thr3ads.net - Rails spinoffs - SAU/Prototype performance [Jan 2007]

If this information is useful, please help other people find it:
Share via:

Colin Mollenhour

2007-Jan-04 12:19 UTC

SAU/Prototype performance

I''ve noticed that the performance of my web apps is very slow when 
dealing with large amounts of data (e.g. generating >100 tree nodes, 
each of which includes InPlaceEditors, Tooltips (my own using 
Prototype/SAU), lots of various nodes built with Builder.node, etc...).  
I started using the new Firebug 1.0 beta''s profiling and discovered a 
few things.
1) Builder.node calls escapeHTML on all attribute values when parsing 
the attributes
2) escapeHTML creates a div node and a text node each time it is called
3) The combination of 1) and 2) yields a massive performance hit

So what is the reason for 1)?  I don''t know of any *standard* attribute
values offhand that use special HTML characters, nor can I think of a 
good reason to. I don''t think it''d be too much to ask for the
programmer
to realize when he may be using HTML characters and have the foresight 
to add .escapeHTML to the string when it is passed.  Maybe there is a 
common case that makes this worth while that I don''t know of?
------------For reference----------
  _attributes: function(attributes) {
    var attrs = [];
    for(attribute in attributes)
      attrs.push((attribute==''className'' ?
''class'' : attribute) +
          ''="'' +
attributes[attribute].toString().escapeHTML() + ''"'');
    return attrs.join(" ");
  },
-----------------------------------

On 2), why can''t the escapeHTML function create a div node and a text 
node once one the first call and reuse it on subsequent calls?  Also, is 
this not a leak?
------------For reference---------
  escapeHTML: function() {
    var div = document.createElement(''div'');
    var text = document.createTextNode(this);
    div.appendChild(text);
    return div.innerHTML;
  },
----------------------------------

Thoughts?

Thanks,
Colin


--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Thomas Fuchs

2007-Jan-05 10:55 UTC

head link

Re: SAU/Prototype performance

Am 04.01.2007 um 13:19 schrieb Colin Mollenhour:
> On 2), why can''t the escapeHTML function create a div node and a  
> text node once one the first call and reuse it on subsequent  
> calls?  Also, is this not a leak?
> ------------For reference---------
>  escapeHTML: function() {
>    var div = document.createElement(''div'');
>    var text = document.createTextNode(this);
>    div.appendChild(text);
>    return div.innerHTML;
>  },
> ----------------------------------
That''s something we''d need to measure. The text node creation
wouldn''t go away, and reusing the div node would mean to empty it out  
before adding the text node. I don''t really think that it would make  
a real difference.

Best,
Thomas

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-05 14:04 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Thomas Fuchs wrote:> Am 04.01.2007 um 13:19 schrieb Colin Mollenhour:
>> On 2), why can''t the escapeHTML function create a div node and
a  text
>> node once one the first call and reuse it on subsequent  calls?  Also,
>> is this not a leak?a) this is veeeery browser dependent :(
b) leak? can''t see a reason why.
> 
> That''s something we''d need to measure.Let''s measure. Note that I threw in two more "old school"
methods
(string manipulation based).


Browser: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1)
Gecko/20061222 Firefox/2.0.0.1
Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
String.split...
       Prototype:   2.829s @   3535/s (n=10000)
Prototype-oneDiv:   3.795s @   2635/s (n=10000)
  String.replace:   0.640s @  15625/s (n=10000)
    String.split:   1.303s @   7675/s (n=10000)
                    Rate Prototype-oneDiv Prototype String.split
String.replace
Prototype-oneDiv  2635/s               --      -25%         -66%
   -83%
       Prototype  3535/s              34%        --         -54%
   -77%
    String.split  7675/s             191%      117%           --
   -51%
  String.replace 15625/s             493%      342%         104%
     --
Tests duration: 8.567 seconds.
Total duration: 13.127 seconds.


Browser: Opera/9.10 (X11; Linux i686; U; en)
Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
String.split...
       Prototype:   0.554s @  18051/s (n=10000)
Prototype-oneDiv:   0.569s @  17575/s (n=10000)
  String.replace:   1.082s @   9242/s (n=10000)
    String.split:   0.355s @  28169/s (n=10000)
                    Rate String.replace Prototype-oneDiv Prototype
String.split
  String.replace  9242/s             --             -47%      -49%
   -67%
Prototype-oneDiv 17575/s            90%               --       -3%
   -38%
       Prototype 18051/s            95%               3%        --
   -36%
    String.split 28169/s           205%              60%       56%
     --
Tests duration: 2.560 seconds.
Total duration: 6.118 seconds.


Browser: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR
2.0.50727; FDM)
Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
String.split...
       Prototype:   2.296s @   4355/s (n=10000)
Prototype-oneDiv:   1.889s @   5294/s (n=10000)
  String.replace:   0.626s @  15974/s (n=10000)
    String.split:   1.110s @   9009/s (n=10000)
                    Rate Prototype Prototype-oneDiv String.split
String.replace
       Prototype  4355/s        --             -18%         -52%
   -73%
Prototype-oneDiv  5294/s       22%               --         -41%
   -67%
    String.split  9009/s      107%              70%           --
   -44%
  String.replace 15974/s      267%             202%          77%
     --
Tests duration: 5.921 seconds.
Total duration: 10.234 seconds.


