> The problem is that read operations on usb printers might just
> block/hang with no data from the printer (?). ulpt doesn't have
> non-blocking I/O, so I've made a patch that simply times out read
> operations, and disables further reads if it detects a blocking/stall
> condition. It is possible that this breaks the back-channel, as I'm
> unsure if we can expect a printer to send inbound data before we
> actually write anything out?
>
> It looks like there are similar problems with other backends? I've only
> looked at the usb backend yet.
>
>   
Sorry, the attachment got cleared by mailman.
http://www.radiotube.org/patch-backend_usb-unix.c

Put it into /usr/ports/print/cups-base/files if you would like to test it.



Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to