On Sep 15, 4:44 am, Paul Kulchenko <[email protected]> wrote:
> > Kindly suggest me how i can implement 1000 circle in different
> > location with fast speed atleast in 5-10 second on map.
>
> I don't think this is even possible. I have a canvas based
> implementation that only has one canvas element per tile and even in
> this case it takes about 3s per 1000 circles to be drawn. You're doing
> more work in your case as you're calculating points to draw the circle
> AND you will have (at least) one HTML element for each circle (as far
> as I understand), which is going to make it even slower. I ran the
> test in IE7 on a dual-core laptop using ExCanvas.
I have never used ExCanvas but I hear it is really slow. It converts
from CANVAS to VML. CANVAS has a built-in "arc" method for drawing
circles. It is fast. I do not know whether it translates well to
VML.
> You'll need to use clustering, flash, tile/tileLayer-based drawing or
> come up with your own (better) way of doing this. The performance hit
> your get may be only for the first run; if you pre-load or pre-
> calculate what you generate it may be not that bad even in IE. I can
> display 5500 points with a decent speed and I cache all canvas
> elements I generate, which means visiting the same tile/location at
> the same zoom level second time is as fast as swapping HTML elements
> in and out of the document.
>
> Also, you may consider not having all points presented at the same
> time; you can increase *perceived* responsiveness by loading first
> 50-100 points immediately and then loading, calculating and drawing
> the rest in the background, especially if you sort them by distance
> from the center. I only draw those points that are on tiles being
> displayed.
>
> If you really want to stick to your algorithm, then pre-calculate as
> much as possible. Remove all the calculations that doesn't involve lat/
> lng variables and turn all the math you have into a table lookup for
> values from 0 to 32 (btw, you may need to have <32 instead of <33 in
> the loop as you start from 0). Consider allocating array elements
> right away and using [] instead of push(). I doubt this will get you
> where you want to be, as the slowness is probably at the VML/engine
> level, but you may get 10-20% off.
>
> Paul.
It looks like you & I may be involved in similar pursuits. I have am
developing a "Light Weight Poly" facility for V3.
www.polyarc.us/polycluster
Do you have a link to your work I might look at ?
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Google Maps API" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/google-maps-api?hl=en
-~----------~----~----~----~------~----~------~--~---