Browser: Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
String.split...
       Prototype:   2.324s @   4303/s (n=10000)
Prototype-oneDiv:   1.761s @   5679/s (n=10000)
  String.replace:   0.492s @  20325/s (n=10000)
    String.split:   1.350s @   7407/s (n=10000)
                    Rate Prototype Prototype-oneDiv String.split
String.replace
       Prototype  4303/s        --             -24%         -42%
   -79%
Prototype-oneDiv  5679/s       32%               --         -23%
   -72%
    String.split  7407/s       72%              30%           --
   -64%
  String.replace 20325/s      372%             258%         174%
     --
Tests duration: 5.927 seconds.
Total duration: 10.194 seconds.


Browser: Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5
(like Gecko)
Timing 10000 iterations of Prototype-oneDiv, Prototype, String.replace,
String.split...
Prototype-oneDiv:   2.457s @   4070/s (n=10000)
       Prototype:   2.545s @   3929/s (n=10000)
  String.replace:   2.732s @   3660/s (n=10000)
    String.split:   2.557s @   3911/s (n=10000)
                   Rate String.replace String.split Prototype
Prototype-oneDiv
  String.replace 3660/s             --          -6%       -7%
  -10%
    String.split 3911/s             7%           --       -0%
   -4%
       Prototype 3929/s             7%           0%        --
   -3%
Prototype-oneDiv 4070/s            11%           4%        4%
    --
Tests duration: 10.291 seconds.
Total duration: 32.202 seconds.


Test it yourself using your own system/browser:
  http://gfx.neohub.com/benchmark/t/escapeHTML.html

cheers
- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFnlrqtZHp/AYZiNkRAsISAJwNZn6KHQA8Cz64R1xieZjL+unUaQCgsTRI
o1u0l0bFW2stZOZaEFlAfdU=Vkpv
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Christophe Porteneuve

2007-Jan-05 21:51 UTC

head link

Re: SAU/Prototype performance

Marius Feraru a crit :> Browser: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1)
> Gecko/20061222 Firefox/2.0.0.1
...> Tests duration: 8.567 seconds.
> Total duration: 13.127 seconds.
Great page, Marius, but I find oddly faster results in my Firefox 2
(Iceweasel branded, I run Debian):

Browser: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.8.1) Gecko/20061024
Iceweasel/2.0 (Debian-2.0+dfsg-1)
Timing 1000 iterations of Prototype, Prototype-oneDiv, String.replace,
String.split...
       Prototype:   0.162s @   6173/s (n=1000)
Prototype-oneDiv:   0.219s @   4566/s (n=1000)
  String.replace:   0.042s @  23810/s (n=1000)
    String.split:   0.135s @   7407/s (n=1000)
...
Tests duration: 0.558 seconds.
Total duration: 3.720 seconds.

Granted, this is not exactly your version (I use 20061024), but it''s 4+
times faster, with performance comparable to your Opera''s.  Weird!

-- 
Christophe Porteneuve a.k.a. TDD
"[They] did not know it was impossible, so they did it." --Mark Twain
Email: tdd-x+CfDp/qHev2eFz/2MeuCQ@public.gmane.org

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Colin Mollenhour

2007-Jan-06 02:21 UTC

head link

Re: SAU/Prototype performance

Great page, that is a cool utility that I''m sure I will use in the
future!

You''ve shown the "old school" methods are clearly faster for
the top
browsers (FF,IE6/7) which I think would make the change to 
String.replace or String.split worthwhile. (I was able to reproduce your 
results, only mine were a lot faster across the board, what CPU are you 
using?)
Even though String.replace is slower in Opera I''d be in favor of it 
since Opera clearly has some room for improvement with that function so 
I''m guessing this may be improved in the future.

