about sluggines on linux

>
>
> regards.
> art
>
>
>
> On Sat, Nov 28, 2009 at 1:52 PM, Arturo Zambrano <[hidden email]> wrote:
> Hi all,
> Sorry, I was busy and I could not respond before.
>
> I found it really slow while using compiz at 24bpp, on top a of Gentoo 64 bit box (turion 2dual core 2Ghz).
>
> I will use it now without compiz and let you know if it behaves better.
>
> Do you have any performance test that I can run so  that I have "numbers" to report? Because
> the slowness could be a subjective impression as I usually run VisualWorks.
>
> regards
> art
>
>
> On Sat, Nov 28, 2009 at 6:41 AM, Damien Cassou <[hidden email]> wrote:
> Hi Arturo,
>
> On Thu, Nov 26, 2009 at 10:07 PM, Stéphane Ducasse
> <[hidden email]> wrote:
> > Arturo told me that pharo was looking slow on his machine.
>
> I'm using Pharo on Ubuntu 9.10 and experience no problem. Could you
> please try by turning off Compiz?
>
> --
> Damien Cassou
> http://damiencassou.seasidehosting.st
>
> "Lambdas are relegated to relative obscurity until Java makes them
> popular by not having them." James Iry
>
>

>
> On Nov 29, 2009, at 3:07 PM, Arturo Zambrano wrote:
>
>> Ok, I tried without compiz and both Pharo and X11 using 16 bpp: It works better (still slow if compared to VW). For example: when pressing right button on the world, it takes almost a second  for the popup menu to show up. Is that normal?
>
> normally not
>
>> I didn't used it in another platform, I will try to give it a try under Windows to see if speed is the same.
>
> Thanks

>>
>> Is there a particular reason for Pharo not supporting 24bpp?
>
> time and money.
>
>> I think all nvidia drivers for linux works up to 24bpp, but this depth is not listed
>> in Pharo (using 16 bits looks a bit ugly)
>
>
> normally displayScreen only supports ... 8, 16 or 32 bits

>
>> Try doing
>> Transcript cr;show:
>>       [Display fillWhite. 1 to: 4 do: [:i | Pen new color: i; turn: 90*i; dragon: 14] ] timeToRun.
>>       Display restore.
>> report back how big the number is printed to the transcript
>>
>> 1897
>
> This number is fine

>>> The conclusion is that we should continue to work on making it faster.
>>>
>>> Arthuro
>>> using 1.1
>>> changing the theme
>>> using StrikeFont installDejaVu
>>> and setting up the fonts give a better feel
>>>
>>> Does anybody know a software slowing down a mac (because the problem is that
>>> I'm not sure that I can run a closure VM on a old mac I have)
>>> so that we could do some tests on slow machines?
>>>
>>> Stef
>>>

