From owner-cvs-ports  Tue Jul  8 02:22:17 1997
Return-Path: <owner-cvs-ports>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id CAA07398
          for cvs-ports-outgoing; Tue, 8 Jul 1997 02:22:17 -0700 (PDT)
Received: from spinner.dialix.com.au (spinner.dialix.com.au [192.203.228.67])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id CAA07390;
          Tue, 8 Jul 1997 02:22:03 -0700 (PDT)
Received: from spinner.dialix.com.au (localhost.dialix.com.au [127.0.0.1])
          by spinner.dialix.com.au with ESMTP id RAA00247;
          Tue, 8 Jul 1997 17:21:57 +0800 (WST)
Message-Id: <199707080921.RAA00247@spinner.dialix.com.au>
X-Mailer: exmh version 2.0delta 6/3/97
To: asami@cs.berkeley.edu (Satoshi Asami)
cc: peter@FreeBSD.ORG, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG,
        cvs-ports@FreeBSD.ORG
Subject: Re: cvs commit: ports/mail/exmh2 Makefile 
In-reply-to: Your message of "Tue, 08 Jul 1997 01:23:34 MST."
             <199707080823.BAA06371@silvia.HIP.Berkeley.EDU> 
Date: Tue, 08 Jul 1997 17:21:56 +0800
From: Peter Wemm <peter@spinner.dialix.com.au>
Sender: owner-cvs-ports@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Satoshi Asami wrote:
>  *   Run scripts with /bin/sh explicitly rather than depending on the x bits
>  *   being set.  I don't know what isn't copying the x bits, but don't depend
>  *   on it.
> 
> ftp maybe?

Oh yeah, I knew that we can't depend on it, I was just curious, that's
all..  Do we do a ctm of the ports tree?  perhaps that's it?  I know cvs 
preserves the x bits on checkout, but if the ,v files were un-x-ed that 
would do it too.

> Actually we should drop all the x bits in the ports' scripts (if we
> can -- will something like "find /home/ncvs/ports -type f | xargs
> chmod -x" work?) to make sure nothing breaks.  It is not recommended
> to rely on the x bit (this has been mentioned on this list many
> times).

Yep, Satoshi, you can do this if you like..  You'll need to do it as root
though.  Hmm.. we already do a find for other ncvs permissions problems, we
could add this to the list.  Hmm... or even, tweak the post-commit stuff to
do it...  But that won't help the imports, the post commit stuff isn't run
the same way for imports.  (I've gone on record before saying I'd slightly
prefer new ports to be done with 'cvs add' and have the modules, parent
Makefile etc all done in one commit message).

> Satoshi

Cheers,
-Peter