Thanks,
Colin

Marius Feraru wrote:>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Thomas Fuchs wrote:
>> Am 04.01.2007 um 13:19 schrieb Colin Mollenhour:
>>> On 2), why can''t the escapeHTML function create a div node
and a  text
>>> node once one the first call and reuse it on subsequent  calls? 
Also,
>>> is this not a leak?
> a) this is veeeery browser dependent :(
> b) leak? can''t see a reason why.
>
>>
>> That''s something we''d need to measure.
> Let''s measure. Note that I threw in two more "old
school" methods
> (string manipulation based).
>
>
> Browser: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1)
> Gecko/20061222 Firefox/2.0.0.1
> Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
> String.split...
>       Prototype:   2.829s @   3535/s (n=10000)
> Prototype-oneDiv:   3.795s @   2635/s (n=10000)
>  String.replace:   0.640s @  15625/s (n=10000)
>    String.split:   1.303s @   7675/s (n=10000)
>                    Rate Prototype-oneDiv Prototype String.split
> String.replace
> Prototype-oneDiv  2635/s               --      -25%         -66%
>   -83%
>       Prototype  3535/s              34%        --         -54%
>   -77%
>    String.split  7675/s             191%      117%           --
>   -51%
>  String.replace 15625/s             493%      342%         104%
>     --
> Tests duration: 8.567 seconds.
> Total duration: 13.127 seconds.
>
>
> Browser: Opera/9.10 (X11; Linux i686; U; en)
> Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
> String.split...
>       Prototype:   0.554s @  18051/s (n=10000)
> Prototype-oneDiv:   0.569s @  17575/s (n=10000)
>  String.replace:   1.082s @   9242/s (n=10000)
>    String.split:   0.355s @  28169/s (n=10000)
>                    Rate String.replace Prototype-oneDiv Prototype
> String.split
>  String.replace  9242/s             --             -47%      -49%
>   -67%
> Prototype-oneDiv 17575/s            90%               --       -3%
>   -38%
>       Prototype 18051/s            95%               3%        --
>   -36%
>    String.split 28169/s           205%              60%       56%
>     --
> Tests duration: 2.560 seconds.
> Total duration: 6.118 seconds.
>
>
> Browser: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR
> 2.0.50727; FDM)
> Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
> String.split...
>       Prototype:   2.296s @   4355/s (n=10000)
> Prototype-oneDiv:   1.889s @   5294/s (n=10000)
>  String.replace:   0.626s @  15974/s (n=10000)
>    String.split:   1.110s @   9009/s (n=10000)
>                    Rate Prototype Prototype-oneDiv String.split
> String.replace
>       Prototype  4355/s        --             -18%         -52%
>   -73%
> Prototype-oneDiv  5294/s       22%               --         -41%
>   -67%
>    String.split  9009/s      107%              70%           --
>   -44%
>  String.replace 15974/s      267%             202%          77%
>     --
> Tests duration: 5.921 seconds.
> Total duration: 10.234 seconds.
>
>
> Browser: Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
> Timing 10000 iterations of Prototype, Prototype-oneDiv, String.replace,
> String.split...
>       Prototype:   2.324s @   4303/s (n=10000)
> Prototype-oneDiv:   1.761s @   5679/s (n=10000)
>  String.replace:   0.492s @  20325/s (n=10000)
>    String.split:   1.350s @   7407/s (n=10000)
>                    Rate Prototype Prototype-oneDiv String.split
> String.replace
>       Prototype  4303/s        --             -24%         -42%
>   -79%
> Prototype-oneDiv  5679/s       32%               --         -23%
>   -72%
>    String.split  7407/s       72%              30%           --
>   -64%
>  String.replace 20325/s      372%             258%         174%
>     --
> Tests duration: 5.927 seconds.
> Total duration: 10.194 seconds.
>
>
> Browser: Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5
> (like Gecko)
> Timing 10000 iterations of Prototype-oneDiv, Prototype, String.replace,
> String.split...
> Prototype-oneDiv:   2.457s @   4070/s (n=10000)
>       Prototype:   2.545s @   3929/s (n=10000)
>  String.replace:   2.732s @   3660/s (n=10000)
>    String.split:   2.557s @   3911/s (n=10000)
>                   Rate String.replace String.split Prototype
> Prototype-oneDiv
>  String.replace 3660/s             --          -6%       -7%
>  -10%
>    String.split 3911/s             7%           --       -0%
>   -4%
>       Prototype 3929/s             7%           0%        --
>   -3%
> Prototype-oneDiv 4070/s            11%           4%        4%
>    --
> Tests duration: 10.291 seconds.
> Total duration: 32.202 seconds.
>
>
> Test it yourself using your own system/browser:
>  http://gfx.neohub.com/benchmark/t/escapeHTML.html
>
> cheers
> - --
> Marius Feraru
> -----BEGIN PGP SIGNATURE-----
>
> iD8DBQFFnlrqtZHp/AYZiNkRAsISAJwNZn6KHQA8Cz64R1xieZjL+unUaQCgsTRI
> o1u0l0bFW2stZOZaEFlAfdU> =Vkpv
> -----END PGP SIGNATURE-----
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Colin Mollenhour

2007-Jan-06 02:36 UTC

head link

Re: SAU/Prototype performance

So Marius debunked my "improvement" to escapeHTML but did prove that 
other methods can be much faster.

However, what are your thoughts on removing the usage of escapeHTML from 
_attributes altogether? I think Builder.node needs to be optimized for 
speed as much as possible since it is something that could be called 
thousands of times easily. I will be looking into the usage of cloneNode 
to help alleviate this, but the code probably won''t be as neat and I 
won''t always be able to use cloneNode.

