Why Squeak is so sloooow?

I don't know if occurs the same in windows, but in
linux squeak is very slow.
I mean that all the thinks related to move pixels over
the screen causes cpu overheat in a desmesurated way
and a big latency.
It is specially bad for videogames developed in squeak
because the movement of a picture imply a speed down
of other pictures in the game.

I thought that it was caused by my old PC. But I have
tested in five diferents machines, included a powerful
P4 with 2.8Ghz and 512 Mb RAM, with the same results.

Is X system (xfree86 and xorg) a bad way for execute
squeak? Occur the same in windows?



       
       
               
______________________________________________
LLama Gratis a cualquier PC del Mundo.
Llamadas a fijos y móviles desde 1 céntimo por minuto.
http://es.voice.yahoo.com
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Hi Antonio,

Do you have any of these games to test? I have been using Squeak on
Linux without any slowness problem.

Cheers,

Offray

Antonio San. escribió:

___________________________________

AVISO LEGAL: El presente correo electronico no representa la opinion o el consentimiento oficial de la PONTIFICIA UNIVERSIDAD JAVERIANA. Este mensaje es confidencial y puede contener informacion privilegiada la cual no puede ser usada ni divulgada a personas distintas de su destinatario. Esta prohibida la retencion, grabacion, utilizacion, aprovechamiento o divulgacion con cualquier proposito. Si por error recibe este mensaje, por favor destruya su contenido y avise a su remitente.
En este aviso legal se omiten intencionalmente las tildes.

Este mensaje ha sido revisado por un sistema antivirus, por lo que su contenido esta libre de virus.
This e-mail has been scanned by an antivirus system, so its contents is free of viruses.
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

I haven't any game, indeed I'm going to look for a
good environment for game-develop learning ... and I
think that squeak could be a very very useful project
for spending efforts and time.
 
But I have done the test with the "games" inside
squeak: BouncingAtoms and Blob.

Only 1 BouncingAtoms and 3 Blobs are enough for
getting an important slowdown in all objets in the
screen (Centrino 1.8 Ghz with 512Mb RAM)

It would be interesting check if the high CPU use
continues when you enter in an other project giving in
background blobs working. But I don't know how to do
that, because when I enter in an other project blobs
stop.


PD. How can I do for getting my reply under the same
topic in the list and not in other thread?



>Hi Antonio,

>Do you have any of these games to test? I have been
>using Squeak on
>Linux without any slowness problem.

>Cheers,

>Offray


               
______________________________________________
LLama Gratis a cualquier PC del Mundo.
Llamadas a fijos y móviles desde 1 céntimo por minuto.
http://es.voice.yahoo.com
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Hi Antonio,

The German company Impara has developed a lot of games in Squeak[1]. I
don't know if any of the Impara people is here, but if this is the case
may be they can make a point about Squeak as a game development platform

[1] http://www.impara.de/references.html

I think that Squeak is nice for games development and some friends and
me will try to learn and use it for a Free Software CreativeCommons Game
(but this is a slow process, don't hold your breath). May be you can try
Python and Panda3D also. But from a cognitive point of view, I can't see
any more friendlier that Squeak.

Cheers,

Offray

Pdt: I don't know what's wrong with the threads of your mails, other
seems tho behave properly... :-/

Antonio San. escribió:

___________________________________

AVISO LEGAL: El presente correo electronico no representa la opinion o el consentimiento oficial de la PONTIFICIA UNIVERSIDAD JAVERIANA. Este mensaje es confidencial y puede contener informacion privilegiada la cual no puede ser usada ni divulgada a personas distintas de su destinatario. Esta prohibida la retencion, grabacion, utilizacion, aprovechamiento o divulgacion con cualquier proposito. Si por error recibe este mensaje, por favor destruya su contenido y avise a su remitente.
En este aviso legal se omiten intencionalmente las tildes.

Este mensaje ha sido revisado por un sistema antivirus, por lo que su contenido esta libre de virus.
This e-mail has been scanned by an antivirus system, so its contents is free of viruses.
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Antonio San. wrote: Maybe you should do some proper scientific tests on your specific needs.
What type of game are you going to create?

--
brad fuller
sonaural: www.sonaural.com
personal: www.bradfuller.com
          www.oreillynet.com/pub/au/2184



_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Brad Fuller wrote:
> Maybe you should do some proper scientific tests on your specific needs.
> What type of game are you going to create?
>
>  
You're certainly right Brad but let me add that the Squeak 3.9 image is
unbearably slow on my linux-based laptop (2Ghz Pentium 4M, 1Gb RAM).
I've even tried VM's from 3.8 to Ian's recent 3.9 release.  It is too
slow for simple window operations and causes a gastly amount of CPU
utilization even under lite GUI use.  Antonio might be reflecting on
this same perception of slowness and frankly I'm sure that any amount of
"proper scientific testing" is going to reveal that it feels like
walking though mud ;-)

Anyway I'm still using a 3.7 image for development.  Nice and snappy.

