HMH wrote: > ^portmap$ > ^rpc\. I just looked at killall5.c and it shouldn't be too hard to implement a feature whereby a given list of process names or numbers is excluded. So let's talk about the semantics for the new operand list.
1. killall5 [[-]SIGNUM] [NAME]... 2. killall5 [[-]SIGNUM] [-n NAME]... 3. killall5 [[-]SIGNUM] [-n NAME...]... 4. killall5 [[-]SIGNUM] [PID]... 5. killall5 [[-]SIGNUM] [-p PID]... 6. killall5 [[-]SIGNUM] [-p PID...]... 7. killall5 [[-]SIGNUM] [-n NAME|-p PID]... 8. killall5 [[-]SIGNUM] [-n NAME...|-p PID...]... Question: Should the program accept process names or pids? killall5 is the same program is pidof, so the resources of pidof are available internally; so it wouldn't be hard to accept names. However, accepting pids would give the user more control and she can easily do "killall5 $(pidof NAME)". Accepting either names or pids would save us from having to decide, but would rule out (1) and (4) since integers are valid as program names. Question: Can we just accept the list as additional non-option arguments, or do you think we have to implement this as an option? Question: Are (3), (6) and (8) reasonable? That is, a single "-p" followed by multiple space-separated PIDs? Or do you find that too non-standard? -- Thomas Hood -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