Thanks,
Colin

Thomas Fuchs wrote:>
>
> Am 04.01.2007 um 13:19 schrieb Colin Mollenhour:
>
>> On 2), why can''t the escapeHTML function create a div node and
a
>> text node once one the first call and reuse it on subsequent  calls?  
>> Also, is this not a leak?
>> ------------For reference---------
>>  escapeHTML: function() {
>>    var div = document.createElement(''div'');
>>    var text = document.createTextNode(this);
>>    div.appendChild(text);
>>    return div.innerHTML;
>>  },
>> ----------------------------------
>
> That''s something we''d need to measure. The text node
creation
> wouldn''t go away, and reusing the div node would mean to empty it
out
> before adding the text node. I don''t really think that it would
make
> a real difference.
>
> Best,
> Thomas
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Fabian Lange

2007-Jan-06 06:43 UTC

head link

Re: SAU/Prototype performance

Hi,
If not removing it completely, I would go for a way (options, second version
etc) to turn it off. Sometimes I even think of a global flag for that.  Have
been playing around with symfony now for a while. I guess Ruby includes it
same way. Shouldn''t in the framework case the framework escape
everything
while generating the Javscript code? So I would go further and say that
scriptaculous can even expect getting escaped text. But since there are
manual users who might need it I would like to have a switch for it, as it
is, like Colin mentioned, a big hit there.

.: Fabian

-----Original Message-----
From: Colin Mollenhour
Sent: Samstag, 6. Januar 2007 03:36
To: rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
Subject: [Rails-spinoffs] Re: SAU/Prototype performance


So Marius debunked my "improvement" to escapeHTML but did prove that 
other methods can be much faster.

However, what are your thoughts on removing the usage of escapeHTML from 
_attributes altogether? I think Builder.node needs to be optimized for 
speed as much as possible since it is something that could be called 
thousands of times easily. I will be looking into the usage of cloneNode 
to help alleviate this, but the code probably won''t be as neat and I 
won''t always be able to use cloneNode.

Thanks,
Colin


--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-06 08:00 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Christophe Porteneuve wrote:> Great page, Marius, but I find oddly faster results in my Firefox 2
> (Iceweasel branded, I run Debian):
> 
> Browser: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.8.1) Gecko/20061024
> Iceweasel/2.0 (Debian-2.0+dfsg-1)
> Timing 1000 iterations of Prototype, Prototype-oneDiv, String.replace,
> String.split...
^ 1000 iterations. I used 10000 (ten thousands) iterations.

- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFn1cOtZHp/AYZiNkRAuWMAJ9GH49EazRLg+KP3EcEXs/824aUMQCfaubg
ZWJbemGuNwGGH+7HmO7Mb4Y=+e4V
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Sébastien Gruhier

2007-Jan-06 08:13 UTC

head link

Re: SAU/Prototype performance

I have tried on Safari and Firefox MAC, safari rocks with prototype :)


Browser: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv: 
1.8.1.1) Gecko/20061204 Firefox/2.0.0.1
Timing 10000 iterations of Prototype, Prototype-oneDiv,  
String.replace, String.split...
        Prototype:   1.027s @   9737/s (n=10000)
Prototype-oneDiv:   1.407s @   7107/s (n=10000)
   String.replace:   0.350s @  28571/s (n=10000)
     String.split:   0.648s @  15432/s (n=10000)
                     Rate Prototype-oneDiv Prototype String.split  
String.replace
Prototype-oneDiv  7107/s               --      -27%          
-54%           -75%
        Prototype  9737/s              37%        --          
-37%           -66%
     String.split 15432/s             117%       58%            
--           -46%
   String.replace 28571/s             302%      193%           
85%             --
Tests duration: 3.432 seconds.
Total duration: 6.867 seconds.


Browser: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en) AppleWebKit/ 
418.9.1 (KHTML, like Gecko) Safari/419.3
Timing 10000 iterations of Prototype-oneDiv, Prototype,  
String.replace, String.split...
Prototype-oneDiv:    .385s @  25974/s (n=10000)
        Prototype:    .365s @  27397/s (n=10000)
   String.replace:    .591s @  16920/s (n=10000)
     String.split:    .421s @  23753/s (n=10000)
                     Rate String.replace String.split Prototype- 
oneDiv Prototype
   String.replace 16920/s             --         -29%              
-35%      -38%
     String.split 23753/s            40%           --               
-9%      -13%
Prototype-oneDiv 25974/s            54%           9%                
--       -5%
        Prototype 27397/s            62%          15%                
5%        --
Tests duration: 1.762 seconds.
Total duration: 5.619 seconds.

On Jan 6, 2007, at 9:00 AM, Marius Feraru wrote:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Christophe Porteneuve wrote:
>> Great page, Marius, but I find oddly faster results in my Firefox 2
>> (Iceweasel branded, I run Debian):
>> Browser: Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.8.1) Gecko/ 
>> 20061024
>> Iceweasel/2.0 (Debian-2.0+dfsg-1)
>> Timing 1000 iterations of Prototype, Prototype-oneDiv,  
>> String.replace,
>> String.split...
>
> ^ 1000 iterations. I used 10000 (ten thousands) iterations.
>
> - --
> Marius Feraru
> -----BEGIN PGP SIGNATURE-----
>
> iD8DBQFFn1cOtZHp/AYZiNkRAuWMAJ9GH49EazRLg+KP3EcEXs/824aUMQCfaubg
> ZWJbemGuNwGGH+7HmO7Mb4Y> =+e4V
> -----END PGP SIGNATURE-----
>
> >

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-06 08:55 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Colin Mollenhour wrote:> (I was able to reproduce your results, only mine were a lot faster
> across the board, what CPU are you using?)
As already mentioned, I run all the tests with ten thousands iterations
(vs the default "one thousand" value which I suppose you used too).

