From michael.j.wittbrodt at stanfordalumni.org  Mon Apr  2 09:14:15 2001
From: michael.j.wittbrodt at stanfordalumni.org (michael.j.wittbrodt at stanfordalumni.org)
Date: 2 Apr 2001 09:14:15 -0700
Subject: Remote coral access problem
Message-ID: <20010402161415.19666.cpmta@c004.sfo.cp.net>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://snf.stanford.edu/pipermail/coral/attachments/20010402/72735a54/attachment.pl>

From bmurray at snf.stanford.edu  Thu Apr  5 08:33:59 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Thu, 5 Apr 2001 08:33:59 -0700 (PDT)
Subject: Coral Downtime Rescheduled (fwd)
Message-ID: <Pine.LNX.4.04.10104050832210.1615-100000@snf.stanford.edu>

John and Mike,

It seems to me that if you're going to be using equipment between 7 and 8
that 6:30 isn't an unreasonable time to enable your equipment?

Bill

---------- Forwarded message ----------
Date: Wed, 4 Apr 2001 17:43:24 -0700
From: Linda Whittelsey <linwhit at nortelnetworks.com>
To: 'Bill Murray' <bmurray at snf.stanford.edu>
Subject: RE: Coral Downtime Rescheduled

Is it possible for you to enable all the equipment.  Or do we have to drag
our butts outa bed at an ungodly hour to get there early to enable our own
equipment? : )

Linda W.

-----Original Message-----
From: Bill Murray [mailto:bmurray at snf.stanford.edu]
Sent: Wednesday, April 04, 2001 5:33 PM
To: labmembers at snf.stanford.edu
Subject: Coral Downtime Rescheduled


Lab members,

We have rescheduled our Coral database maintenance for Friday instead of
Thursday.  Coral will be unavailable on Friday, April 6, from 6:30am to
8:00am.  Please enable any equipment that you will be using during this 
time before 6:30am.

Thanks,
Coral Development Team




From shott at snf.stanford.edu  Thu Apr  5 08:44:01 2001
From: shott at snf.stanford.edu (John Shott)
Date: Thu, 05 Apr 2001 08:44:01 -0700
Subject: Coral Downtime Rescheduled (fwd)
References: <Pine.LNX.4.04.10104050832210.1615-100000@snf.stanford.edu>
Message-ID: <3ACC92C1.D5185A4D@snf.stanford.edu>

Bill:

I think there is no good reason to do an "All On" for a previously announced
one hour outage ... there is a simple choice: come in early or wait until 8
a.m.  I don't think that is unreasonable.

Do you want me to send e-mail to Linda?

Thanks,

John

p.s. Since Coral runs 24x7 with no pre-planned time for maintenance activity,
I think this is a perfectly reasonable thing for us to ask for.  Also, I
personally would have not agreed to delay it to simply miss a users meeting
... if our lab members can think of nothing to gripe about better than we
scheduled one hour of downtime for the first time in several months, then we
must be doing a damn good job.


From mbell at snf.stanford.edu  Thu Apr  5 12:13:51 2001
From: mbell at snf.stanford.edu (Mike Bell)
Date: Thu, 05 Apr 2001 12:13:51 -0700
Subject: clever PC Leland error message
Message-ID: <3ACCC3EF.547E2468@snf.stanford.edu>

John and Bill,

Turns out the problem with Tony's machine (PC Leland quit working) was
probably 2 fold. The version he had was not the most current - although
it was a version that did not "require" an update per the university. I
was having the same problem suddenly on my machine and fixed it by going
from 2.0 to 2.1.1. When I upgraded Tony's machine to 2.1.1 the install
finished successfully, but the application bombed with a "missing
xxxxxxxx.dll" message. Might have been nice to get a "IE version is not
current - load IE 5.0 or greater". Turns out the fix was to load a more
current version of IE. I can only feel sorry for the help desk folks who
apparently were buried with calls on this. In addition, I hate to think
how much time I lost on this.

If you hear of this problem elsewhere I can fix it quickly this time.

Mike
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://snf.stanford.edu/pipermail/coral/attachments/20010405/e4fb68c5/attachment.html>

From bmurray at snf.stanford.edu  Thu Apr  5 23:09:00 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Thu, 5 Apr 2001 23:09:00 -0700 (PDT)
Subject: Coral 1.3 Ready
Message-ID: <Pine.LNX.4.04.10104052300450.5209-100000@snf.stanford.edu>

John and Mike,

I have finished my testing, checked all the code in, and tagged the latest
version as Coral 1.3, the database and code cleanup.  I think we're in
good shape.  

