[667] in athena10
Re: attach -e
daemon@ATHENA.MIT.EDU (Jonathan Reed)
Tue Dec 2 17:29:57 2008
Cc: Greg Price <price@mit.edu>, Anders Kaseorg <andersk@mit.edu>,
Bill Cattey <wdc@mit.edu>, athena10@mit.edu
Message-Id: <AA26FEFC-6EDC-4129-B8BF-F98CEA531B4D@mit.edu>
From: Jonathan Reed <jdreed@MIT.EDU>
To: Evan Broder <broder@mit.edu>
In-Reply-To: <4935B58F.8010602@mit.edu>
Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes
Content-Transfer-Encoding: 7bit
Mime-Version: 1.0 (Apple Message framework v919.2)
Date: Tue, 2 Dec 2008 17:28:45 -0500
On Dec 2, 2008, at 5:24 PM, Evan Broder wrote:
> Greg Price wrote:
>> On Tue, Dec 02, 2008 at 04:47:00PM -0500, Jonathan Reed wrote:
>>
>>> Changing a behavior that has existed for years is just as
>>> complicated. And this won't affect scripts or anything. If you
>>> want
>>> to use, say, "hes", in a script, then you explicitly refer to /mit/
>>> consult/bin/hes, and it works.
>>>
>>
>> No, it doesn't work in Athena 9 and it wouldn't work in Athena 10 if
>> we modified detach to start working again and someone (or the same
>> user, in broder's proposal) had run `detach consult`.
>>
>> detach should probably become a complete no-op, because of the UI bug
>> it now causes for some experienced Athena <=9 users like wdc.
>>
>> Greg
>>
> We...could make detach a shell alias for add -r...
There should ideally be some way to run "add -r" that doesn't require
the locker to be attached. As I mentioned on Zephyr, I don't think
the problem is that "detach foo" doesn't work, since for both Bill and
I, the first instinct is to say "Oh, right, I should run add -r". The
problem is that "add -r consult" then says "Not attached". Which it
shouldn't do, since all lockers are in theory "attached".
I know some people strenuously object to making detach frob your PATH,
and that's fine, as long as "add -r" works even after a detach, or we
create "unadd" or something which just frobs your PATH.
-Jon
