Effect of Cursor Movement

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Effect of Cursor Movement

Dan Norton
Hello,

To see this, Bring up BouncingAtomsMorph, enlarge the window, set a high atoms count so
that you can easily sense atom movement.

Notice how movement speeds up when the cursor is moved and slows when it stops? Turn
on FPS. On my machine, frame rate is about 22 with no cursor movement and increases to
around 36 when the cursor is moving. The cursor does not have to be in the window. It can
be anywhere.

Why should that be? It may not be a problem, but it is a curiosity. Any explanation?

BTW this is Win7, cogwin-15.33.3427, Cuis 2478.

 - Dan

_______________________________________________
Cuis mailing list
[hidden email]
http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
Reply | Threaded
Open this post in threaded view
|

Re: Effect of Cursor Movement

Juan Vuletich-4
Hi Dan,

(below)
On 9/11/2015 1:19 PM, Dan Norton wrote:

> Hello,
>
> To see this, Bring up BouncingAtomsMorph, enlarge the window, set a high atoms count so
> that you can easily sense atom movement.
>
> Notice how movement speeds up when the cursor is moved and slows when it stops? Turn
> on FPS. On my machine, frame rate is about 22 with no cursor movement and increases to
> around 36 when the cursor is moving. The cursor does not have to be in the window. It can
> be anywhere.
>
> Why should that be? It may not be a problem, but it is a curiosity. Any explanation?
>
> BTW this is Win7, cogwin-15.33.3427, Cuis 2478.
>
>   - Dan

Haven't even taken a look, so if I did maybe I would find something
else... but for sure Cuis is triggering MouseMoveEvents, and looking for
possible interested parties. That, on each Morphic cycle. So, each
Morphic cycle might take a bit longer, and FPS can go a bit down.

In any case, if you are curious, you can take a closer look. Often,
questions like this lead to possible optimizations!

Cheers,
Juan Vuletich

_______________________________________________
Cuis mailing list
[hidden email]
http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
Reply | Threaded
Open this post in threaded view
|

Re: Effect of Cursor Movement

Phil B
On Fri, 2015-09-11 at 13:57 -0300, Juan Vuletich wrote:

> Hi Dan,
>
> (below)
> On 9/11/2015 1:19 PM, Dan Norton wrote:
> > Hello,
> >
> > To see this, Bring up BouncingAtomsMorph, enlarge the window, set a
> > high atoms count so
> > that you can easily sense atom movement.
> >
> > Notice how movement speeds up when the cursor is moved and slows
> > when it stops? Turn
> > on FPS. On my machine, frame rate is about 22 with no cursor
> > movement and increases to
> > around 36 when the cursor is moving. The cursor does not have to be
> > in the window. It can
> > be anywhere.
> >
> > Why should that be? It may not be a problem, but it is a curiosity.
> > Any explanation?
> >
> > BTW this is Win7, cogwin-15.33.3427, Cuis 2478.
> >
> >   - Dan
>
> Haven't even taken a look, so if I did maybe I would find something
> else... but for sure Cuis is triggering MouseMoveEvents, and looking
> for
> possible interested parties. That, on each Morphic cycle. So, each
> Morphic cycle might take a bit longer, and FPS can go a bit down.
>
> In any case, if you are curious, you can take a closer look. Often,
> questions like this lead to possible optimizations!
>

The issue he's describing sounds like the opposite of that.  Any chance
this is related to the issue we discussed on vm-dev back in 2012? (The
thread was titled "Cog VM idling performance")  Your fix back then
mostly addressed the issue (I've noticed some minor idling slowdowns
since then, but nothing as extreme as when I originally reported the
issue)

Funny thing is I think I'm now dealing with the other side of this
problem in Seaside: I have a situation where an idling VM is consuming
too much CPU... (nothing specifically related to Cuis other than the
warning to 'be careful of what you wish for' :-)

> Cheers,
> Juan Vuletich
>
> _______________________________________________
> Cuis mailing list
> [hidden email]
> http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org

_______________________________________________
Cuis mailing list
[hidden email]
http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org
Reply | Threaded
Open this post in threaded view
|

Re: Effect of Cursor Movement

Juan Vuletich-4
Hi Folks,

(inline)

On 10/23/2015 6:10 PM, Phil (list) wrote:

> On Fri, 2015-09-11 at 13:57 -0300, Juan Vuletich wrote:
>> Hi Dan,
>>
>> (below)
>> On 9/11/2015 1:19 PM, Dan Norton wrote:
>>> Hello,
>>>
>>> To see this, Bring up BouncingAtomsMorph, enlarge the window, set a
>>> high atoms count so
>>> that you can easily sense atom movement.
>>>
>>> Notice how movement speeds up when the cursor is moved and slows
>>> when it stops? Turn
>>> on FPS. On my machine, frame rate is about 22 with no cursor
>>> movement and increases to
>>> around 36 when the cursor is moving. The cursor does not have to be
>>> in the window. It can
>>> be anywhere.
>>>
>>> Why should that be? It may not be a problem, but it is a curiosity.
>>> Any explanation?
>>>
>>> BTW this is Win7, cogwin-15.33.3427, Cuis 2478.
>>>
>>>    - Dan
>> Haven't even taken a look, so if I did maybe I would find something
>> else... but for sure Cuis is triggering MouseMoveEvents, and looking
>> for
>> possible interested parties. That, on each Morphic cycle. So, each
>> Morphic cycle might take a bit longer, and FPS can go a bit down.
>>
>> In any case, if you are curious, you can take a closer look. Often,
>> questions like this lead to possible optimizations!
>>
> The issue he's describing sounds like the opposite of that.  Any chance
> this is related to the issue we discussed on vm-dev back in 2012? (The
> thread was titled "Cog VM idling performance")  Your fix back then
> mostly addressed the issue (I've noticed some minor idling slowdowns
> since then, but nothing as extreme as when I originally reported the
> issue)

Thanks for the heads-up. You are right. The issue is related to that
code. I fixed it, and now BouncingAtomsMorph fps is quite better and
independent of user activity!

> Funny thing is I think I'm now dealing with the other side of this
> problem in Seaside: I have a situation where an idling VM is consuming
> too much CPU... (nothing specifically related to Cuis other than the
> warning to 'be careful of what you wish for' :-)

Check for #serverMode. This sould fix that problem. You can tweak it to
devote very little cpu to Morphic (with sluggish Morphic when you need
it). Squeak also has it.

Cheers,
Juan Vuletich

_______________________________________________
Cuis mailing list
[hidden email]
http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org