My tests were done on a very "experimentalish" Linux system, host
processors are "Intel(R) Pentium(R) 4 CPU 3.00GHz", each guest host
having 512MB RAM allocated (no swapping occured during tests).

Sorry about this misleading "trick", 1K is a safe default for people
who
want to get a quick glance, but I used a larger samples number to get
more trustworthy results. OFC, I could have used even more iterations,
but 10K was enough to get a smooth variation between different
executions (at least for those "Prototypish" methods).
For instance, a 100K run on Firefox will show similar results (comparing
to previous rating):

Browser: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1)
Gecko/20061222 Firefox/2.0.0.1
Timing 100000 iterations of Prototype, Prototype-oneDiv, String.replace,
String.split...
       Prototype:  27.664s @   3615/s (n=100000)
Prototype-oneDiv:  39.626s @   2524/s (n=100000)
  String.replace:   8.138s @  12288/s (n=100000)
    String.split:  12.519s @   7988/s (n=100000)
                    Rate Prototype-oneDiv Prototype String.split
String.replace
Prototype-oneDiv  2524/s               --      -30%         -68%
   -79%
       Prototype  3615/s              43%        --         -55%
   -71%
    String.split  7988/s             217%      121%           --
   -35%
  String.replace 12288/s             387%      240%          54%
     --
Tests duration: 87.947 seconds.
Total duration: 103.594 seconds.
> Even though String.replace is slower in Opera I''d be in favor of
it
> since Opera clearly has some room for improvement with that function
> so I''m guessing this may be improved in the future.
Hm, as previously stated, browsers behave very different, and
Sbastien''s tests just completed the picture. Thanks.
I wonder, even after running all these tests, which would be the really
best overall choice? I for one cannot clearly see a winner. OFC, YMMV.

I remember I worked long time ago on a tool that consolidates different
runs of the same benchmark, I guess it could help us getting a better
overview... but I''m a mess, have no idea about the when, where or for.
Anyway, should be simple enough to write from scratch (maybe even in
JavaScript)... we''ll see. ;-)

Browser dependent implementations could be a choice, but really!?! Is
this the way to go? I really doubt it, and I''m sure many will second me
on this one, based on previous conversation, both on this list and on
our bug tracker.

Sure, we should work on optimizing things, but I''m not too sure about
this particular case ("escapeHTML"). In this case, maybe settling to
your previous solution (avoiding escapeHTML for Builder.node attributes)
would be a safer resolution for now, even if it''s plain wrong (purist
talk here: if one test fails, everything should get back to the drawing
board). I know, web developers are not generally like that, as they''re
getting disappointed every day by existent "platforms".

cheers.
- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFn2P3tZHp/AYZiNkRAqKPAKCI3V2dyByhVHgfcs67TUPTZLlRAACgtE+t
HHrFr6kNgBzUmoYQ7wlAG8Q=Xz9L
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-06 11:10 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marius Feraru wrote:> I remember I worked long time ago on a tool that consolidates different
> runs of the same benchmark, I guess it could help us getting a better
> overview... but I''m a mess, have no idea about the when, where or
for.
> Anyway, should be simple enough to write from scratch (maybe even in
> JavaScript)... we''ll see. ;-)Hm, reply to myself, not good. But anyway, curiosity killed the cat. So
I baked a quick hack to see what are we talking about. Collin, prepare
to laugh :)

A. Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1) Gecko/20061222
Firefox/2.0.0.1
B. Opera/9.10 (X11; Linux i686; U; en)
C. Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR
2.0.50727; FDM)
D. Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
E. Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5 (like Gecko)
F. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv: 1.8.1.1)
Gecko/20061204 Firefox/2.0.0.1
G. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en) AppleWebKit/ 418.9.1
(KHTML, like Gecko) Safari/419.3

Method execution rate (per second) per browser:
          Method      A      B      C      D      E      F      G
       Prototype   3535  18051   4355   4303   3929   9737  27397
Prototype-oneDiv   2635  17575   5294   5679   4070   7107  25974
  String.replace  15625   9242  15974  20325   3660  28571  16920
    String.split   7675  28169   9009   7407   3911  15432  23753

Top method per browser:
          Method      A      B      C      D      E      F      G   AVG
       Prototype      3      2      4      4      2      3      1  2.71
Prototype-oneDiv      4      3      3      3      1      4      2  2.86
  String.replace      1      4      1      1      4      1      4  2.29
    String.split      2      1      2      2      3      2      3  2.14

Collin, checkout the average column (AVG)? :)

So, String.replace gets most tops, but also gets most bottoms. :)

The funniest thing is that String.split looks like being the best
choice. UGLY! I always hated that solution (for its ugliness if nothing
else) ;-)

... please prove me I''m wrong :)

- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFn4OvtZHp/AYZiNkRAtWjAJ49RF5MXAFcccHbXNnjciurcRybtQCgjAqO
6//6zcb7E/ncp1oXGR7cXbs=l9DZ
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-06 13:48 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


Colin, I have great news :)

After more brainstorming, we (mishoo and I) came up with some updates to
my previous test, the most important thing being that we have a new
winner :)

We baked two more methods: String.replaceMap (a sloppy variant of
String.replace), and Prototype.oneDiv2 (a variant of
Prototype.oneDiv which does caches the textNode too).