David

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

David Shaffer wrote: Ok, so I haven't used 3.9 really. What's the problem? What changed from
3.7 to 3.9. 3.8 used the 3.7 VM so that's probably not it.
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Hi Antonio,

Which VM version are you using - could you try to run, in command line:

squeak -version

And report results, thanks. Older versions of VM caused high CPU use, but I
never had a "cpu overheat in a desmesurated way" :) - new version 3.9 is
fine.

Many people on squeak-dev including me use Squeak on Linux, it certainly is
reasonably fast for me (e.g. faster than Ruby, although I think you are
talking mostly about graphical stuff?)

Also, in case you are using OpenCroquet http://opencroquet.org make sure you
have OpenGL driver installed correctly. If you use straight squeak, is it
slow for you even with a fresh image? http://squeak.org/Download



On 2006 September 21 10:18, Antonio San. wrote: _______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

  Antonio,

  If you evaluate the following (and print-it):

  | a b |
  a := FloatArray new: (16 * 1024*1024).
  b := FloatArray new: (16 * 1024*1024).
  [a += b] timeToRun.

you probably get a number around 100 or 200.  This means that Squeak
can add two 32-bit float arrays with 16M entries in 100 milliseconds
or such.  This is basicaly comparable Java-performance.

  Yes, some areas in Squeak is slow, but there are a few ways to make
particular stuff faster.

-- Yoshiki


At Thu, 21 Sep 2006 16:18:48 +0200 (CEST),
Antonio San. wrote: _______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Hi david
this is really strange because 3.9 is much snappier than 3.8 on our  
machines. This is strange.
Do you use a special settings (such as some strange smallland setup?)

Stef

On 21 sept. 06, at 17:55, David Shaffer wrote:

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

On 2006 September 21 11:55, David Shaffer wrote:
>
> You're certainly right Brad but let me add that the Squeak 3.9 image is
> unbearably slow on my linux-based laptop (2Ghz Pentium 4M, 1Gb RAM).
> I've even tried VM's from 3.8 to Ian's recent 3.9 release.  It is too
> slow for simple window operations and causes a gastly amount of CPU
> utilization even under lite GUI use.  

David,

This is suprising, on my:
        - old Athlon 2400
        - SuSE9.3
        - VM Squeak3.9alpha of 4 July 2005 [latest update: #7021],
        - latest 3.9 gama 7058 image
        - 1280x960

I was just recently thinking "this is best speed I have seen since 3.7". When
doing simple windows operations (moving around, resizing etc) top reports
below 30%, always for a very brief period when doing the work (typically
below 20%).

Yoshiki's test (higlight, alt-p)
| a b |
a := FloatArray new: (16 * 1024*1024).
b := FloatArray new: (16 * 1024*1024).
[a += b] timeToRun.
268

I do not know how to run any "oficial" graphical tests, but with
MathMorphRevival installed ( plug :) ) , time for creating 60 graphs it takes
17 seconds, usage around 80%, never above 85% (ok my top sampling 5s), goes
down to almost 0% after the UI work is done.

David or Antonio, if you would like to compare:
(higlight, alt-p)
[
1 to: 10 do: [ :i |
XYPlotter example1 plot asMorph openInWorld.
XYPlotter example2 plot asMorph openInWorld.
XYPlotter example3 plot asMorph openInWorld.
XYPlotter example4 plot asMorph openInWorld.
XYPlotter example5 plot asMorph openInWorld.
XYPlotter example6 plot asMorph openInWorld.
]
] timeToRun.
===> RESULT 17295 (hardware as above)

Milan

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Hi antonio

You should pay attention that some aspects of morphic are slow.
Lot of events are generated.... Now squeak is reasonably fast (much  
faster than python and ruby in general).
Can you tell us more about the kind of game you want to build.
You can also use VisualWorks non commercial which is quite fast.

Stef


On 21 sept. 06, at 17:22, Antonio San. wrote:

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Milan Zimmermann wrote:
shaffer@localhost ~/Squeak/Squeak3.9g-7055 $ squeak -version
3.9-7 #1 Mon Sep  4 17:01:02 EDT 2006 gcc 3.4.6
Squeak3.9alpha of 4 July 2005 [latest update: #7021]
Linux localhost 2.6.17-gentoo-r4 #1 PREEMPT Thu Aug 24 18:31:06 EDT 2006
i686 Mobile Intel(R) Pentium(R) 4 - M CPU 2.00GHz GNU/Linux
default plugin location: /usr/local/lib/squeak/3.9-7/*.so


> Yoshiki's test (higlight, alt-p)
> | a b |
> a := FloatArray new: (16 * 1024*1024).
> b := FloatArray new: (16 * 1024*1024).
> [a += b] timeToRun.
> 268
>  
Under 3.9g-7055 I get numbers in the hundreds.  Average of about 157 ms.



