[16606] in Athena Bugs
Re: netscape-fix
daemon@ATHENA.MIT.EDU (t. belton)
Fri Jan 15 12:10:10 1999
Date: Fri, 15 Jan 1999 12:10:04 -0500 (EST)
From: "t. belton" <tbelton@MIT.EDU>
To: Jeremy Daniel <jdaniel@MIT.EDU>
Cc: web-agents@MIT.EDU, seph@MIT.EDU, bugs@MIT.EDU
In-Reply-To: <199901140452.XAA07676@hal-2000.mit.edu>
The original poster also asked (later) if we could just have the script do
the 'rm'.
I replied (to the bug list) that I'd rather not have the script do the
'rm' itself because the user might WANT to keep the cache in their home
locker for some reason - which we don't disallow, we just strongly
discourage it.
However, I was thinking of the 'netscape' wrapper script, and now that I
consider it, I obviously misunderstood his suggestion - he must have been
talking about 'netscape-fix', as you were ... If the user is running
'netscape-fix', then clearly she has already decided to nuke whatever
cache is in her home locker. Hmm.
The problem here is that you don't like telling the user to use 'rm',
because of what can go wrong, and I am equally leery of putting 'rm' in a
script anywhere, because that seems like a bigger risk. If the user goofs,
the Help Gods get a frantic call. If WE goof in the script (somehow - some
unanticipated glitch), the Help Gods get a frantic call AND it's our
fault. I guess I'd rather the user took the risk of messing up her own
space ... but that's not the greatest attitude in the world, I know.
Anyone want to weigh in? Should we correct the message and leave it to
the user to actually do the 'rm', or make it so 'netscape-fix' does the
'rm' automagically? I'd like to change it one way or another today, so
speak up.
-Todd
On Wed, 13 Jan 1999, Jeremy Daniel wrote:
> Seeing this transaction made me think about the issues closer, is
> there any reason we don't just have the netscape-fix script do the rm
> itself? (possibly after prompting the user). It's generally a good
> idea to think very hard about finding a different way rather than
> telling a user to use rm -rf
>
>
> Jer
>
>
> ------- Forwarded transaction
>
> [16600] daemon@ATHENA.MIT.EDU (Joseph Sokol-Margolis) Athena Bugs 01/13/99 15:02 (28 lines)
> Subject: a poorly phrased message
> To: bugs@MIT.EDU
> Date: Wed, 13 Jan 1999 15:02:51 EST
> From: Joseph Sokol-Margolis <seph@MIT.EDU>
>
> So I was in the sipb office today, and a user came needing help. She
> had logged in, and gotten the message about netscape eating her quota,
> then tried to type the commands as suggested. Suddenly nothing worked.
> She had meant to type "rm -rf ~/.netscape/cache" but had actually
> typed "rm -rf ~ /.netscape/cache" As many users seem confused about
> where to put spaces, and where not to, I suggest changing the message
> from:
>
> 14:15 **** Netscape is eating 5751K of your quota! ****
> 14:15 To correct, this situation, quit Netscape, and then run
> 14:15 add infoagents
> 14:15 netscape-fix
> 14:15 rm -rf ~/.netscape/cache
>
> to:
>
> 14:15 **** Netscape is eating 5751K of your quota! ****
> 14:15 To correct, this situation, quit Netscape, and then run
> 14:15 add infoagents
> 14:15 netscape-fix
> 14:15 cd
> 14:15 rm -rf .netscape/cache
>
> seph
> --[16600]--
>
> ------- End forwarded transaction
>
If you believe that mathematician Andrew Wiles' proof of
Fermat's last theorem has caused you or a member of your
family to dress too provocatively, call (800) 555-9455.
