unless tp_alloc succeeds.

Enable the man page generation from docbook y uncommenting in the
makefile.

Add the xml version. Minor changes, but still quite incomplete.
This is just a dummy manpage for now.

also logged it when such an event happens.

reference leak.  To test this case, introduced a framework for testing
process_xfrin().  This setup will be used for subseuent related fixes.

also fixed another possible (but quite unlikely) leak in case tp_alloc() fails.

this problem surfaced when 1246 got merged; the ModuleCCSession start in bind10 actually relied on some environment variable leaking out (which 1246 fixed).
This change makes it an explicit optional argument to ModuleCCSession (which can still be overridden by the environment variable)

and fix category of an earlier entry i made

like missing NULL check or exception safety.  also as a cleanup got rid
of unnecessary reinterpret_cast.

Proper solution (enabled, passing tests that use generateTestPacket2()
function) is available in #1226 that is currently under review.

The ccsession is not a process, so it should not have a "process started"
message associated with it.

Although -v was passed through to other processes, it did not affect
the main bind10 process.  -v now enables full debugging on bind10
when it starts up.

Boss now waits to receive a message from the configuration manager
to say that it is running before starting other processes.

this is a workaround for build problem for the solaris buildbot (both g++ and
sunstudio); they don't seem to work well with the const iterator in operator!=.