There is no XYPlotter in my 3.9g-7055 image.  I tried loading Plot Morph
and Plot Morph 2 from SqueakMap.  Neither worked so I'm guessing that's
not the source of this class for 3.9.  I'd be happy to run this
benchmark on the off chance it will lead to some information about why
things are sluggish.  I'm guessing, though, that it is a dead-end.
"feels slow" is often hard to track down in benchmarks.  Probably
something that does window manipulation would reveal the problem...

David

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

can you report that on squeak-dev?

On 21 sept. 06, at 21:07, David Shaffer wrote:

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Yoshiki Ohshima wrote: Ouch! I just did this (twice, just to make sure) in the 7058 imagine and
squeak bombed with the output below.
This was with the latest Linux 3.9 VM. I tried with the 3.7-7 VM and it
bombed too.
kernel: 2.6.16-1.2080.16.rrt.rhfc5.ccrma (realtime kernel from Stanford
- for audio)

what's the problem?

(Note Andreas said to Avi in a Sep12 msg regarding similar error:
"I think you're seeing the old problem of signed vs. unsigned ints when
accessing memory locations (the negative numbers seem to indicate that
much). Try a later VM (3.9-7 is the latest according to SqueakVM.org) -
IIRC, then this problem got fixed in later versions. )

Just to show I'm using the 3.9 version:

[bfuller@ives images]$ which squeak
/usr/local/bin/squeak

[bfuller@ives images]$ ls -lsa /usr/local/bin/squeak
0 lrwxrwxrwx 1 bfuller sonaural 26 Sep 14 16:56 /usr/local/bin/squeak ->
  ../lib/squeak/3.9-7/squeak

[bfuller@ives images]$ ls /usr/local/lib/squeak/3.9-7
AioPlugin             Squeak3D             vm-display-X11
B3DAcceleratorPlugin  UnixOSProcessPlugin  vm-sound-null
npsqueak.so           UUIDPlugin           vm-sound-OSS
PseudoTTYPlugin       vm-display-fbdev     XDisplayControlPlugin
squeak                vm-display-null


------------------------------------------------------
sweep failed to find exact end of memory

-2128024296 Point>+
-2128024388 Rectangle>center
-2128026592 Morph>layoutInBounds:
-2128027312 Morph>layoutProportionallyIn:
-2128046872 [] in ProportionalLayout>layout:in:
-2128046964 SequenceableCollection>do:
-2128047056 Morph>submorphsDo:
-2128047148 ProportionalLayout>layout:in:
-2128054100 Morph>doLayoutIn:
-2128055204 [] in Morph>fullBounds
-2128055020 BlockContext>on:do:
-2128055388 Morph>fullBounds
-2128055296 [] in Morph>submorphBounds
-2128055480 SequenceableCollection>do:
-2128055572 Morph>submorphBounds
-2128057332 Morph>privateFullBounds
-2128065988 Morph>doLayoutIn:
-2128067000 [] in Morph>fullBounds
-2128066816 BlockContext>on:do:
-2128067092 Morph>fullBounds
-2128067184 MorphicEventDispatcher>dispatchDefault:with:
-2128067276 MorphicEventDispatcher>dispatchEvent:with:
-2128067384 Morph>processEvent:using:
-2128067812 PasteUpMorph>processEvent:using:
-2128067960 Morph>processEvent:
-2128068052 HandMorph>sendEvent:focus:clear:
-2128068296 HandMorph>sendMouseEvent:
-2128068388 HandMorph>handleEvent:
-2128068768 MouseOverHandler>processMouseOver:
2032428416 HandMorph>handleEvent:
2032428116 HandMorph>processEvents
2032428208 [] in WorldState>doOneCycleNowFor:
2032428024 SequenceableCollection>do:
2032427932 WorldState>handsDo:
2032427840 WorldState>doOneCycleNowFor:
2032427748 WorldState>doOneCycleFor:
2032427656 WorldState>doOneSubCycleFor:
2032427564 PasteUpMorph>doOneSubCycle
2032409876 MenuMorph>invokeModalAt:in:allowKeyboard:
2032409784 MenuMorph>invokeModal:
2032409416 MenuMorph>invokeModal
2032402804 PluggableTextMorph>yellowButtonActivity:
2032402712 TextMorphForEditView>mouseDown:
2032402620 Morph>handleMouseDown:
2032402528 MouseButtonEvent>sentTo:
2032402436 Morph>handleEvent:
2032402332 MorphicEventDispatcher>dispatchMouseDown:with:
2032402200 MorphicEventDispatcher>dispatchEvent:with:
2032402108 Morph>processEvent:using:
2032401976 MorphicEventDispatcher>dispatchMouseDown:with:
2032401844 MorphicEventDispatcher>dispatchEvent:with:
2032401752 Morph>processEvent:using:
2032401660 MorphicEventDispatcher>dispatchMouseDown:with:
2032401568 MorphicEventDispatcher>dispatchEvent:with:
2032401476 Morph>processEvent:using:
2032401344 MorphicEventDispatcher>dispatchMouseDown:with:
2032401252 MorphicEventDispatcher>dispatchEvent:with:
2032401160 Morph>processEvent:using:
2032401068 MorphicEventDispatcher>dispatchMouseDown:with:
2032400976 MorphicEventDispatcher>dispatchEvent:with:
2032400868 Morph>processEvent:using:
2032400776 PasteUpMorph>processEvent:using:
2032400592 Morph>processEvent:
2032400500 HandMorph>sendEvent:focus:clear:
2032400408 HandMorph>sendMouseEvent:
2032400264 HandMorph>handleEvent:
2032400044 HandMorph>processEvents
2032400136 [] in WorldState>doOneCycleNowFor:
2032399952 SequenceableCollection>do:
2032399860 WorldState>handsDo:
2032399768 WorldState>doOneCycleNowFor:
2032399676 WorldState>doOneCycleFor:
2032399584 PasteUpMorph>doOneCycle
2025718576 [] in >spawnNewProcess
2025718760 [] in BlockContext>newProcess

