From mboxrd@z Thu Jan  1 00:00:00 1970
MIME-Version: 1.0
In-Reply-To: <0fdf1724083db9d68cc2e1d3c5a0bc7e@vrtra.net>
References: <041beafe28d237c7ab5edbd2ad94bace@sp.inri.net>
	<0fdf1724083db9d68cc2e1d3c5a0bc7e@vrtra.net>
Date: Wed, 16 May 2012 09:02:42 +0200
Message-ID: <CACm3i_jvY206Gs=GjxXaLN4p-_aTe+v6uASBLPb_MGeF+x1wZw@mail.gmail.com>
From: Gorka Guardiola <paurea@gmail.com>
To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Subject: Re: [9fans] rc behavior with rfork
Topicbox-Message-UUID: 8e2e315a-ead7-11e9-9d60-3106f5b1d025

> On the otherhand, putting any command in the chain makes the behavior
> disappear.
>
> cpu% @{rfork e; echo hi} |cat > /env/hi cpu% cat /env/hi hi
>
> My question is, is this intensional? =A0It feels as if there is a
> leakage here of the rfork when its effect is felt beyond the braces,
> and it feels odd for the two fs interfaces to behave differently (even
> though one of them is special)
>

When the subshell executes the rfork, how is it to know that the /env is
"outside of the braces"?

Another way of asking it is:

if I have a program with an open file descriptor in /env and calls rfork RF=
ENVG
what should happen with its /env?

That program is the shell and any command it executes that inherits
the file descriptor.

G.