[cups.bugs] [MOD] STR #3753: Failure to resolve dnssd hostname should retry not fail

[twaugh.redhat]

DO NOT REPLY TO THIS MESSAGE.  INSTEAD, POST ANY RESPONSES TO THE LINK BELOW.

[STR New]

Failing to resolve a DNS-SD hostname should probably act in the same way as
failing to connect to a resolved DNS hostname.

Reasoning is as follow: when the device is not powered, the DNS-SD
hostname will no longer exist, although it is the correct and valid name. 
Jobs queued for the printer will attempt to resolve the name, fail, and (by
default) cause the queue to become stopped.  The queue will remain stopped
even when the printer is powered on.

Compare with a queue defined using a DNS hostname: when the device is not
powered, the backend will continue to attempt to connect to the device,
finally succeeding when the device is powered on.

My suggestion is to alter dnssd.c:exec_backend() so that it retries if
cupsBackendDeviceURI() fails, rather than exiting with
CUPS_BACKEND_FAILED.

Link: http://www.cups.org/str.php?L3753
Version: 1.4.5