GitLab

Advertise is not perfect yet:
- server-id is missing
- client-id conveys invalid content

 - IfaceMgr now uses shared_ptr.
 - skeleton functions for specific message processing added.
 - tests fixed.

 - IA support (Option6IA class) added.
 - Support for suboptions in IA option.
 - IAADDR support (Option6IAAddr class) added.
 - Initial tests added for new classes added.

The FreeBSD distcheck failed during uninstall time as it attempted to
regenerate the file. Its dependency was from the same second.

Running make with debugging showed:

Examining bind10-guide.txt...non-existent...modified before source (../../../doc
/guide/bind10-guide.html)...out-of-date.

Solaris make says:

make: Fatal error in reader: Makefile, line 397: Badly formed macro assignment

It is fine with NetBSD's bmake and GNU gmake.

Anyways, it is not used by any target.

This was to choose the HTML-renderer.

See http://bind10.isc.org/wiki/ChangeLogDetails

use a tab before the keyword type.
use two tabs before the committer username.
line too long

catch up on ixfr config change

bump date in XML too.

As discussed at ISC all hands shortened README a lot.

Reduced by almost 200 lines.

Mention DHCP echo server.

Point to guide in txt format (now in repo).

Removed some redundant details that are in guide.

NOTE: This removed details about 1) gtest / make check 2) coverage
testing; 3) run_*.sh ; 4) running (already in guide); 5) bindctl
config examples.

I will make sure these are documented in guide, in the wiki,
or at least re-added to the doc/ directory. But some may not be
in this week's snapshot tarball.

I opened one ticket to handle the bindctl config examples (#1302).

This was discussed a little on jabber.

mostly catching up on IXFR documentation additions.

As asked at ISC All Hands.

We can later discuss if we don't want rendered documents in repo.

This was brought up at ISC All-Hands.

This will be included in the tarball.

For now this will be included in the repo too
(a different commit will add it).

I am using elinks for now. Note for now nothing in make forces it
to be rebuilt, so no tool is needed for this.

This was from jinmei.

The real fix will be done in #1292.

also, commented out now-meaningless tests rather than tweaking them
to let them pass in a superficial manner (we'll discuss this).

(the previous test succeeds just by accident)

not just on retransfer and refresh

against the master address in the config. error if they don't match, start transfer if they do

Use python_PYTHON for the python scripts. This results in
installing byte compiled version.

Removed files from EXTRA_DIST as they are now in dist by default.

Added CLEANFILES to clean up .pyc in build directory.

Note that I am keeping the installation directory the same.

and reversed its interpretation
also removed the additional None check (it was checked twice)

trivial, directly pushing.

the spec default is used when unspecified.

directly pushing.