GitLab

Use "make maintainer-clean" instead of "make distclean" to remove the manpages.
This fixes problem where enduser using tarball does a "make clean"
and loses the maintainer-tarball-provided manpages.

Fix a misspelling in a log output.
Also fix same misspelling in some comments and unittest output.

This has not been reviewed. Let's consider it a stub for now.

The tests didn't fail due to missing file. See memory_datasrc_unittest.cc:

EXPECT_THROW(zone_finder_.load(TEST_DATA_DIR "/duplicate_rrset.zone"),
...
EXPECT_THROW(rootzone.load(TEST_DATA_DIR "/duplicate_rrset.zone"),

I assume this is a false-negative because the test successds because
loading still fails.

Maybe that test or "load" should be extended for this, but for now I add
the missing file to EXTRA_DIST.

detected by distcheck failure:
http://git.bind10.isc.org/~tester/builder/BIND10-distcheck/20120831093000-FreeBSD8-i386/logs/distcheck.out

discussed on jabber

distcheck failed since wasn't listed

okayed on jabber.

some stricter compilers require this.
this fixes a build failure: http://git.bind10.isc.org/~tester/builder/BIND10/20120829105133-Solaris10-sparc-Sunstudio/logs/build.out
okayed on jabber.

added some test cases to confirm the behavior; update the doc.

with notes about the intent.

hopefully the new name better represents the intent.

not sure if it's absolute necessary for doxygen to process it, but I thought
it's more readable anyway.

I forgot to add it before.

Obviously, it fails.

Since creating new file is suspicious, the administrator may have
misspelled some file name. The message explains the problem and shows
the file name as well.

Also, removed a duplicate message, created a new instance for it.

Mostly by jinmei, small changes by vorner (assert->throw, as assert can
be turned off).

... just for clarity that the length comes first, and that this is not
some zero terminated buffer.

and tested for both versions.

okayed on jabber