I was not able to fix all the bugs in the resource (add) client.  However,
I didn't introduce any new ones.  Since the add client is only used by
a few people and is scheduled to be replaced on April 15,  I couldn't hold
things up to fix existing bugs.  The only remaining problem is that 
accounts need to be added before adding projects that charge to the 
accounts.  After we get everything installed tomorrow, maybe John and I
can take a minute to make sure that the administrative staff will be ok
with this. 

See you all at 7:00am tomorrow.

Thanks,
Bill



From s.vargo at siwaveinc.com  Sat Apr  7 07:49:51 2001
From: s.vargo at siwaveinc.com (Stephen Vargo)
Date: Sat, 07 Apr 2001 07:49:51 -0700
Subject: Coral enable problem
Message-ID: <986654991.3acf290f9b904@www.electricwebmail.com>

Coral team,

	I have tried 6 times this morning (from 6 am until 7:30 am on 4/7/01) to enable several items using Coral. After waiting 3-4 minutes a window appears that says "Hardware manager error:Error connecting to daemon. See lab staff."

When will coral be up and running?

Thanks.
Steve Vargo
svargo at snf.stanford.edu


From shott at snf.stanford.edu  Sat Apr  7 09:48:27 2001
From: shott at snf.stanford.edu (John Shott)
Date: Sat, 07 Apr 2001 09:48:27 -0700
Subject: Coral enable problem
References: <986654991.3acf290f9b904@www.electricwebmail.com>
Message-ID: <3ACF44DB.D12C0BA7@snf.stanford.edu>

Steve:

We have networking problems ... unfortunately, I'm not Mr. Network but I'm
trying to see what I can do to get us back on the air.

I'll keep you posted ...

John


From shott at snf.stanford.edu  Sat Apr  7 10:54:51 2001
From: shott at snf.stanford.edu (John Shott)
Date: Sat, 07 Apr 2001 10:54:51 -0700
Subject: Ongoing network problems ...
Message-ID: <3ACF546B.E5F5F2A@snf.stanford.edu>

Bill and Mike:

We continue to have network problems that appear to result in Walker being
completely cut off from the rest of the world.  At this point I have no means
of determining whether it is a fiber problem, a switch problem, or a network
card problem.  Power cycling the switch seems to help briefly ... but not long
enough to be useful.  Switching the fiber from one interface to the other
doesn't seem to help.  Jiggling the fiber into the switch doesn't seem to
help.  I tried to reboot walker ... only to find that it can't find "linux"
(AKA snf, I think) to mount stuff, can't make a ntp connection, etc.

So, here is what I've done:

1. I issued a hardware "All On".  Because nfs wasn't mounted and walker itself
is pretty dead (even though I think the daemon is actually running), I
couldn't run my script that allows be to test the daemon with scripts called
all_on_0, all_on_1, all_on_2, and all_on_3 ... that is all on for each board
one at a time.

2. So, I disconnected the 4 walker ribbon cables from the patch panel and
briefly connected the 4 ribbon cables from the old interlock control box.

3. From the Sunray in the basement, I sshed to crystal and issued (as root, I
think) "eqcntl allon" from the command line.

4. Then I disconnected the old interlock box ribbon cables and re-connected
the walker ribbon cables so that we are ready to go when we resolve the
networking problem.

5. Finally, in the equipment table, I used sqlplus to set interlocked = 0
where interlocked = 1 so that we effectively keep software enabling and
disabling (and accounting records).  Note: I also have a log file that is the
result of "select name from equipment where interlocked = 1" that I ran before
I did this so that I have a list of the 62 pieces of equipment that are
normally interlocked.  (Note: this file is in
~shott/labnet/hardware/dirver/interlock.log .... I've also got a file called
interlock_return.sql in that directory that will return the interlocked 0 to a
1 for those 62 pieces of equipment when we get this problem figured out.

Note: I also told people that they had to continue to enable/disable and that
I would double monthly charges for anyone caught cheating the system.

I may send out e-mail to jason and charlie orgish to ask if they know of any
simple things we can to do resolve whether our problem is with Walker's NIC,
wiring, fiber, or the switch.

