bin/52432: /etc/dumpdates format egregious

[greywolf%starwolf.com@localhost]

>Number:         52432
>Category:       bin
>Synopsis:       /etc/dumpdates format egregious
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    bin-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Jul 26 21:30:00 +0000 2017
>Originator:     Greywolf
>Release:        NetBSD 8.99.1
>Organization:

				--*greywolf;
>Environment:
System: NetBSD eddie.starwolf.com 8.99.1 NetBSD 8.99.1 (EDDIE) #11: Fri Jul 21 13:55:41 PDT 2017 greywolf%eddie.starwolf.com@localhost:/sys/arch/amd64/compile/EDDIE amd64
Architecture: x86_64
Machine: amd64
>Description:
	/etc/dumpdates: padding device out to 511 chars seems excessive
>How-To-Repeat:
	do a dump, check /etc/dumpdates
>Fix:
	Why not just use "%s" instead of "%-511s", considering that on read-
	back, "%s" would be simpler for scanf?  The padding just seems a bit
	egregiously long (it used to be 32 chars).  Are there devices with
	white space in them such that we need to be this cautious on reading
	back the device name?

	[this bug report may also be egregious, but it seemed to me worth
	 noting.  I have had occasion to edit /etc/dumpdates to recover full
	 + incremental backup sets (read: I was stupid and forgot the
	 'u' flag).]