There's also a 1-button sync using ivman, that might help resolve this, because it (ivman) executes jpilot-sync after the device has been registered on the system, at which point it picks up the event via a device event notification from the udev daemon that it subscribes to, I surmise, and runs the jpilot-sync command. -TR On Thursday 17 January 2008, David A. Desrosiers wrote: > On Thu, 2008-01-17 at 07:34 -0800, MYRON SUSSMAN wrote: > > In summary: > > 1. Hook Centro up to USB port, press sync button, wait a few seconds, > > press jpilot "sync with backup" button. No response. > > > > 2. Close jpilot. Press sync button, then do > > pilot-xfer -p usb: --list > > This gives a file list. > > This is most-likely a timing issue. By the time you try to sync with > J-Pilot, your hardware -> kernel -> udev -> J-Pilot layers haven't seen > the presence of the USB device yet, so J-Pilot fails. By the time you > close J-Pilot, open a shell, run pilot-xfer, it HAS seen the device, and > you can sync with it. > > I don't have access to a Centro, nor do I know what its processor speed > is, but I bet if you play with the length of time you wait on the > J-Pilot side before hitting the Sync button in J-Pilot, you might have a > higher rate of success. > > Anywhere from 2-10 seconds is normal with these devices, and it is > entirely dependent on the speed of your PC, speed of your USB interface > and speed of the processor on the Palm itself. > > I've been trying to pinpoint where the best place is to implement a > hit-and-wait style polling mechanism inside libpisock-using-libusb, so > the whole timing issue just goes away completely with these devices.
More information about the jpilot mailing list