The following reply was made to PR lib/33569; it has been noted by GNATS.

From: matthew green <mrg@eterna.com.au>
To: Jaromir Dolecek <jdolecek@NetBSD.org>
Cc: lib-bug-people@NetBSD.org, gnats-admin@NetBSD.org,
	netbsd-bugs@NetBSD.org, kristaps@gradient-enterprises.com,
	gnats-bugs@NetBSD.org
Subject: re: lib/33569 (systrace.h is not bundled with NetBSD distribution) 
Date: Tue, 30 May 2006 03:56:24 +1000

 i think kristaps has this one right.  why do we have a systrace(4)
 if there is no way to use that information?  what exactly is the
 win of not exporting this header file?  it means that people have
 to use systrace(1) OR just grab the kernel sources.  since grabbing
 the sources is so easy, clearly there is no 'security' issue in
 exporting it.  we have at least one external user of this interface
 what is the benefit of not exporting it?  i don't understand why
 this one device should be so restricted.
 
 
 
 if it *is* decided that no sys/systrace.h is to be exported to
 userland then systrace(4) needs to have all the programming
 information, FILES reference and some of the xrefs removed and just
 the kernel options to remain.  i think 80% of that manual is about
 programming it, which depends on the definitions in systrace.h.
 
 
 .mrg.