Hi,
I am new to this list (albeit not to Wine :-) my Wine usage dates back to 1993 when I interviewed Bob Amstadt in his home... )
Anyway - I have a problem with Palm USB hotsync which could perhaps be solved on the userspace level. As you all know, the Linux USB kernel code creates an in-kernel device (in this case, ttyUSB1) only when a connected device actually talks. With a Palm, this is only the case if one presses the hotsync button. This keeps me from running Hotsync permanently which I need to sync my Palm against Lotus Notes. As soon as I start Hotsync, it complains that the device is not there at all. Even if I press "hotsync" and start Hotsync immediately after that, no communication takes place. (pilot-xfer -p /dev/ttyUSB1 works just fine)
Now, for Wine / Cxoffice, ttyUSB is merely a serial device. Wouldn't it be possible to modify the Wine code in a way to fake the open() call and return success to the caller, poll the actual device and open it as soon as it actually comes up? I would be willing to hack this in but I have no real idea where to look regarding the sheer size of the code. I suspect it's in server/serial.c but I am not sure. Any hint is welcome.
Or does somebody have an idea how to work around this problem? Until recently, I used a serial hotsync cable but my new notebook has no serial port any more... and carrying an USBserial converter isn't really an option on the road.
(IrDA could be an alternative but I couldn't get this to work with my T41 either, at least not with FIR).
"Harald" == Harald Milz milz@seneca.muc.de writes:
Harald> Hi, I am new to this list (albeit not to Wine :-) my Wine usage Harald> dates back to 1993 when I interviewed Bob Amstadt in his Harald> home... )
Harald> Anyway - I have a problem with Palm USB hotsync which could Harald> perhaps be solved on the userspace level. As you all know, the Harald> Linux USB kernel code creates an in-kernel device (in this case, Harald> ttyUSB1) only when a connected device actually talks. With a Harald> Palm, this is only the case if one presses the hotsync Harald> button. This keeps me from running Hotsync permanently which I Harald> need to sync my Palm against Lotus Notes. As soon as I start Harald> Hotsync, it complains that the device is not there at all. Even Harald> if I press "hotsync" and start Hotsync immediately after that, Harald> no communication takes place. (pilot-xfer -p /dev/ttyUSB1 works Harald> just fine)
Harald> Now, for Wine / Cxoffice, ttyUSB is merely a serial Harald> device. Wouldn't it be possible to modify the Wine code in a way Harald> to fake the open() call and return success to the caller, poll Harald> the actual device and open it as soon as it actually comes up? I Harald> would be willing to hack this in but I have no real idea where Harald> to look regarding the sheer size of the code. I suspect it's in Harald> server/serial.c but I am not sure. Any hint is welcome.
Perhaps try a WINEDEBUG=+relay,+server,+file run, to see whats happening. Grepping through the source will help to find out whats happening.
Keep us informed about your progress.
Good luck