> Ok, well you don' t need to run special software. You just need to use a non-optimized VM
> The *worst* case is turn 32bit mode on, and use Rosetta
> I've put a non-optimized VM in the experimental folder/COCOA
> via
> http://smalltalkconsulting.com/squeak.html
> called
> Squeak V5 SLOW VERSION FOR TESTING DO NOT USE FOR GENERAL USE.app.zip
> That gives you something that is like 8 or 10 times slower than normal.
>
> Perhaps someone can run some benchmarks other than tinybenchmarks to confirm?
>
> 64bit mode, 32bit image
>
> '52805280 bytecodes/sec; 1763285 sends/sec'
> '55030094 bytecodes/sec; 1738242 sends/sec'
> '57040998 bytecodes/sec; 1704137 sends/sec'
> '57091882 bytecodes/sec; 1741615 sends/sec'
> '57040998 bytecodes/sec; 1736000 sends/sec'
> '56888888 bytecodes/sec; 1765598 sends/sec'
> '57193923 bytecodes/sec; 1749536 sends/sec'
> '57142857 bytecodes/sec; 1732649 sends/sec'
> '56487202 bytecodes/sec; 1722672 sends/sec'
> '55797733 bytecodes/sec; 1629865 sends/sec'
>
> 32bit mode via Get Info , 32bit image
> '62439024 bytecodes/sec; 1779601 sends/sec'
> '62439024 bytecodes/sec; 1780778 sends/sec'
> '63054187 bytecodes/sec; 1787873 sends/sec'
> '61895551 bytecodes/sec; 1796222 sends/sec'
> '62930186 bytecodes/sec; 1804649 sends/sec'
> '62439024 bytecodes/sec; 1818053 sends/sec'
> '62317429 bytecodes/sec; 1826687 sends/sec'
> '62930186 bytecodes/sec; 1831657 sends/sec'
> '63178677 bytecodes/sec; 1807071 sends/sec'
> '62015503 bytecodes/sec; 1815601 sends/sec'
>
> 32bit mode via Get Info , 32bit image,  Rosette mode aka PowerPC emulation on MacIntel
> '38905775 bytecodes/sec; 1264497 sends/sec'
> '39191671 bytecodes/sec; 1265459 sends/sec'
> '39215686 bytecodes/sec; 1235396 sends/sec'
> '39384615 bytecodes/sec; 1241850 sends/sec'
> '39000609 bytecodes/sec; 1274179 sends/sec'
> '39384615 bytecodes/sec; 1256857 sends/sec'
> '39287906 bytecodes/sec; 1247435 sends/sec'
> '39825762 bytecodes/sec; 1238154 sends/sec'
> '39312039 bytecodes/sec; 1245568 sends/sec'
> '38787878 bytecodes/sec; 1260665 sends/sec'
>
> Normally, optimized
> 64bit mode, 32bit image CHEATING
> '433530906 bytecodes/sec; 13051576 sends/sec'
> '432432432 bytecodes/sec; 13042352 sends/sec'
> '433898305 bytecodes/sec; 13051576 sends/sec'
> '433530906 bytecodes/sec; 13042352 sends/sec'
> '433898305 bytecodes/sec; 12941745 sends/sec'
> '433898305 bytecodes/sec; 13023944 sends/sec'
> '414910858 bytecodes/sec; 12815922 sends/sec'
> '433898305 bytecodes/sec; 13005587 sends/sec'
> '433164128 bytecodes/sec; 13005587 sends/sec'
> '434266327 bytecodes/sec; 12444321 sends/sec'
>
>
>
> On 2009-12-06, at 2:58 AM, Stéphane Ducasse wrote:
>
>> Ok excellent then. I should chekc because I have first generation powerbook around.
>>
>> On PC they have a tools that slow donw a computer to get the same power as an old one.
>
>>>> The conclusion is that we should continue to work on making it faster.
>>>>
>>>> Arthuro
>>>> using 1.1
>>>> changing the theme
>>>> using StrikeFont installDejaVu
>>>> and setting up the fonts give a better feel
>>>>
>>>> Does anybody know a software slowing down a mac (because the problem is that
>>>> I'm not sure that I can run a closure VM on a old mac I have)
>>>> so that we could do some tests on slow machines?
>>>>
>>>> Stef
>>>>
>
> --
> ===========================================================================
> John M. McIntosh <[hidden email]>   Twitter:  squeaker68882
> Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
> ===========================================================================
>
>
>
>

