[26113] in Athena Bugs

home help back first fref pref prev next nref lref last post

Re: linux 9.3.11: matlab

daemon@ATHENA.MIT.EDU (Alex T Prengel)
Wed Oct 13 12:14:48 2004

Message-Id: <200410131614.i9DGECbN016516@astrophel.mit.edu>
To: cfox@mit.edu
In-reply-to: "[26112] in Athena Bugs"
Date: Wed, 13 Oct 2004 12:14:12 -0400
From: Alex T Prengel <alexp@mit.edu>
cc: alexp@mit.edu
cc: bugs@mit.edu
Errors-To: bugs-bounces@mit.edu

Hi,

>I get a copy of the message dated Oct. 6 2004, each time I run matlab; I
>think I should see it only once.

Shouldn't be happening- I can't reproduce it.

>The message also seems to be incorrectly edited from the previous
>version; the message I am seeing contains:

>> This message is in file /mit/matlab/Athena/messages/ver7avail in case you 
>> want to read it again.

>But it's clearly actually /mit/matlab/Athena/messages/ver701avail

Yes- sorry about that; it's fixed (in the R/W volume).

>(If it's intended to be displayed every time, then this line isn't
>needed, in any case.)

Should only display once.

>My ~/.sw_messages has:
>  matlab61av
>  matlab531noaway
>(among many other things) and lost the lines:
>  openafsavail
>  scqilab
>when I ran matlab just now.

The latter is intentional- cookies are usually deleted after the corresponding
message is no longer being sent (with the delete_cookie subroutine in the
launch script), to prevent the ~/.sw_messages file from being cluttered
with obsolete cookies- but yours seems to be, perhaps because a given
cookie is deleted only for a limited time (because otherwise the launch script
gets cluttered with delete_cookie calls, which also slows down launch).

Is your ping time to your home dir AFS server long? That might explain
the problem- the script looks for the matching cookie in order to decide
if it should send the message or not- if it doesn't find it in time it
sends the message before the answer comes back... you can delete
all lines with "matlab" in your ~/.sw_messages file, except
matlab7.0.1_avail which is the only current, active Matlab cookie- that 
might help shorten cookie search time enough to make the problem go away.

                                         Alex

home help back first fref pref prev next nref lref last post