[48357] in SIPB bug reports
Re: A proposal to encourage more use of /mit/sipb/bin
daemon@ATHENA.MIT.EDU (Geoffrey Thomas)
Mon Aug 19 17:11:50 2013
From geofft@MIT.EDU Mon Aug 19 21:11:49 2013
Return-Path: <geofft@MIT.EDU>
Delivered-To: bug-sipb-mtg@CHARON.mit.edu
Received: (qmail 15542 invoked from network); 19 Aug 2013 21:11:49 -0000
Received: from mailhub-auth-2.mit.edu (18.7.62.36)
by charon.mit.edu with SMTP; 19 Aug 2013 21:11:49 -0000
Received: from outgoing.mit.edu (outgoing-auth-1.mit.edu [18.9.28.11])
by mailhub-auth-2.mit.edu (8.13.8/8.9.2) with ESMTP id r7JLBgV5012655;
Mon, 19 Aug 2013 17:11:43 -0400
Received: from localhost (team-rocket.mit.edu [18.96.2.200])
(authenticated bits=0)
(User authenticated as geofft@ATHENA.MIT.EDU)
by outgoing.mit.edu (8.13.8/8.12.4) with ESMTP id r7JLBfUc004840;
Mon, 19 Aug 2013 17:11:42 -0400
Date: Mon, 19 Aug 2013 14:11:41 -0700 (PDT)
From: Geoffrey Thomas <geofft@MIT.EDU>
To: Benjamin Kaduk <kaduk@mit.edu>
cc: bug-sipb@mit.edu
Subject: Re: A proposal to encourage more use of /mit/sipb/bin
In-Reply-To: <alpine.GSO.1.10.1308191407210.24720@multics.mit.edu>
Message-ID: <alpine.DEB.2.00.1308191405030.1881@team-rocket.mit.edu>
References: <alpine.DEB.2.00.1308182122200.15460@team-rocket.mit.edu> <alpine.GSO.1.10.1308191407210.24720@multics.mit.edu>
User-Agent: Alpine 2.00 (DEB 1167 2008-08-23)
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
On Mon, 19 Aug 2013, Benjamin Kaduk wrote:
> On Sun, 18 Aug 2013, Geoffrey Thomas wrote:
>
>> What do folks think about this plan? Note that it'd require shaking up
>> everything in /mit/sipb/bin, although if folks feel like it would reduce
>
> /mit/sipb/bin is a disaster of a symlink farm, if I remember correctly. I
> managed to break the entire locker for a couple hours some years ago by
> adding a new sysname and only putting a single binary in it, without
> rebuilding the symlink farm.
> What I did then replicates the poor state of many sysnames, leading to chains
> of many symlinks to get to the actual executable. I'd be happy to see this
> get cleaned up into a more sensible form.
I can't quite tell if you are making an argument in support of, against,
or orthogonal to my proposal. :)
I'd really like to see some automated scripting for making / updating
symlinks, or better yet just doing one-command rebuilds for each new
sysname, but a) I continue to be remarkably unlikely to actually do thiis
work, and b) I _believe_ that's orthogonal.
Under this proposal there'll be one top-level symlink farm, where
/mit/sipb/arch/$sysname/thing/$foo symlinks to
/mit/sipb/project/$project/arch/$sysname/thing/$foo (where $sysname is
constant). It's up to the individual subproject for how they want to
organize /mit/sipb/project/$project/arch.
Presumably when someone (on system:administrators, presumably?) creates a
new top-level sysname, they should also run some sort of script to cause,
for each project, /mit/sipb/project/$project/arch/$sysname to symlink
to /mit/sipb/project/$project/arch/$old_sysname. The project can then
remove that symlink and replace it with a proper directory at its
convenience. This script would be much simpler to write than a
fully-general arch symlink handling script, so it can be part of this
reorganization effort.
--
Geoffrey Thomas
geofft@mit.edu
