added method to go back to the original
implementation). 

What would be nice is for ruby to have built in
something like the feature described in
http://www.rubygarden.org/ruby?EventOrientedRuby.

Is it supported by yarv/rite?

Can something like it be implemented with a "pure
jarv" solution? (i.e. without modifying the
interpreter or adding C modules).

Thank you very much for your time,
Aureliano.


	


	
		
___________________________________________________________ 
1GB gratis, Antivirus y Antispam 
Correo Yahoo!, el mejor correo web del mundo 
http://correo.yahoo.com.ar 


From ko1 at atdot.net  Mon Oct 24 20:01:31 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Tue, 25 Oct 2005 09:01:31 +0900
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <20051018220911.23091.qmail@web50405.mail.yahoo.com>
References: <20051018220911.23091.qmail@web50405.mail.yahoo.com>
Message-ID: <435D75DB.80306@atdot.net>

Hi,

I want to support re-writing instruction sequence on Ruby level.

ex)
   [ruby program] ->
   YARV compiler ->
   [YARV instruction sequence (base)] ->
   Ruby program (you can add it) ->
   [YARV instruction sequence (changed)]

Using this feature, I think you can make security manager.  Of course, 
you must know YARV instruction set.

Aureliano Calvo wrote:
> Hi all,
> I'm trying to implement a security manager for ruby
> like the java security manager. In order to achieve
> it, I need to be able to selectively prevent some
> things to ocurr:
> 
> * Method definition
> * Method undefinition
> * Method redefinition
> * Constant definition
> * Constant modification
> * Constant undefinition
> * Module/Class definition
> * Module/Class modification
> * Module/Class undefinition
> 
> I've tried to implement a "pure ruby" solution, but
> I've found that it's impossible to prevent the
> definition, redefinition and undefinition of a
> constant and it's very very tricky to prevent the
> other operations (for instance method redefinition can
> be prevented by aliasing every method and overloading
> the method_added method to go back to the original
> implementation). 
> 
> What would be nice is for ruby to have built in
> something like the feature described in
> http://www.rubygarden.org/ruby?EventOrientedRuby.
> 
> Is it supported by yarv/rite?
> 
> Can something like it be implemented with a "pure
> jarv" solution? (i.e. without modifying the
> interpreter or adding C modules).
> 
> Thank you very much for your time,
> Aureliano.
> 
> 
> 	
> 
> 
> 	
> 		
> ___________________________________________________________ 
> 1GB gratis, Antivirus y Antispam 
> Correo Yahoo!, el mejor correo web del mundo 
> http://correo.yahoo.com.ar 
> 


-- 
SASADA Koichi at atdot dot net

From aurelianocalvo at yahoo.com.ar  Mon Oct 24 22:33:35 2005
From: aurelianocalvo at yahoo.com.ar (Aureliano Calvo)
Date: Tue, 25 Oct 2005 02:33:35 +0000 (GMT)
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <435D75DB.80306@atdot.net>
Message-ID: <20051025023335.36869.qmail@web50411.mail.yahoo.com>

Is there a way to modify the YARV instruction sequence
in runtime? I would like to modify the instructions of
an eval statement.

Is this feature implemented or is on the to do list?

Thank's in advance,
Aureliano.

