I assume you mean map in the geographical sense and not in the
programming sense...
- You likely aren't putting 5,000 images on a map at once, as even if the
images were 5 pixels by 5 pixels that would be mean that the screen
would need to have a resolution of 25,000 by 25,000 and that is with
no space between the images.
- HTTP not going to be all that efficient with 5K requests for images. Lots
of overhead for the calls on a per image basis...
- If the images are mostly from a limited set of objects, embed the common
ones them via css and I believe they will get stuffed in your SWF. I think you
can rename your .swf file .zip and open it up and look at the content.
- WebORB? I can't imagine (I hope you aren't) serving images as
binary content from
WebOrb, it could be done, but it probably shouldn't be done. I assume
your data lat/long
and url for the image you want to retrieve via WebORB. That's fine,
but you might also look
at some sort of simple REST style or WS calls that return xml of that
information
it might be easier for you initially if the WebORB tutorials aren't helping.
- Regardless of whether you are mashing up or even if you are using
navtec or maps
from other providers, you should be able to easily get the coordinates
for at least 0,0, and x,y and you can then convert lat long
appropriately to place
images on map or if you are mashing up, you can call a library to do it.
- On the subject of retrieving those images, you might want to do calls that
return the images within the lat/long boundaries of the map image you are
displaying and as the user pans or moves to different parts of the map,
you call and ask your database (via rest/ws or WebORB or BlaseDS ,etc)
for new images and their coordinates within the lat/long of the image.
There are a lot of easy to use geocoding extensions for databases and libraries
to help you access the data. Postgres has a GIS extension that is nice, I
am sure mysql has something, but I don't use mysql.
-Doug