More updates:
- - change text after each test (adding/changing trailing digits - to
avoid any hypothetical browsers'' cache tricks);
- - dropped escaping "&quot;" (to play similar to
"oneDiv" methods);
- - changed default iterations number to "10000" (to avoid future
confusion).

OK, without further ado, here are the consolidated results:

Benchmarks run using 10000 iterations on these browsers:
A. Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1) Gecko/20061222
Firefox/2.0.0.1
B. Opera/9.10 (X11; Linux i686; U; en)
C. Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR
2.0.50727; FDM)
D. Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5 (like Gecko)
E. Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
F. Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en) AppleWebKit/312.8.1
(KHTML, like Gecko) Safari/312.6
G. Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.1)
Gecko/20061010 Firefox/2.0

Method execution rate (per second) per browser:
           Method      A      B      C      D      E      F      G
        Prototype   3197  17094   4182   3959   4704   3598   3162
 Prototype.oneDiv   2548  16529   5005   4006   5283   5851   1826
Prototype.oneDiv2   8170  51020  13405   8258  10953  13477   6588
   String.replace  15175   6739  15699   4468  15949   1604  10526
String.replaceMap   8905   9699  12804   4008  16584   1694   6540
     String.split  10163  27322   9681   4983  10373   1730   6954

Top method per browser:
           Method      A      B      C      D      E      F      G   AVG
        Prototype      5      3      6      6      6      3      5  4.86
 Prototype.oneDiv      6      4      5      5      5      2      6  4.71
Prototype.oneDiv2      4      1      2      1      3      1      3  2.14
   String.replace      1      6      1      3      2      6      1  2.86
String.replaceMap      3      5      3      4      1      5      4  3.57
     String.split      2      2      4      2      4      4      2  2.86

Sorry, I have no Intel based Mac available at this hour, maybe Sbastien
could help us. ;-)

cheers
- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFn6i1tZHp/AYZiNkRAgcuAKDvaLyJrfZ4BvCh/Q5eIaZvSuy1vQCgxfCu
6o4cDXbL6QMgO0GVPJt0vJY=abUL
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-06 14:07 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marius Feraru wrote:> Sorry, I have no Intel based Mac available at this hour, maybe Sbastien
> could help us. ;-)Talked too soon, my folks are still alive ;-)

Benchmarks run using 10000 iterations on these browsers:
A. Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1) Gecko/20061222
Firefox/2.0.0.1
B. Opera/9.10 (X11; Linux i686; U; en)
C. Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 2.0.50727; FDM)
D. Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5 (like Gecko)
E. Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
F. Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en) AppleWebKit/312.8.1 (KHTML, like
Gecko) Safari/312.6
G. Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.1)
Gecko/20061010 Firefox/2.0
H. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.1) Gecko/20061204
Firefox/2.0.0.1
I. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en) AppleWebKit/418 (KHTML, like
Gecko) Safari/417.9.2

Method execution rate (per second) per browser:
           Method      A      B      C      D      E      F      G      H      I
        Prototype   3197  17094   4182   3959   4704   3598   3162   8803  24096
 Prototype.oneDiv   2548  16529   5005   4006   5283   5851   1826   6017  20833
Prototype.oneDiv2   8170  51020  13405   8258  10953  13477   6588  16667  39841
   String.replace  15175   6739  15699   4468  15949   1604  10526  32051  20619
String.replaceMap   8905   9699  12804   4008  16584   1694   6540  15221  24213
     String.split  10163  27322   9681   4983  10373   1730   6954  17241  27548

Top method per browser:
           Method      A      B      C      D      E      F      G      H      I
AVG
        Prototype      5      3      6      6      6      3      5      5      4
4.78
 Prototype.oneDiv      6      4      5      5      5      2      6      6      5
4.89
Prototype.oneDiv2      4      1      2      1      3      1      3      3      1
2.11
   String.replace      1      6      1      3      2      6      1      1      6
3.00
String.replaceMap      3      5      3      4      1      5      4      4      3
3.56
     String.split      2      2      4      2      4      4      2      2      2
2.67


So this new method really looks like being our winner ;-)

- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFn60gtZHp/AYZiNkRAm7PAJ9lOvg+9xB95P0UrgB+c3QlqK94ZgCfWslH
n3KwHx5c+goOZeVUY26dJfU=8PFS
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Colin Mollenhour

2007-Jan-07 03:11 UTC

head link

Re: SAU/Prototype performance

Excellent work! The oneDiv2 function really does shine but considering 
the prevalence of FF and IE6/7, I would still be hard pressed to choose 
it over String.replace. Either way, they are both a big improvement over 
the original Prototype method (with the exception of String.replace for 
Opera and Intel Safari).  Take the last step and submit a patch with 
your results to back it up :)