Thanks,

John


From shott at snf.stanford.edu  Sun Apr  8 10:53:07 2001
From: shott at snf.stanford.edu (John Shott)
Date: Sun, 08 Apr 2001 10:53:07 -0700
Subject: Minor addition to equipment server ...
Message-ID: <3AD0A583.910E5A7A@snf.stanford.edu>

Bill and Mike:

Based on our recent experience with wanting to continue software
enabling/disabling while turning off hardware enabling/disabling, I think that
it might be useful to have an addtion to the equipment server configuration
file that includes a line like:

USE_HARDWARE_INTERLOCK=true

Then, assuming during initializition this would be turned into a Boolean
use_hardware_interlock value, the line that currently reads:

if (new Boolean(mA.getValue("interlocked")).booleanValue()) {
	// Then either do a hardware disable or enable ...
}

could be replaced with:

if (use_hardware_interlock && new
Boolean(mA.getValue("interlocked")).booleanValue()) {
	// Then either do a hardware disable or enable ...
}

This would seem to have 2 advantages:

1. We could temporarily go from normally using hardware interlock to not in
the event of a network/walker problem without changing the database values.

2. For sites that didn't have hardware interlocking at all, they wouldn't even
have to check the database for the interlock status of each piece of equipment
... and would be immune to garbage in those fields.

I haven't tried to do this myself for fear that I might screw up ... plus I
assume that we don't want to mess with much until we are 100% sure that all of
the database/sofware changes are consistent. Although, thus far ... we sure
seem to be in good shape.

Thanks,

John


From bmurray at snf.stanford.edu  Sun Apr  8 15:16:41 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Sun, 8 Apr 2001 15:16:41 -0700 (PDT)
Subject: Minor addition to equipment server ...
In-Reply-To: <3AD0A583.910E5A7A@snf.stanford.edu>
Message-ID: <Pine.LNX.4.04.10104081510400.17064-100000@snf.stanford.edu>

John,

This sounds like a great idea.  I'm sure you wouldn't screw anything up,
but I do agree we shouldn't make any changes to Coral until we're certain
that the database/code upgrade is just right.  I'd be happy to make and
test this change when I make the resource manager changes that Mike will
require for the new resource client.

Have a great week and thanks to Mike for the quick fix of the network!

Bill 


On Sun, 8 Apr 2001, John Shott wrote:

> Bill and Mike:
> 
> Based on our recent experience with wanting to continue software
> enabling/disabling while turning off hardware enabling/disabling, I think that
> it might be useful to have an addtion to the equipment server configuration
> file that includes a line like:
> 
> USE_HARDWARE_INTERLOCK=true
> 
> Then, assuming during initializition this would be turned into a Boolean
> use_hardware_interlock value, the line that currently reads:
> 
> if (new Boolean(mA.getValue("interlocked")).booleanValue()) {
> 	// Then either do a hardware disable or enable ...
> }
> 
> could be replaced with:
> 
> if (use_hardware_interlock && new
> Boolean(mA.getValue("interlocked")).booleanValue()) {
> 	// Then either do a hardware disable or enable ...
> }
> 
> This would seem to have 2 advantages:
> 
> 1. We could temporarily go from normally using hardware interlock to not in
> the event of a network/walker problem without changing the database values.
> 
> 2. For sites that didn't have hardware interlocking at all, they wouldn't even
> have to check the database for the interlock status of each piece of equipment
> ... and would be immune to garbage in those fields.
> 
> I haven't tried to do this myself for fear that I might screw up ... plus I
> assume that we don't want to mess with much until we are 100% sure that all of
> the database/sofware changes are consistent. Although, thus far ... we sure
> seem to be in good shape.
> 
> Thanks,
> 
> John
> 



From mtung at iridigm.com  Mon Apr  9 17:14:22 2001
From: mtung at iridigm.com (Ming-Hau Tung)
Date: Mon, 9 Apr 2001 17:14:22 -0700 
Subject: Remote Access Not Working
Message-ID: <A705C922ABE1D411982C00B0D079E00E051B86@MAIL>

To Coral Help Staff,
I am an Industrial User, and trying to get Remote access to Coral.
I have followed the procedure listed on the web.
I have downloaded the Java file, and created the remote password at the CIS
facility.
When I launch the Java Web start, the User Login and Password prompt
appears.
I then enter the ID info, and wait for ~5min, and then a failure notice
appears
"Cannot Access Equipment Information"
I believe this to be my network's error, please inform me so that I can tell
the network admin.
people here at Iridigm how to remedy this wall
Best Regards,
ming


