Re: Initial sync pickup - maybe fixed

[Jason 'vanRijn' Kasper <vR movingparts net>]

how are things going with getting the patches into fedora core2?

On Thu, 2004-06-03 at 09:15, Nigel Metheringham wrote:
> Just spent a while attempting to track this down - for those not paying
> attention the basic problem was that sometimes (often even) USB palm
> devices are not picked up by gpilotd when you hit the sync button.
> 
> However it tends to be a Heisenbug - trying to instrument whats
> happening makes it work :-(
> 
> The device is being seen by the USB subsystem and the visor module put
> in place correctly.  gpilotd is seeing the entries in
> /proc/bus/usb/devices and attempting to start a sync.... which fails to
> establish a connection.
> 
> What appears to be happening is that gpilotd opens the USB tty device
> and waits for Timeout* seconds for data to appear (using select).  It
> then gives up, closes the device, loops around sees the device is in
> /proc/bus/usb/devices and loops through the open/wait again.
> 
> [*Timeout is set in the gpilotd config applet, and defaults to 2
> seconds]
> 
> However we have a race condition here since the palm device also appears
> to send its introductory data packet about once every 2 seconds. 
> Believe it or not it appears the palm is carefully sending a data packet
> around the time that gpilotd is closing its USB device and retrying....
> 
> If you set the Timeout to something bigger - ie 4 seconds - in the
> config GUI things seem to work much better.  This is working for me at
> present.  We will see if it continues to work...
> 
> 	Nigel.
-- 

,-----------------------------------------------------------------//
| Jason 'vanRijn' Kasper ::  Numbers 6:22-26 
 `
 | All brontosauruses are thin at one end, much MUCH thicker 
 | in the middle, and then thin again at the far end.  That is 
 | the theory that I have and which is mine, and what it is too.  
 ,
| bash$ :(){ :|:&};:
`----------------------//