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

On 2006 September 21 15:07, David Shaffer wrote:
>
> There is no XYPlotter in my 3.9g-7055 image.

> Milan Zimmermann wrote:
<<snip>>
> I do not know how to run any "oficial" graphical tests, but with
> MathMorphRevival installed ( plug :):) )

You would have to install the above package from Squeakmap to run the
"graphics test" I mentioned, but please note it's alpha, so install to a 3.9
image without important stuff.

I am not sure this is a good "graphical test", it does create graphs, so it
should have some value, it is non-interactive graphics though. Apart from
that, 3.9 feels as good as 3.7 here, but as you said .. it's my feelin' only

Milan
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Brad Fuller wrote: just a note to say that I tried it on kernel:
2.6.17-1.2174_FC5
with no difference.... same problem.
Anyone have an idea for this?

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Milan Zimmermann wrote: Gasp, sorry, didn't see that.  I've loaded it now.  Here are the
benchmark results:

[
1 to: 10 do: [ :i |
XYPlotter example1 plot asMorph openInWorld.
XYPlotter example2 plot asMorph openInWorld.
XYPlotter example3 plot asMorph openInWorld.
XYPlotter example4 plot asMorph openInWorld.
XYPlotter example5 plot asMorph openInWorld.
XYPlotter example6 plot asMorph openInWorld.
]
] timeToRun.

15000

At this point the UI is so slow that moving a window or morph results in
a noticable lag (nearly a second between releasing the mouse button
after a drag until the object that was dragged is drawn in its new
location).

>
> You would have to install the above package from Squeakmap to run the
> "graphics test" I mentioned, but please note it's alpha, so install to a 3.9
> image without important stuff.
>
> I am not sure this is a good "graphical test", it does create graphs, so it
> should have some value, it is non-interactive graphics though. Apart from
> that, 3.9 feels as good as 3.7 here, but as you said .. it's my feelin' only
>  

Agreed...it is subjective.

David

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

Slightly off topic, but this is a noob list, maybe some will find this
interesting.  If you aren't using the index in #to:do:, why not simply...

10 timesRepeat: [
    XYPlotter example1 plot asMorph openInWorld.
    XYPlotter example2 plot asMorph openInWorld.
    XYPlotter example3 plot asMorph openInWorld.
    XYPlotter example4 plot asMorph openInWorld.
    XYPlotter example5 plot asMorph openInWorld.
    XYPlotter example6 plot asMorph openInWorld.
]

_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners

On 2006 September 21 18:43, Ramon Leon wrote:
>
> Slightly off topic, but this is a noob list, maybe some will find this
> interesting.  If you aren't using the index in #to:do:, why not simply...

Ramon - yes, hehe i do qualify for a newbie, my repertoir is limited :) -
thanks for the suggestion
_______________________________________________
Beginners mailing list
[hidden email]
http://lists.squeakfoundation.org/mailman/listinfo/beginners