I have restructured the polling loop from 20 polls + 10 ms sleep to 1 poll + 1 ms sleep, the latter is more along how the bus works. The load on mu machine was about the same for both solutions, but some of my queries to the device went from 20-21 ms to 2-3 ms, so this was an improvement.
A couple of things though: 1. Perhaps we should do more than one poll per sleep, in case the 1 ms sleep is slightly longer which could cause packet buildup on a heavily utilized bus. I'm going to do some testing with fetching all pending packets, perhaps that will work. 2. if we only do one poll per sleep, then we have to use some other code where the sleep takes 10 or 20 ms as the bus will appear very slow otherwise. //Roger ------------------------------------------------------------------------- SF.Net email is sponsored by: The Future of Linux Business White Paper from Novell. From the desktop to the data center, Linux is going mainstream. Let it simplify your IT future. http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4 _______________________________________________ javax-usb-devel mailing list javax-usb-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/javax-usb-devel
