[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: new redirection operator seems broken
From: |
Pierre Gaston |
Subject: |
Re: new redirection operator seems broken |
Date: |
Fri, 13 Mar 2009 10:07:05 +0200 |
On Thu, Mar 12, 2009 at 10:13 PM, Chet Ramey <chet.ramey@case.edu> wrote:
> I wonder if I should modify it so the implicit 2>&1 happens first, right
> after the pipe, so any user-specified redirections can override it. That
> doesn't seem that radical a change. Opinions? (I know what you think,
> Andreas ;-) ).
If i understand correctly it will only change something if you do:
command [n]>&2 |& command
As this is not a useful construct , I think that it is most likely
going to be used
only by people wanting to experiment with |& that are going to
sendbug reports
because the result is not what you would expect because it's not consistent
with "command 2>&1 >& command."
On the other hand zsh, after you disable multios, seems to behave
like bash does
now ...
I personally vote for changing it.
(Though I'm not really a big fan of |& anyway, that's so different
from ksh |&....)
As an aside the CHANGES file talks about >>& while in fact only &>> has been
implemented.
my 2 cents...
Pierre
- new redirection operator seems broken, Matt, 2009/03/12
- Re: new redirection operator seems broken, Greg Wooledge, 2009/03/12
- Re: new redirection operator seems broken, Chet Ramey, 2009/03/12
- Re: new redirection operator seems broken, Andreas Schwab, 2009/03/12
- Re: new redirection operator seems broken, Greg Wooledge, 2009/03/12
- Re: new redirection operator seems broken, Andreas Schwab, 2009/03/12
- Re: new redirection operator seems broken, Chet Ramey, 2009/03/12
- Re: new redirection operator seems broken, Greg Wooledge, 2009/03/12
- Re: new redirection operator seems broken, Chet Ramey, 2009/03/12
- Re: new redirection operator seems broken,
Pierre Gaston <=