GitLab

mysql_lease_mgr.

I introduced problem by fixing a typo in a comment in a zone file.
The character positions or file sizes of the file were hardcoded
in the unit test. I had seen the failure on my system but didn't
associate it. (My mistake for not looking closely.)

The typo fix added one character so changed the hardcoded 549 to 500
and 506 to 507.

(Later maybe some macros should be used for this?)

merged

mostly in documentation or in comments

reviewed by muks via jabber

includes output changes for two isc_throw.

Fix failure to connect in msgq tests.

Added braces to a single-line "if" statement.  Also corrected
some comments and removed trailing spaces from some statements.

Conflicts:
	ChangeLog

and also remove dash before beta revision in version naming

unit tests.

trivial, no review

I wrote this some month ago.
Didn't get reviewed.

some compilers claim it causes a copy whie it's defined non copyable,
triggering an error.  this fixes some build failure reported by the bots.
should mostly trivial, so committing at my discretion.

acknowledged on jabber

(used tools/reorder_message_file.py)

instead of the "static" BIND zone.  there's no difference between these two
any more, so this is just a matter of style.

 - using namespace boost is no longer used in pkt6_unittest.cc
 - copyright years updated
 - Pkt6::getRelayOption() documented
 - several methods are now better commented
 - ChangeLog now refers to libdhcp++, not b10-dhcp6

Even if we fail to connect, close the socket. It is loosely related to
the previous commit, as that one used unsuccessful connection attempts
to discover msgq is not ready yet and the unit tests complained. It
should have no real effect, since the garbage collector would reclaim
the socket after a while anyway.

This should solve the race condition when the socket file is created,
but connect to it does not work yet, because listen() was not called on
it yet. Really connecting ensures it is possible to connect.

may need some more discussion, but not the point for this branch anyway,
and at least it's more consistent with rest of the source code.

Okayed via jabber yesterday.

I am not adding a changelog entry as it is minor.