> Ok, well you don' t need to run special software. You just need to use a non-optimized VM
> The *worst* case is turn 32bit mode on, and use Rosetta
> I've put a non-optimized VM in the experimental folder/COCOA
> via
> http://smalltalkconsulting.com/squeak.html
> called
> Squeak V5 SLOW VERSION FOR TESTING DO NOT USE FOR GENERAL USE.app.zip
> That gives you something that is like 8 or 10 times slower than normal.
>
> Perhaps someone can run some benchmarks other than tinybenchmarks to confirm?
>
> 64bit mode, 32bit image
>
> '52805280 bytecodes/sec; 1763285 sends/sec'
> '55030094 bytecodes/sec; 1738242 sends/sec'
> '57040998 bytecodes/sec; 1704137 sends/sec'
> '57091882 bytecodes/sec; 1741615 sends/sec'
> '57040998 bytecodes/sec; 1736000 sends/sec'
> '56888888 bytecodes/sec; 1765598 sends/sec'
> '57193923 bytecodes/sec; 1749536 sends/sec'
> '57142857 bytecodes/sec; 1732649 sends/sec'
> '56487202 bytecodes/sec; 1722672 sends/sec'
> '55797733 bytecodes/sec; 1629865 sends/sec'
>
> 32bit mode via Get Info , 32bit image
> '62439024 bytecodes/sec; 1779601 sends/sec'
> '62439024 bytecodes/sec; 1780778 sends/sec'
> '63054187 bytecodes/sec; 1787873 sends/sec'
> '61895551 bytecodes/sec; 1796222 sends/sec'
> '62930186 bytecodes/sec; 1804649 sends/sec'
> '62439024 bytecodes/sec; 1818053 sends/sec'
> '62317429 bytecodes/sec; 1826687 sends/sec'
> '62930186 bytecodes/sec; 1831657 sends/sec'
> '63178677 bytecodes/sec; 1807071 sends/sec'
> '62015503 bytecodes/sec; 1815601 sends/sec'
>
> 32bit mode via Get Info , 32bit image,  Rosette mode aka PowerPC emulation on MacIntel
> '38905775 bytecodes/sec; 1264497 sends/sec'
> '39191671 bytecodes/sec; 1265459 sends/sec'
> '39215686 bytecodes/sec; 1235396 sends/sec'
> '39384615 bytecodes/sec; 1241850 sends/sec'
> '39000609 bytecodes/sec; 1274179 sends/sec'
> '39384615 bytecodes/sec; 1256857 sends/sec'
> '39287906 bytecodes/sec; 1247435 sends/sec'
> '39825762 bytecodes/sec; 1238154 sends/sec'
> '39312039 bytecodes/sec; 1245568 sends/sec'
> '38787878 bytecodes/sec; 1260665 sends/sec'
>
> Normally, optimized
> 64bit mode, 32bit image CHEATING
> '433530906 bytecodes/sec; 13051576 sends/sec'
> '432432432 bytecodes/sec; 13042352 sends/sec'
> '433898305 bytecodes/sec; 13051576 sends/sec'
> '433530906 bytecodes/sec; 13042352 sends/sec'
> '433898305 bytecodes/sec; 12941745 sends/sec'
> '433898305 bytecodes/sec; 13023944 sends/sec'
> '414910858 bytecodes/sec; 12815922 sends/sec'
> '433898305 bytecodes/sec; 13005587 sends/sec'
> '433164128 bytecodes/sec; 13005587 sends/sec'
> '434266327 bytecodes/sec; 12444321 sends/sec'
>
>
>
> On 2009-12-06, at 2:58 AM, Stéphane Ducasse wrote:
>
>> Ok excellent then. I should chekc because I have first generation powerbook around.
>>
>> On PC they have a tools that slow donw a computer to get the same power as an old one.
>
>>>> The conclusion is that we should continue to work on making it faster.
>>>>
>>>> Arthuro
>>>> using 1.1
>>>> changing the theme
>>>> using StrikeFont installDejaVu
>>>> and setting up the fonts give a better feel
>>>>
>>>> Does anybody know a software slowing down a mac (because the problem is that
>>>> I'm not sure that I can run a closure VM on a old mac I have)
>>>> so that we could do some tests on slow machines?
>>>>
>>>> Stef
>>>>
>
> --
> ===========================================================================
> John M. McIntosh <[hidden email]>   Twitter:  squeaker68882
> Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
> ===========================================================================
>
>
>
>

> No problem.
>
>
>> Hi Stef,
>> I dind't see your email until today, when I came back to this topic! Don't know why it was marked as read :(.
>> Just to report the same test under Vista (obviously, on the same machine).
>>
>> '225948808 bytecodes/sec; 6804203 sends/sec'
>>
>
> they are coherent with the previous ones.
>
>> '533333333 bytecodes/sec; 12157397 sends/sec'
>
> and clearly and indication that we should get also slower machines to kick us to optimize
> or rethink certain decisions.
>
> Stef