> Hi,
> 
> I want to support re-writing instruction sequence on
> Ruby level.
> 
> ex)
>    [ruby program] ->
>    YARV compiler ->
>    [YARV instruction sequence (base)] ->
>    Ruby program (you can add it) ->
>    [YARV instruction sequence (changed)]
> 
> Using this feature, I think you can make security
> manager.  Of course, 
> you must know YARV instruction set.
> 
> Aureliano Calvo wrote:
> > Hi all,
> > I'm trying to implement a security manager for
> ruby
> > like the java security manager. In order to
> achieve
> > it, I need to be able to selectively prevent some
> > things to ocurr:
> > 
> > * Method definition
> > * Method undefinition
> > * Method redefinition
> > * Constant definition
> > * Constant modification
> > * Constant undefinition
> > * Module/Class definition
> > * Module/Class modification
> > * Module/Class undefinition
> > 
> > I've tried to implement a "pure ruby" solution,
> but
> > I've found that it's impossible to prevent the
> > definition, redefinition and undefinition of a
> > constant and it's very very tricky to prevent the
> > other operations (for instance method redefinition
> can
> > be prevented by aliasing every method and
> overloading
> > the method_added method to go back to the original
> > implementation). 
> > 
> > What would be nice is for ruby to have built in
> > something like the feature described in
> > http://www.rubygarden.org/ruby?EventOrientedRuby.
> > 
> > Is it supported by yarv/rite?
> > 
> > Can something like it be implemented with a "pure
> > jarv" solution? (i.e. without modifying the
> > interpreter or adding C modules).
> > 
> > Thank you very much for your time,
> > Aureliano.
> > 
> > 
> > 	
> > 
> > 
> > 	
> > 		
> >
>
___________________________________________________________
> 
> > 1GB gratis, Antivirus y Antispam 
> > Correo Yahoo!, el mejor correo web del mundo 
> > http://correo.yahoo.com.ar 
> > 
> 
> 
> -- 
> SASADA Koichi at atdot dot net
> 



	


	
		
___________________________________________________________ 
1GB gratis, Antivirus y Antispam 
Correo Yahoo!, el mejor correo web del mundo 
http://correo.yahoo.com.ar 


From ko1 at atdot.net  Tue Oct 25 01:19:21 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Tue, 25 Oct 2005 14:19:21 +0900
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <20051025023335.36869.qmail@web50411.mail.yahoo.com>
References: <20051025023335.36869.qmail@web50411.mail.yahoo.com>
Message-ID: <435DC059.70004@atdot.net>

Hi,

Aureliano Calvo wrote:
> Is there a way to modify the YARV instruction sequence
> in runtime? I would like to modify the instructions of
> an eval statement.
> 
> Is this feature implemented or is on the to do list?

on to do list :)

Regards,
-- 
// SASADA Koichi at atdot dot net
//


From wilsonb at gmail.com  Tue Oct 25 12:47:20 2005
From: wilsonb at gmail.com (Wilson Bilkovich)
Date: Tue, 25 Oct 2005 12:47:20 -0400
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <435DC059.70004@atdot.net>
References: <20051025023335.36869.qmail@web50411.mail.yahoo.com>
	<435DC059.70004@atdot.net>
Message-ID: <d4e4955b0510250947t2340f4faxe6855d0ec554a453@mail.gmail.com>

Does this mean that you could use RubyInline to write YARV VM
instructions directly, as well?
Some people are doing very interesting things by writing Java bytecode by hand.

On 10/25/05, SASADA Koichi <ko1 at atdot.net> wrote:
> Hi,
>
> Aureliano Calvo wrote:
> > Is there a way to modify the YARV instruction sequence
> > in runtime? I would like to modify the instructions of
> > an eval statement.
> >
> > Is this feature implemented or is on the to do list?
>
> on to do list :)
>
> Regards,
> --
> // SASADA Koichi at atdot dot net
> //
>
>


From surrender_it at yahoo.it  Tue Oct 25 13:55:04 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Tue, 25 Oct 2005 10:55:04 -0700 (PDT)
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <d4e4955b0510250947t2340f4faxe6855d0ec554a453@mail.gmail.com>
Message-ID: <20051025175504.98115.qmail@web26201.mail.ukl.yahoo.com>


--- Wilson Bilkovich <wilsonb at gmail.com> ha scritto: 

> Does this mean that you could use RubyInline to
> write YARV VM
> instructions directly, as well?
> Some people are doing very interesting things by
> writing Java bytecode by hand.

