On 08/01/2012 01:02 PM, Jeffrey S. Haemer wrote:
David,
> I didn't intend for there to be any bash-isms. The only things I
> know of that I use are $(echo hi) and similar - which I didn't think
> were bash-only. If they are, then I should get rid of them.
I haven't checked lately, but /bin/sh (as in "Bourne") doesn't do
$(). Should probably stick to backticks for command substitution.
Depends on your definition of "lately." :-)
$() became a requirement for /bin/sh implementations in IEEE
1003.2-1992. That standard codified existing practice, so you'll find
$() in every /bin/sh that's under the drinking age.
(Me, I stopped using backticks for command substitution in the late 80's.)
You daredevil, you! ;-)
On even older, non-POSIX Unix-y boxes -- Unix v7 boxes, say -- we'd
still have a problem. Are such boxes a target?
No. Only interested in boxes within the last 10 years or so (maybe
less?). We already know there are appliances we can't inject nanoprobes
into. Anything we don't support properly, we can always pretend is an
appliance ;-)
Boxes that lack /bin/sh, like out-of-the-box Windows machines, are
also a problem.
Well, those boxes lack a lot more than that - and they have some of
their own unique features.
As a result, I'm not worried about bash-compatibility on Windows
machines. We need to make a special discovery case for them, and that's
that. If we ever do mainframes, they would also present unique problems.
But for everyone else, we ought to strive for as much in common as makes
sense.
--
Alan Robertson <[email protected]> - @OSSAlanR
"Openness is the foundation and preservative of friendship... Let me claim from you
at all times your undisguised opinions." - William Wilberforce
_______________________________________________
Assimilation mailing list - Discovery-Driven Monitoring
[email protected]
http://lists.community.tummy.com/cgi-bin/mailman/listinfo/assimilation
http://assimmon.org/