Thanks,
Colin

Marius Feraru wrote:>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Marius Feraru wrote:
>> Sorry, I have no Intel based Mac available at this hour, maybe S
bastien
>> could help us. ;-)
> Talked too soon, my folks are still alive ;-)
>
> Benchmarks run using 10000 iterations on these browsers:
> A. Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1) Gecko/20061222 
> Firefox/2.0.0.1
> B. Opera/9.10 (X11; Linux i686; U; en)
> C. Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 
> 2.0.50727; FDM)
> D. Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5 (like 
> Gecko)
> E. Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
> F. Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en) AppleWebKit/312.8.1 
> (KHTML, like Gecko) Safari/312.6
> G. Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.1) 
> Gecko/20061010 Firefox/2.0
> H. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.1) 
> Gecko/20061204 Firefox/2.0.0.1
> I. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en) AppleWebKit/418 
> (KHTML, like Gecko) Safari/417.9.2
>
> Method execution rate (per second) per browser:
>           Method      A      B      C      D      E      F      G      
> H      I
>        Prototype   3197  17094   4182   3959   4704   3598   3162   
> 8803  24096
> Prototype.oneDiv   2548  16529   5005   4006   5283   5851   1826   
> 6017  20833
> Prototype.oneDiv2   8170  51020  13405   8258  10953  13477   6588  
> 16667  39841
>   String.replace  15175   6739  15699   4468  15949   1604  10526  
> 32051  20619
> String.replaceMap   8905   9699  12804   4008  16584   1694   6540  
> 15221  24213
>     String.split  10163  27322   9681   4983  10373   1730   6954  
> 17241  27548
>
> Top method per browser:
>           Method      A      B      C      D      E      F      G      
> H      I   AVG
>        Prototype      5      3      6      6      6      3      5      
> 5      4  4.78
> Prototype.oneDiv      6      4      5      5      5      2      6      
> 6      5  4.89
> Prototype.oneDiv2      4      1      2      1      3      1      
> 3      3      1  2.11
>   String.replace      1      6      1      3      2      6      1      
> 1      6  3.00
> String.replaceMap      3      5      3      4      1      5      
> 4      4      3  3.56
>     String.split      2      2      4      2      4      4      2      
> 2      2  2.67
>
>
> So this new method really looks like being our winner ;-)
>
> - --
> Marius Feraru
> -----BEGIN PGP SIGNATURE-----
>
> iD8DBQFFn60gtZHp/AYZiNkRAm7PAJ9lOvg+9xB95P0UrgB+c3QlqK94ZgCfWslH
> n3KwHx5c+goOZeVUY26dJfU> =8PFS
> -----END PGP SIGNATURE-----
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Marius Feraru

2007-Jan-07 10:17 UTC

head link

Re: SAU/Prototype performance

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Colin Mollenhour wrote:> The oneDiv2 function really does shine but considering the prevalence of 
> FF and IE6/7, I would still be hard pressed to choose it over 
> String.replace.And if we notice how different FF behaves on each platform (comparing to
Opera for instance), the choice becomes even more difficult, at least if we
would have chosen to implement different solutions for each browser/platform.
> Take the last step and submit a patch with your results to back it up :)OK. Submitted as ticket #6937: http://dev.rubyonrails.org/ticket/6937

- --
Marius Feraru
-----BEGIN PGP SIGNATURE-----

iD8DBQFFoMjOtZHp/AYZiNkRAiUeAKCvKWIGvoYRu6FrEvWWrCCuHaoJzACgr6ZB
ZiiB2wHFNHiXmYilnNig5aw=98h3
-----END PGP SIGNATURE-----

--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Fred

2007-Jan-07 12:28 UTC

head link

Re: SAU/Prototype performance

Colin Mollenhour wrote:> Excellent work! The oneDiv2 function really does shine but considering
> the prevalence of FF and IE6/7, I would still be hard pressed to choose
> it over String.replace.
Also consider that while innerHTML has been widely copied, it has no
public standard whereas String.replace does.  It is known that
different browsers will return different innerHTML given identical HTML
source or piece of DOM tree.  Having it as the basis for a kind of
''normalise'' function doesn''t make sense - speed
should not be the sole
criterion.


-- 
Fred


--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Thomas Fuchs

2007-Jan-08 12:41 UTC

head link

Re: SAU/Prototype performance

Here are the results from my Quad G5:

Safari:
---------

Browser: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; de-de) AppleWebKit/ 
418.9.1 (KHTML, like Gecko) Safari/419.3
Timing 10000 iterations of Prototype, Prototype.oneDiv,  
Prototype.oneDiv2, String.replace, String.replaceMap, String.split...
         Prototype:    .610s @  16393/s (n=10000)
  Prototype.oneDiv:    .594s @  16835/s (n=10000)
Prototype.oneDiv2:    .331s @  30211/s (n=10000)
    String.replace:    .871s @  11481/s (n=10000)
String.replaceMap:    .789s @  12674/s (n=10000)
      String.split:    .975s @  10256/s (n=10000)


