Comment 5 for bug 1258919

Sorry I have been busy all week.

Before posting to the usb list, do you confirm my first guess that this may be a kernel bug ? I saw that cups does not support usblp anymore. Can we be sure that this is not a cups issue ? I just mentioned the kernel because it seemed one possible thing to do. But is usblp is not used, how can we diagnose eiher the kernel or cups ? There are no /dev/usb/lp? devices anymore.

I did another test : I generate a long list in the following way :

line 1
line 2
etc ...

I store it in a Postcsript file with groff

If the list is 20 lines long, it is printinted completely and the printer releases the page. If it is 56 lines or more, one or two lines at the end are not printed and the printer does not release the page. In both situations, lpstat -t returns :

root@rubedo:~# lpstat -t
scheduler is running
system default destination: DESKJET-710C
device for DESKJET-710C: usb://HP/DESKJET%20710C
DESKJET-710C accepting requests since Sun 15 Dec 2013 09:34:18 PM CET
printer DESKJET-710C is idle. enabled since Sun 15 Dec 2013 09:34:18 PM CET
        Sending data to printer.

Please note the final "Sending data to printer." which is not removed.

Do we have enough to say that the culprit is the kernel, of cups ?

What should I do now ? Can I write an alternate backend ? I saw an example with usblp, but not with libusb ?

Best regards
R. Grasso