[Chameleon] mousemove and CPU

[Paul Spencer]

Bart,

these tools just exagerate the existing problem.  The fix, I hope, is to 
provide the mouse handling through a single interface provided by the 
map image, and things like GetMousePosition will only be done once 
rather than for every widget that tracks mouse movement.

I am also toying with the idea of only having a single tool active on 
the map to further reduce the amount of javascript that has to be run 
within the event handler.

Cheers,

Paul

Bart van den Eijnden wrote:

> Hi Paul,
> 
> I have the feeling this is especially the case when the ROI tool is in 
> the application. I have 2 applications, one with the ROI Rectangle tool, 
> and one without. And for the application with the ROI tool, the mouse 
> movement is really slower/taking more CPU than in the application 
> without the ROI tool.
> 
> Also when doing some profiling with Venkman in Mozilla (great tool BTW) 
> the ROI mousemove functions are among the most heavy ones.
> 
>     Function Name: ROIRectToolGetMousePosition  (Lines 1148 - 1188)
>     Total Calls: 114 (max recurse 0)
>     Total Time: 761.09 (min/max/avg 10.01/10.02/6.68)
> 
> 
>     Function Name: ROIRectToolMouseMove  (Lines 1190 - 1211)
>     Total Calls: 114 (max recurse 0)
>     Total Time: 761.09 (min/max/avg 10.01/10.02/6.68)
> 
> Best regards,
> Bart
> 
> On Thu, 08 Jan 2004 16:31:21 -0500, Paul Spencer <pagameba@magma.ca> wrote:
> 
>> Bart,
>>
>> we are (hopefully) going to be looking into this issue with JF Doyon 
>> and NRCan.  The current mechanism is to attach to the onmousemove 
>> event handler of the document object.  We are proposing to replace 
>> this with an event handler on the map image (or layer or whatever 
>> works in most DOMs) with backwards compatibility to use the existing 
>> mechanism.  This requires some rework since all the move movement 
>> handling is done in slightly different ways and I would like to 
>> consolidate this into one centralized handling function.
>>
>> As to why it happens over the document and not the map image, I have 
>> no clue!  Any hints on this would be great.
>>
>> Cheers,
>>
>> Paul
>>
>> Bart van den Eijnden wrote:
>>
>>> Hi list,
>>>
>>> I have been looking a little bit into the CPU usage of a Chameleon 
>>> application and see that the things that occur on the mouse move 
>>> event take a lot of CPU (50-100% depending on the speed of movement). 
>>> Perhaps some of these functions could be on intervals which are more 
>>> friendly to the CPU usage?
>>>
>>> Also, I noticed that this does not only occur when moving over the 
>>> map area, but this occurs when moving over any part of the page. 
>>> That's quite strange, because no coordinates etc. need to be updated 
>>> when moving the mouse around outside the map area.
>>>
>>> Best regards,
>>> Bart _______________________________________________
>>> Chameleon mailing list
>>> Chameleon@lists.maptools.org
>>> http://lists.maptools.org/mailman/listinfo/chameleon
>>>
>>
> 
> 
> 

-- 
  -----------------------------------------------------------------
|Paul Spencer                           spencer@dmsolutions.ca    |
|-----------------------------------------------------------------|
|Applications & Software Development                              |
|DM Solutions Group Inc                 http://www.dmsolutions.ca/|
  -----------------------------------------------------------------