I dunno, I see "hacking" as poking at something without a great understanding or a strict methodology. In the case of "hacking a system", this would include probing the ports of a system to see what's open and vulnerable. In the case of "hacking away at some code", it's cutting corners and using trial and error rather than having a clear plan in mind. "Hacking code", in my mind, is about just getting it to work and worrying about finesse later.

Example:
I was doing some performance testing, and I found that the odbc.ini file on a server kept getting corrupted when the system was under heavy load. I filed a bug on it, but I didn't want to halt testing until I could get a fix. I wrote a quick perl script to check the file every second to see if it existed and replace it if it was corrupted. With this hack in place I was able to continue testing.

A few builds later, the problem had been fixed. I assumed that someone had gone into the code, found the cause, and fixed the fault that caused the file corruption. Wrong. My hack was simply rolled into the code base. That is the danger of hacking code: temporary fixes often become permanant.