Hi!

I have asked this question on the Ubuntu forum but since none has answered I'll give it a try here

I've been messing around with netcat and found a little problem that I haven't been able to workaround. Using netcat I've connected to another computer (where I have another netcat listening for incoming traffic) and got a shell. So far everything is great - I can connect and run programs without any problems. But I haven't figured out how to escape applications that are running on the remote computer. On my local computer I would escape by using ctrl + c to kill the programs but when connected to a remote computer this will kill netcat and therefor I lose my connection to the remote computer. There must be a way – but how?

/MJ

Well that works but you should do that with ssh. with ssh ctrl-c will terminate the program on the remote computer as expected plus it is much more secure then what you are doing. what you are doing is basicly what a hacker would do to set up a backdoor into your box. if you are connected to the Internet or wifi with that box you are just asking for it. ssh is encrypted and can be set up to use RSA keys instead of a password.

Open another conn and pkill whatever is running, next time run things in the background or use 'screen'? Wrt encr that's a good point: see socat?

> johnson_steve
Thank you for the warning. I agree 100% with you on this matter. Netcat is insecure in the way that is used it. However, what I'm trying to do is not to set up some kind of remote login to my system but merely learn the in's and out's of netcat. I've heard so much about it so thought it would be cool to learn it.

> unSpawn
Is it possible to run an application such as 'vi' in the background? I'm not sure what 'screen' does so I'll have to look it up before I comment on it. I also thought of using two connections but I wanted to know if there was a way to kill it directly. But curiosity keeps killing the netcat.

I will learn both ssh and socat in the future. But right now I'm only trying to master netcat

you can run apps inside screen and then press ctrl-a, d to dissconnect from screen and reattach with 'screen -r' you could also change the escape character on your local terminal to something else so it doesn't close when you press ctrl-c

>johnson_steve
I've never understood what screen was good for from just reading the manual. Now when you explain it and I tried it: fantastic! Thank you. This does a great job of bypassing everything

Most programs which you usually run in a terminal will terminate when the terminal they are connected to is hung up. When this happens, Linux sends the process HUP signal (HUP = HangUP). There are a few ways to prevent this.

Some programs simply ignore this signal, and in that way can continue without impediment. You can do this in your own scripts by trapping the signal.

A 'traditional' unix way to do it is to use the nohup command. This sends any program output to a file called nohup.out, and prevent the HUP signal from reaching a program. You have to invoke the program with nohup, e.g.
You can then press control-Z to suspend the process, enter "bg" to background it, and log out. The program will continue. You cannot re-attach to the process, but it works for many tasks, and has the advantage of being installed pretty much everywhere.

Screen is the ultimate solution for this, allowing you to re-connect to sessions after disconnecting, and even letting multiple views on the same session. The only problem with screen is that it is not as commonly installed (especially on more traditional unixes), and it's a bit bigger and more resource hungry than nohup. It's not big, but on embedded or other very light weight systems it might be an issue.