From shott at snf.stanford.edu  Mon Apr  9 17:27:41 2001
From: shott at snf.stanford.edu (John Shott)
Date: Mon, 09 Apr 2001 17:27:41 -0700
Subject: Remote Access Not Working
References: <A705C922ABE1D411982C00B0D079E00E051B86@MAIL>
Message-ID: <3AD2537D.A3128333@snf.stanford.edu>

Ming:

Yes, in most cases where people get the "Cannot Access Equipment Information"
error, we believe that network connectivity has been the issue.  Do you know
whether your network connection is behind a firewall?  If so, it is our
belief  that the CORBA connections that represent the communication between
your client and our server is stopped by the firewall.  Do you have any
alternative network connectivitiy that would not be behind a firewall?  If so,
that may be a good way of testing whether a firewall is contributing to these
problems.  Let me know if I can be of further assistance.

Good luck,

John


From shott at snf.stanford.edu  Tue Apr 10 18:31:38 2001
From: shott at snf.stanford.edu (John Shott)
Date: Tue, 10 Apr 2001 18:31:38 -0700
Subject: Adding and de-activating lab members.
Message-ID: <3AD3B3FA.7813F501@snf.stanford.edu>

Mike and Bill:

I've been receiving requests in response to some of my e-mails sent to
labmembers at snf of the form "I don't use the lab any more, please remove me
from the mailing lists ...".  I've been doing that manually in the following
way ... but this is something that should probably be a feature of the new
resource client.

Here is what I currently do:

1. Remove them from all mailing lists ... I currently do this by issuing the
command (as user "alias") "/app/ezmlm/bin/ezmlm-remove member".  An argument
could be made, I think, that a better way to do this would be to remove them
from the qualify list for each piece of equipment. If this, in turn, then did
an "unsub" (that is the manual unsubscribe rather than one that requires
confirmation) for each piece of equipment, then they would automatically be
removed from the labmembers list when it is regenerated each night ...).

2. I next edit member and change their status from "a" (active) to "i"
(inactive). (Actually, come to think of it, I should be setting edate for the
member as well ... I haven't been doing that).

3. Finally, I go in and change works_on by the sql command:
update works_on
set revoker = 'shott',
edate = SYSDATE,
active = 0
where master = 'member_name_goes_here' and active = 1;
which kills all of their currently active projects and would keep them from
getting a valid ticket.

At some point, we may wish to move inactive users to the old_members table ...
which, I suppose could either be done at this point or at some later date.

One think that I don't currently do is wipe out or disable their login
accounts on sunray and snf.  My guess is that doing a "userdel" would be a bit
drastic ... while we should probably do that at some point, my guess is that
we should wait.  However, I think that preventing them from logging in would
be a good thing.  The way that we used to do that on crystal, as I recall, was
to replace thier default shell in /etc/passwd with a script that would be
called something like /home/labnet/sbin/nologin.  Following is the copy of the
script that we used to use on crystal (it was called
/micro/micro/labroot/etc/nologin)

#!/bin/sh
echo
echo "===================================================="
echo "Your laboratory account has been inactivated."
echo "Your Crystal account still exists but your lab"
echo "materials and priviledges will have to be reissued"
echo "if you want to use the laboratory again."
echo "===================================================="
echo

Finally, just as a reminder, currently, the generation of accounts on sunray
and snf, mail forwarding, etc. are done manually ... that this should be a
part of the new client.  The scripts that we currently use are:
/home/labnet/sbin/newuser member "Real Name" optional_mail_forwarding at aol.com

This, in turn, calls "setpasswd member" which makes use of the linux feature
of allowing a random password to be set ... this random passwd is then
returned to the person running it and they are asked to type that name in
twice (ostensibly to prove that they read it correctly) but, under the covers
it is also doing "passwd member" (as root) on sunray to set their password
there.  This is clearly a kludge, not very secure, etc.  In the ideal world,
we would let these folks pre-select a password that they want when they fill
out their forms ... securely of course ... and then have a secure way of
setting it as they wish.

In any event, those are my thoughts for the day.

Thanks,

John


From shott at snf.stanford.edu  Tue Apr 10 19:00:47 2001
From: shott at snf.stanford.edu (John Shott)
Date: Tue, 10 Apr 2001 19:00:47 -0700
Subject: Java Regular Expressions ...
Message-ID: <3AD3BACF.23650FD1@snf.stanford.edu>

Bill and Mike:

I know that we raised the question of whether there was a regular expression
package for Java. I have come across several that (I think) are GNU or BSD
style licenses ...