I wonder how much of that is needed in javaland but
possible in ruby with metaprogramming :)

icq #69488917


	

	
		
___________________________________ 
Yahoo! Mail: gratis 1GB per i messaggi e allegati da 10MB 
http://mail.yahoo.it

From aurelianocalvo at yahoo.com.ar  Tue Oct 25 17:25:28 2005
From: aurelianocalvo at yahoo.com.ar (Aureliano Calvo)
Date: Tue, 25 Oct 2005 18:25:28 -0300 (ART)
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <20051025175504.98115.qmail@web26201.mail.ukl.yahoo.com>
Message-ID: <20051025212528.50079.qmail@web50401.mail.yahoo.com>

> > Does this mean that you could use RubyInline to
> > write YARV VM
> > instructions directly, as well?
> > Some people are doing very interesting things by
> > writing Java bytecode by hand.
> 
> I wonder how much of that is needed in javaland but
> possible in ruby with metaprogramming :)
> 
> icq #69488917

I agree that there are a lot of tasks that can be
written in plain ruby but not in plain java, but I've
found one thing that can't (prevent method/constant
redefinition). I also believe that might be possible
to give ruby better AOP support with it.

Aureliano.


	


	
		
___________________________________________________________ 
1GB gratis, Antivirus y Antispam 
Correo Yahoo!, el mejor correo web del mundo 
http://correo.yahoo.com.ar 


From ko1 at atdot.net  Tue Oct 25 17:26:41 2005
From: ko1 at atdot.net (SASADA Koichi)
Date: Wed, 26 Oct 2005 06:26:41 +0900
Subject: [Yarv-devel] Prevent operations in yarv
In-Reply-To: <d4e4955b0510250947t2340f4faxe6855d0ec554a453@mail.gmail.com>
References: <20051025023335.36869.qmail@web50411.mail.yahoo.com>	<435DC059.70004@atdot.net>
	<d4e4955b0510250947t2340f4faxe6855d0ec554a453@mail.gmail.com>
Message-ID: <435EA311.7020902@atdot.net>

Hi,

I can't imagine this feature with RubyInline.  How do you use it?


Wilson Bilkovich wrote:
> Does this mean that you could use RubyInline to write YARV VM
> instructions directly, as well?
> Some people are doing very interesting things by writing Java bytecode by hand.
> 
> On 10/25/05, SASADA Koichi <ko1 at atdot.net> wrote:
> 
>>Hi,
>>
>>Aureliano Calvo wrote:
>>
>>>Is there a way to modify the YARV instruction sequence
>>>in runtime? I would like to modify the instructions of
>>>an eval statement.
>>>
>>>Is this feature implemented or is on the to do list?
>>
>>on to do list :)
>>
>>Regards,
>>--
>>// SASADA Koichi at atdot dot net
>>//
>>
>>
> 
> 


-- 
SASADA Koichi at atdot dot net

From surrender_it at yahoo.it  Fri Oct 28 20:29:47 2005
From: surrender_it at yahoo.it (gabriele renzi)
Date: Fri, 28 Oct 2005 17:29:47 -0700 (PDT)
Subject: [Yarv-devel] Warning on win32
Message-ID: <20051029002947.90424.qmail@web26208.mail.ukl.yahoo.com>

Hi list,

I just tried compiling the latest revision on win32,
and I get this warnings:

In file included from thread.c:67:
thread_win32.h: In function `native_thread_crteate':
thread_win32.h:196: warning: passing arg 3 of
`_beginthreadex' from incompatible
 pointer type
thread_win32.h: In function `make_timer_thread':
thread_win32.h:222: warning: passing arg 3 of
`_beginthreadex' from incompatible
 pointer type

which I think means that 
timer_thread_func() and thread_start_func_1() should
be declared as returning "unsigned int" not "int".
 
The point is that all tests pass anyway, whatever the
declaration is, which IMHO means that:
- we can make the compiler happy by changing
declarations
- or the test suite is missing a test that should
cause failing :)

Additional info:
OS: winxp SP2
cc: gcc version 3.4.2 (mingw-special, Thread model:
win32)
CPU: Athlon XP
YARV: revision 283


icq #69488917


	

	
		
___________________________________ 
Yahoo! Mail: gratis 1GB per i messaggi e allegati da 10MB 
http://mail.yahoo.it