From: Dmitri K. <dm...@us...> - 2003-04-05 15:23:29
|
Hi Fernando, Well, device ID should work, this is standard IEEE1284 feature. It does prove that reverse channel is available and operational. More importantly the fact that the device has accepted a command and responded with a list of MFC channels is a good step forward. It indicates that the device is ready to talk MFC protocol. BTW what channels did you get back? Can you please execute "./mfctest /dev/parport0 SCN" and attach the output? Also what kernel and what parpot / ppdev driver version are you running? What is in your /proc/sys/dev/parport/parport0/* ? I can offer a few wild guesses about what might go wrong: * Your kernel and/or driver version is different and there is a subtle difference in handling non-blocking read requests. * Your parport is configured differently in BIOS and/or in Linux (SPP/EPP/ECP, irq, dma support etc) * Your parport cable is not IEEE1284 compliant. * Your device needs to be switched into parallel port mode explicitly, unplugged from USB, power-cycled or something stupid like that. * Your computer is just too fast (don't laugh I've seen it happening) * I misunderstood the protocol completely and my MFC works by sheer luck. BTW if anyone out there is reading this and has access to MFC with parallel port, I appreciate if you can try mfctest program fot yourself and let us know the results. It is a very simple program, just one file. It's in the project's CVS. Regards, Dmitri ----- Original Message ----- From: "Fernando Trias" <ft...@pe...> To: "Dmitri Katchalov" <dm...@us...>; <bro...@li...> Sent: Saturday, 5 April 2003 7:29 Subject: RE: [Brother-mfc-devel] parport test code in CVS Neither the mfctest.c program not the code snippet I sent earlier to this list seem to work on my MFC-6800. I get the channels fine, the send_* functions succeed and nothing is returned by the printer. I did try some code (below) to get the printer's identifier and it works, so the printer seems to be able to talk back. [snip] |