How to (all too) easily crash ods-forcerd

Run an instance of OpenDNSSEC 2.1.4, configured with a few hundred zones.

The enforcer can be crashed with

`ods-enforcer zone list | less`

and exit'ing the `less` process before paging to the last page.

The enforcer process will die with SIGPIPE:

```
23816 8 ods-enforcerd RET write 3
23816 8 ods-enforcerd CALL write(0xb,0x759b80fff170,0x83)
23816 8 ods-enforcerd RET write -1 errno 32 Broken pipe
23816 8 ods-enforcerd PSIG SIGPIPE SIG_DFL: code=SI_NOINFO
23816 2 ods-enforcerd RET __select50 -1 errno 4 Interrupted system call
```

I consider this to be a robustness bug - an operator should not so easily be able to crash a running daemon process.

Also, no warning is left anywhere that I can see - nothing is logged before this process simply exits.

The surprise comes when trying to do the net `ods-enforcer` operation, that it has nothing to connect to.