LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Software (https://www.linuxquestions.org/questions/linux-software-2/)
-   -   Is there any shell with a server/client model? (https://www.linuxquestions.org/questions/linux-software-2/is-there-any-shell-with-a-server-client-model-726003/)

i92guboj 05-14-2009 12:58 PM

Is there any shell with a server/client model?
 
I've been looking around for some time and I feel just curious.

Does anyone around know if there's any shell that can work this way? You launch a daemon, then use light clients to connect to it. Kind of like what you can do with rxvt-unicode (urxvtd/urxvtc), but on a shell interpreter.

This could be a memory saver when you have lots of bash sessions open.

Code:

$ ps -A | grep bash | wc -l
17

A secondary side effect would be a cleaner top/htop/ps output. So far, I've never heard of a shell that can be daemonized like this.

Thanks for reading :)

Valery Reznic 05-15-2009 12:22 AM

Quote:

Originally Posted by i92guboj (Post 3540815)
I've been looking around for some time and I feel just curious.

Does anyone around know if there's any shell that can work this way? You launch a daemon, then use light clients to connect to it. Kind of like what you can do with rxvt-unicode (urxvtd/urxvtc), but on a shell interpreter.

This could be a memory saver when you have lots of bash sessions open.

Code:

$ ps -A | grep bash | wc -l
17

A secondary side effect would be a cleaner top/htop/ps output. So far, I've never heard of a shell that can be daemonized like this.

Thanks for reading :)

Why do you think it will save memory ?
All read-only and all not written to read-write pages are shared anyway
and all written to read-write pages are likely to have different per session data

i92guboj 05-15-2009 01:15 AM

It's more about curiosity than a real need. My theory is purely empirical and based on my experience with urxvt. In the case of urxvt it certainly helps to run urxvtd and 20 urxvtc's instead of 20 plain urxvt's. I haven't audited the urxvt sources to find the 'why'. I guess that sharing a single context has some advantages.

However it would be nice even if it's just to simplify a bit the output of process watchers. :)


All times are GMT -5. The time now is 01:58 PM.