While there appear to be a number of them, here is one that seems interesting
as it appears to behave like a lot of the gnu stuff, can use different regular
expression syntax (with a default to Perl5 systax which seems reasonable),
works on Strings, StringBuffers, characher arrays, and has a useful set of
methods ... Following the web link is a copy of the Javadoc from their RE
class

http://www.cacas.org/java/gnu/regexp/

public class RE
extends gnu.regexp.REToken

RE provides the user interface for compiling and matching regular expressions. 

A regular expression object (class RE) is compiled by constructing it from a
String, StringBuffer or
character array, with optional compilation flags (below) and an optional
syntax specification (see
RESyntax; if not specified, RESyntax.RE_SYNTAX_PERL5 is used). 

Various methods attempt to match input text against a compiled regular
expression. These methods are: 
  isMatch: returns true if the input text in its entirety matches the regular
expression pattern. 
  getMatch: returns the first match found in the input text, or null if no
match is found. 
  getAllMatches: returns an array of all non-overlapping matches found in the
input text. If no matches
are found, the array is zero-length. 
  substitute: substitute the first occurence of the pattern in the input text
with a replacement string
(which may include metacharacters $0-$9, see REMatch.substituteInto). 
  substituteAll: same as above, but repeat for each match before returning. 
  getMatchEnumeration: returns an REMatchEnumeration object that allows
iteration over the
matches (see REMatchEnumeration for some reasons why you may want to do this
instead of using
getAllMatches. 

These methods all have similar argument lists. The input can be a String, a
character array, a
StringBuffer, a Reader or an InputStream of some sort. Note that when using a
Reader or InputStream,
the stream read position cannot be guaranteed after attempting a match (this
is not a bug, but a
consequence of the way regular expressions work). Using an REMatchEnumeration
can eliminate most
positioning problems. 

The optional index argument specifies the offset from the beginning of the
text at which the search should
start (see the descriptions of some of the execution flags for how this can
affect positional pattern
operators). For a Reader or InputStream, this means an offset from the current
read position, so
subsequent calls with the same index argument on a Reader or an InputStream
will not necessarily be
accessing the same position on the stream, whereas repeated searches at a
given index in a fixed string
will return consistent results. 

You can optionally affect the execution environment by using a combination of
execution flags (constants
listed below). 

Three other packages that appear to be affiliated with either gnu or apache
include:

http://jakarta.apache.org/regexp/
http://jakarta.apache.org/oro/
http://www.crocodile.org/~sts/Rex/

While I'm probably not sufficiently expert to select the best of these, it
appears as if we can easily get a package that may help us in terms of
searches as well as data validation ... for example, if I'm not mistaken, a
valid account number for us is: [129][A-Z][A-Z][A-Z][0-9][0-9][0-9] ... I'm
sure that there's a shorter way of saying 1, 2 or 9 followed by 3 capital
letters followed by 3 digits in regular expressionese, but even that isn't too
bad ... Presumably things like this could be stored in the database to make
validation of data easily configurable at different sites.

Thanks,

John


From hopcroft at snf.stanford.edu  Sat Apr 14 05:04:37 2001
From: hopcroft at snf.stanford.edu (Matt Hopcroft)
Date: Sat, 14 Apr 2001 05:04:37 -0700 (PDT)
Subject: Coral Down
Message-ID: <Pine.LNX.4.04.10104140503030.3218-100000@snf.stanford.edu>


	Coral crashed this morning. Various "AdminManager IOR" error
messages. Please disable anything still enabled in my name.


Matt Hopcroft
hopcroft at snf.stanford.edu
650.736.2380



From shott at snf.stanford.edu  Sat Apr 14 11:45:00 2001
From: shott at snf.stanford.edu (John Shott)
Date: Sat, 14 Apr 2001 11:45:00 -0700
Subject: Java 1.3.1 Release Candidate ...
Message-ID: <3AD89AAC.37F237C0@snf.stanford.edu>

Bill and Mike:

I see that Sun released the Release Candidate for Java 1.3.1.  I have
downloaded it, made sure that we have all of the required OS patches
installed, and then installed it in /usr/j2se on sunray, guilden, bopeep, and
rosen.  I also made sure that it didn't overwrite our external packages (ABA,
JCE, etc. in /usr/j2se/jre/lib/ext and that our java.security and java.policy
files in /usr/j2se/jre/lib/security are as we had them prior to this install.