Webkit r18654:
---------------------

rowser: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; de-de) AppleWebKit/ 
420+ (KHTML, like Gecko) Safari/419.3
Timing 10000 iterations of Prototype, Prototype.oneDiv,  
Prototype.oneDiv2, String.replace, String.replaceMap, String.split...
         Prototype:   0.376s @  26596/s (n=10000)
  Prototype.oneDiv:   0.277s @  36101/s (n=10000)
Prototype.oneDiv2:   0.153s @  65359/s (n=10000)
    String.replace:   0.738s @  13550/s (n=10000)
String.replaceMap:   2.062s @   4850/s (n=10000)
      String.split:   0.228s @  43860/s (n=10000)

Firefox:
----------

Browser: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; de; rv: 
1.8.1.1) Gecko/20061204 Firefox/2.0.0.1
Timing 10000 iterations of Prototype, Prototype.oneDiv,  
Prototype.oneDiv2, String.replace, String.replaceMap, String.split...
         Prototype:   1.543s @   6481/s (n=10000)
  Prototype.oneDiv:   2.365s @   4228/s (n=10000)
Prototype.oneDiv2:   0.790s @  12658/s (n=10000)
    String.replace:   0.606s @  16502/s (n=10000)
String.replaceMap:   1.027s @   9737/s (n=10000)
      String.split:   0.953s @  10493/s (n=10000)


Note how the latest Safari Webkit Nightly runs circles around Firefox  
with 65359/s, wow... 8O

Best,
Thomas


Am 06.01.2007 um 15:07 schrieb Marius Feraru:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Marius Feraru wrote:
>> Sorry, I have no Intel based Mac available at this hour, maybe  
>> Sébastien
>> could help us. ;-)
> Talked too soon, my folks are still alive ;-)
>
> Benchmarks run using 10000 iterations on these browsers:
> A. Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.1) Gecko/ 
> 20061222 Firefox/2.0.0.1
> B. Opera/9.10 (X11; Linux i686; U; en)
> C. Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR  
> 2.0.50727; FDM)
> D. Mozilla/5.0 (compatible; Konqueror/3.5; Linux) KHTML/3.5.5 (like  
> Gecko)
> E. Mozilla/4.0 (compatible; MSIE 6.0; Windows 98)
> F. Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en) AppleWebKit/312.8.1  
> (KHTML, like Gecko) Safari/312.6
> G. Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.1)  
> Gecko/20061010 Firefox/2.0
> H. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.1)  
> Gecko/20061204 Firefox/2.0.0.1
> I. Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en) AppleWebKit/418  
> (KHTML, like Gecko) Safari/417.9.2
>
> Method execution rate (per second) per browser:
>           Method      A      B      C      D      E      F       
> G      H      I
>        Prototype   3197  17094   4182   3959   4704   3598   3162    
> 8803  24096
> Prototype.oneDiv   2548  16529   5005   4006   5283   5851   1826    
> 6017  20833
> Prototype.oneDiv2   8170  51020  13405   8258  10953  13477   6588   
> 16667  39841
>   String.replace  15175   6739  15699   4468  15949   1604  10526   
> 32051  20619
> String.replaceMap   8905   9699  12804   4008  16584   1694   6540   
> 15221  24213
>     String.split  10163  27322   9681   4983  10373   1730   6954   
> 17241  27548
>
> Top method per browser:
>           Method      A      B      C      D      E      F       
> G      H      I   AVG
>        Prototype      5      3      6      6      6      3       
> 5      5      4  4.78
> Prototype.oneDiv      6      4      5      5      5      2       
> 6      6      5  4.89
> Prototype.oneDiv2      4      1      2      1      3      1       
> 3      3      1  2.11
>   String.replace      1      6      1      3      2      6       
> 1      1      6  3.00
> String.replaceMap      3      5      3      4      1      5       
> 4      4      3  3.56
>     String.split      2      2      4      2      4      4       
> 2      2      2  2.67
>
>
> So this new method really looks like being our winner ;-)
>
> - --
> Marius Feraru
> -----BEGIN PGP SIGNATURE-----
>
> iD8DBQFFn60gtZHp/AYZiNkRAm7PAJ9lOvg+9xB95P0UrgB+c3QlqK94ZgCfWslH
> n3KwHx5c+goOZeVUY26dJfU> =8PFS
> -----END PGP SIGNATURE-----
>
> >
--
Thomas Fuchs
wollzelle

http://www.wollzelle.com

questentier on AIM
madrobby on irc.freenode.net

http://www.fluxiom.com :: online digital asset management
http://script.aculo.us :: Web 2.0 JavaScript
http://mir.aculo.us :: Where no web developer has gone before





--~--~---------~--~----~------------~-------~--~----~
 You received this message because you are subscribed to the Google Groups
"Ruby on Rails: Spinoffs" group.
To post to this group, send email to
rubyonrails-spinoffs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
To unsubscribe from this group, send email to
rubyonrails-spinoffs-unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org
For more options, visit this group at
http://groups.google.com/group/rubyonrails-spinoffs?hl=en
-~----------~----~----~----~------~----~------~--~---

Rails spinoffs - Jan 2007 - SAU/Prototype performance

SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance

Re: SAU/Prototype performance