On 3/15/07, Bruce Dubbs <[EMAIL PROTECTED]> wrote:
> Dan Nicholson wrote:
>
> > +1. Patch looks good to me. I'm curious to hear about the dropping Ash
> > part of the equation from others, but I think Dash could be added
> > regardless.
>
> I don't have any objections to dropping ash for dash, but I think that
> should be Randy's call.
Right.
> My biggest objection to {,d}ash is that it doesn't use readline. That
> makes it infeasible (for me) to use interactively. Additionally, I like
> the extra flexibility of the bash extensions for scripts.
We already had this conversation, but...
I never intend to use dash interactively. If I'm in a shell, I want
the full glory. For a simple script, though, it's more efficient to
use a lightweight shell. I actually measured the difference a while
back, but never got around to posting it:
Time for sysinit using dash
6.633s
Time for 5 using dash
7.763s
Total rc time for dash
14.396s
Time for sysinit using bash
7.455s
Time for 5 using bash
8.154s
Total rc time for bash
15.609s
Percent difference
7.77%
So, using dash cut off nearly 8% on rc. Whether that's significant or
not is up to you. I can attach the diff to /etc/rc.d/init.d/rc if
anyone else wants to measure this on their system.
As for using bash extensions in scripts, I totally agree there. I have
a whole package management system that flexes all of bash's muscles. I
just put #!/bin/bash at the top. /bin/sh is for POSIX scripts.
--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
