transmote speaks…

yeah, let me know how that goes — i wasn’t entirely sure how best to implement that, along with the downsample ratio, to keep things as simple as possible, because there are three things to consider: the mode at which to run the camera (which i decided to default to 640×480), the downsample ratio (which remains defaulted to 0.5), and then the size at which to display the image.

what i didn’t provide access to is the placement of the clipping rectangle within the video source; it automatically sizes to one dimension and centers on the other. if you need to offset that instead of centering it, try shifting video.x/y in FLARCameraSource.init.

let me know if you run into any trouble.

sorry javier, i don’t quite understand — are you saying that ARToolkit will automatically load all the pattern files you have in a certain directory?

if i’m understanding you correctly, the problem is that flash cannot gain access to a directory listing without a server-side script. it would be possible to write something, in php for example, that collects all the filenames within a directory and passes them into the swf on init, but this is outside of the scope of FLARToolkit.

oh, yeah, sure you could do that with FLARManager. it’s just putting the assets in a separate text file. note that with squidder’s setup, you’d have to make changes to that .config file every time you added/removed a .pat to the folder.

i chose not to do that for readability — i want FLARManager to be as easy to pick up as possible, even for coders who haven’t progressed to the level where they understand how to load an external text / xml file. but, it’s also extensible, so if you want to separate out the asset list and load it into the FLARPatterns array in your own way, that’s completely your option.

sure, just set up a NetStream instance for each stream you want to use, and switch between them based on the FLARMarker.patternId you get back in your FLARMarkerEvent.MARKER_ADDED event handler.