On Wed May 16 11:40:13 EDT 2012, [email protected] wrote:
> hm, wait... that cant be right:
> 
> term% @{echo 1; echo 2 >[1=2]} >[2]/dev/null
> 1
> term% @{echo 1; echo 2 >[1=2]} >/dev/null
> 2
> 
> so this is what i would expect. why is that
> broken with rfork e? theres some bug lurking.

the "bug" is in devenv.  it re-evaluates the environment
group at the time of read/write.  due to the behavior
of rc, this is not going to be too much of a problem,
but on the face of it (that is, without thinking too hard),
it seems incorrect, or at least a minor violation of the
principle of least surprise.

- erik

Reply via email to