My guess is that once we are happy with all of Bill's database/code changes
(which I think we are ... no problems during the week) that we will want to:

1. Go into remaining Java files and "indent region" them using JDE and save
them all ... followed by a tagging of Coral 1.3.1.

2. Then, I think, that we may be able to make yet another attempt at getting
j2se 1.3 to fly by making appropriate changes ... which, I think, I can do and
then see if we can run on guilden without problems. If that works, then we can
see if we can run for more than an hour on sunray/rosen using j2se 1.3 ... and
hopefully get finally upgraded to 1.3.

Not only do I think that we'd like to upgrade to 1.3 and take advantage of the
presumably better system (including the new idlj converter) ... but it would
make it a heck of a lot easier to tell Penn State, go get 1.3.1-RC and install
it instead of 1.2.2 plus you need idltojava, etc.

Comments?

Thanks,

John


From beckwith at snf.stanford.edu  Tue Apr 17 13:51:52 2001
From: beckwith at snf.stanford.edu (Sharleen Beckwith)
Date: Tue, 17 Apr 2001 13:51:52 -0700 (PDT)
Subject: No coral screen appears when I log in. Everything else is here.
Message-ID: <Pine.LNX.4.04.10104171351160.26343-100000@snf.stanford.edu>




From bmurray at snf.stanford.edu  Tue Apr 17 14:31:43 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Tue, 17 Apr 2001 14:31:43 -0700 (PDT)
Subject: Inconsistent Data
Message-ID: <Pine.LNX.4.04.10104171426570.6965-100000@snf.stanford.edu>

John,

According to the resource manager log, we have some inconsistent data.
The following occurred sometime this afternoon:

Default project and/or account are/is not in the works_on or charges_to relation
s.  Inconsistent data for anolley.
Project 2DYA426 has no accounts.
Creating ticket for mmlee without project 2DYA426
Default project and/or account are/is not in the works_on or charges_to relation
s.  Inconsistent data for mmlee.

This error may already be fixed because it was logged earlier:

Project 9VAV411is in works_on, but not in the project table.
Creating ticket for ericp without project 9VAV411

Bill





From hopcroft at snf.stanford.edu  Tue Apr 17 15:22:59 2001
From: hopcroft at snf.stanford.edu (Matt Hopcroft)
Date: Tue, 17 Apr 2001 15:22:59 -0700 (PDT)
Subject: equipment qualification
Message-ID: <Pine.LNX.4.04.10104171519170.30070-100000@snf.stanford.edu>


Hi,

	I'm having trouble getting added to the user list for tylanfga.
The request has been made several times, I've logged in/out, waited
several hours, etc., but I haven't received the confirmation email. Is
the system working? Thanks,

Matt Hopcroft
hopcroft at snf.stanford.edu
650.736.2380



From hopcroft at snf.stanford.edu  Tue Apr 17 16:03:02 2001
From: hopcroft at snf.stanford.edu (Matt Hopcroft)
Date: Tue, 17 Apr 2001 16:03:02 -0700 (PDT)
Subject: equipment qualification
In-Reply-To: <Pine.LNX.4.04.10104171519170.30070-100000@snf.stanford.edu>
Message-ID: <Pine.LNX.4.04.10104171602260.30070-100000@snf.stanford.edu>


	It is working fine, I spoke too soon.

Matt Hopcroft
hopcroft at snf.stanford.edu
650.736.2380

On Tue, 17 Apr 2001, Matt Hopcroft wrote:

> 
> Hi,
> 
> 	I'm having trouble getting added to the user list for tylanfga.
> The request has been made several times, I've logged in/out, waited
> several hours, etc., but I haven't received the confirmation email. Is
> the system working? Thanks,
> 
> Matt Hopcroft
> hopcroft at snf.stanford.edu
> 650.736.2380
> 
> 



From bmurray at snf.stanford.edu  Wed Apr 18 08:49:52 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Wed, 18 Apr 2001 08:49:52 -0700 (PDT)
Subject: SocketPermission
Message-ID: <Pine.LNX.4.04.10104180849010.13407-100000@snf.stanford.edu>

Tom,

Take a look at SocketPermission on this page:

http://java.sun.com/j2se/1.3/docs/guide/security/permissions.html

Bill



From shott at snf.stanford.edu  Wed Apr 18 21:42:28 2001
From: shott at snf.stanford.edu (John Shott)
Date: Wed, 18 Apr 2001 21:42:28 -0700
Subject: gnu.regexp-1.1.2
Message-ID: <3ADE6CB4.A9C8CAA4@snf.stanford.edu>

Bill and Mike:

Although I don't know that it is necessarily the best of the lot, I've
downloaded gnu.regexp (Version 1.1.2, I think) that may be useful with use in
clients for checking the validity of certain input data.

