On Thu, Oct 2, 2014 at 8:47 PM, Stephen John Smoogen <[log in to unmask]> wrote:

> You didn't follow the links in the reference section
>
> https://groups.google.com/forum/?hl=en#!msg/net.sources/BC0V7oosT8k/MKNdzEG_c3AJ
>
> It is hard to rebuild the past before the web. A lot of source code was only
> available either on tape delivery, usenet, or UUCP loops. The VAX 11/750 we
> had in 1987 at college would usually have a basic BSD installed and then
> everything re-compiled to better match the systems eccentricities. And a lot
> of that source code was purged after a year or so because 40 meg washtub
> disks are expensive and tape lasts forever.
>
> I remember that all the profs at college had tcsh instead of csh because
> before the VAX11/750 with BSD 4.2 they had all been using tops-20 and hated
> basic unix lack of features that tcsh gave them.

Oh, dear Lord, you're bringing back memories. BSD 4.2 was the first
UNIX I admin'ed. And yes, I built tcsh on it to get Emacs style
command line editing. When Bash came out with similar features, i
switched happily to it because it was more consistent with actual
Bourne shell based scripts, and maintaining both sets of scripts with
both sets of syntax was a nightmare. Subtle inconsistencies among them
were the bane of my existence, and i wanted to slap harshly the admin
who changed root's default shell from "/bin/sh" to "/bin/tcsh".

The use of such a poerful and flexible default shell as Bash has
admittedly made me nervous in the Linux world, so I'm not deeply
shocked that it's turned out to have a subtle and critical security
flaw. "Inheritance" and parsing of command lines has long been a
dangerous and tricky prospect, one that's not aided by attempts to add
more and more features.

> The link on bash is at
> https://groups.google.com/forum/?hl=en#!msg/gnu.announce/hvhlR1Vn1P0/NYwp-4_0CaUJ
>
>
>
> --
> Stephen J Smoogen.
>