Particularly for remote clients, the size of the jar file seems to make a
difference ... the size of this is slighlty less than 32 kb which strikes me
as being not terribly big.  Particularly if we download it as a separate jar
file (figuring that it will change less frequently than our client), it would
seem that downloading it once in addition to our (currently 500 kb) client
isn't too bad.

A link to the web site that includes API and syntax and usage notes is:

http://www.cacas.org/java/gnu/regexp/

No big deal ... but I thought I'd pass this on while I knew where it was
located.

Talk to you later,

John


From beckwith at snf.stanford.edu  Thu Apr 19 16:01:31 2001
From: beckwith at snf.stanford.edu (Sharleen Beckwith)
Date: Thu, 19 Apr 2001 16:01:31 -0700 (PDT)
Subject: the mouse at the terminal next to wbnonmetal doesn't work, it
Message-ID: <Pine.LNX.4.04.10104191600160.17569-100000@snf.stanford.edu>


only moves to the right, not the left...must be the Bush era!

Sharleen



From bmurray at snf.stanford.edu  Thu Apr 19 16:45:33 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Thu, 19 Apr 2001 16:45:33 -0700 (PDT)
Subject: the mouse at the terminal next to wbnonmetal doesn't work, it
In-Reply-To: <Pine.LNX.4.04.10104191600160.17569-100000@snf.stanford.edu>
Message-ID: <Pine.LNX.4.04.10104191643140.16654-100000@snf.stanford.edu>

Sharleen,

Mike or I will take a look.  

Thanks,
Bill

PS This Bush era is taking it's toll on me!  His environmental policy upset
   me so much that I became a "life member" of the Sierra Club at a cost of
   $1250.  I'm going to go broke if this guy keeps it up!

On Thu, 19 Apr 2001, Sharleen Beckwith wrote:

> 
> only moves to the right, not the left...must be the Bush era!
> 
> Sharleen
> 



From bmurray at snf.stanford.edu  Fri Apr 20 00:26:09 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 20 Apr 2001 00:26:09 -0700 (PDT)
Subject: Coral 1.33 Up & Running
Message-ID: <Pine.LNX.4.04.10104192358340.20285-100000@snf.stanford.edu>

John and Mike,

I just finished testing and installing the last round of "code clean up" 
changes.  These include formatting all files with the new editor, removing
unused code including the accounting and activity servers and clients,
logging all equipment disables, and the addition of the USE_HARDWARE_
INTERLOCK configuration parameter.  For a number of we reasons we can
discuss later, I placed the new parameter in Server.conf not eqmgr.conf.
I set the parameter to true in Server.conf and false in Server.conf.dev.
Now we automatically get the interlocks disabled in the development
version of the code so that we can freely enable/disable any equipment,
not just the bottlewash.

I have tagged the latest version as Coral 1.33 (in CVS: coral-1_33).

When you are ready to get the newest version of the code, I recommend
that you use "cvs update -P" in your labnet directory.  The -P will
prune all empty directories from your working copy.  I have removed
a number of directories from the repository.  However, CVS never gets
rid of these directories in case we wish to return to an old version,
but you don't want them in your working directories.

I have not created a new version of remote Coral because I do not 
believe this is necessary and didn't want to create hassles for the
191 remote users.  Mike: when you arrive tomorrow, could you run
your remote version of Coral to confirm all is ok?  If there are any
problems, please call me at home tomorrow morning.  I may be a little
late arriving tomorrow since I'm still up at midnight.  

Finally, the equipment manager had been up for close to a week
and exceeded maximum open cursors this evening (before I installed the
new version).  I noticed the problem fairly quickly and restarted the
servers.  As result, some transactions were rolled back which means
the database will be completely consistent.  However, the cache may
have gotten out of synch with the database.  This is a relatively minor
problem.  I mention because I expect to see Len Booth at my door tomorrow
to let me know some piece of equipment is enable/disabled and the light
isn't correct in Coral.  I have saved the log in lastlog, but I don't
think it provides any real info.  I may need to take a look at this
probolem?

Good night,
Bill
problem right away and










From mbell at snf.stanford.edu  Fri Apr 20 07:04:34 2001
From: mbell at snf.stanford.edu (Mike Bell)
Date: Fri, 20 Apr 2001 07:04:34 -0700
Subject: Coral 1.33 Up & Running
References: <Pine.LNX.4.04.10104192358340.20285-100000@snf.stanford.edu>
Message-ID: <3AE041F2.2ACF3AAC@snf.stanford.edu>

Bill,

Remote Coral seems to still work.

Mike

Bill Murray wrote:

> John and Mike,
>
> I just finished testing and installing the last round of "code clean up"
> changes.  These include formatting all files with the new editor, removing
> unused code including the accounting and activity servers and clients,
> logging all equipment disables, and the addition of the USE_HARDWARE_
> INTERLOCK configuration parameter.  For a number of we reasons we can
> discuss later, I placed the new parameter in Server.conf not eqmgr.conf.
> I set the parameter to true in Server.conf and false in Server.conf.dev.
> Now we automatically get the interlocks disabled in the development
> version of the code so that we can freely enable/disable any equipment,
> not just the bottlewash.
>
> I have tagged the latest version as Coral 1.33 (in CVS: coral-1_33).
>
> When you are ready to get the newest version of the code, I recommend
> that you use "cvs update -P" in your labnet directory.  The -P will
> prune all empty directories from your working copy.  I have removed
> a number of directories from the repository.  However, CVS never gets
> rid of these directories in case we wish to return to an old version,
> but you don't want them in your working directories.
>
> I have not created a new version of remote Coral because I do not
> believe this is necessary and didn't want to create hassles for the
> 191 remote users.  Mike: when you arrive tomorrow, could you run
> your remote version of Coral to confirm all is ok?  If there are any
> problems, please call me at home tomorrow morning.  I may be a little
> late arriving tomorrow since I'm still up at midnight.
>
> Finally, the equipment manager had been up for close to a week
> and exceeded maximum open cursors this evening (before I installed the
> new version).  I noticed the problem fairly quickly and restarted the
> servers.  As result, some transactions were rolled back which means
> the database will be completely consistent.  However, the cache may
> have gotten out of synch with the database.  This is a relatively minor
> problem.  I mention because I expect to see Len Booth at my door tomorrow
> to let me know some piece of equipment is enable/disabled and the light
> isn't correct in Coral.  I have saved the log in lastlog, but I don't
> think it provides any real info.  I may need to take a look at this
> probolem?
>
> Good night,
> Bill
> problem right away and



From kliu at molecularreflections.com  Fri Apr 20 13:31:01 2001
From: kliu at molecularreflections.com (Kelvin Liu)
Date: Fri, 20 Apr 2001 13:31:01 -0700
Subject: running coral behind a firewall
Message-ID: <A7C127D35E08D511B38500E01830542B5A39@SERVER>

Hello,
	I am an industrial user at snf.  I can successfully access Coral
using my laptop from my broadband connection at home or from my dial-up
service but anytime I try to run Coral from behind our company firewall, it
tells me that it "Cannot access equipment information."  I feel it is
because of the firewall. What can I do to remedy this situation?  We don't
really have a network administrator at our company.  The firewall is a
hardware imposed firewall that is implemented through our router.  Thanks
for your help.

Kelvin Liu
Staff Engineer
Molecular Reflections
6330 Nancy Ridge Drive #107
San Diego, CA 92121
Tel :(858) 623-0259
Fax: (858) 623-0331



From shott at snf.stanford.edu  Fri Apr 20 13:57:06 2001
From: shott at snf.stanford.edu (John Shott)
Date: Fri, 20 Apr 2001 13:57:06 -0700
Subject: running coral behind a firewall
References: <A7C127D35E08D511B38500E01830542B5A39@SERVER>
Message-ID: <3AE0A2A2.CE83016D@snf.stanford.edu>

Kelvin:

Yes, currently Remote Coral doesn't run from behind a firewall ... and the
signature of that failure is the "Cannot access equipment information".  We
probably need to update our warning when that error is generated to suggest
that it is likely a firewall/networking issue.

The reason that this happens is that we use an industry-standard client-server
communication protocal know variously as CORBA (Common Object Request Broker
Architecture) or IIOP (Internet Inter-Orb Protocol) ... unfortunately, this
standard is sufficiently new that most firewalls don't understand this
protocol or have the ability to allow it selectively to pass through the
firewall.  In fact, the only firewall that knows how to do this is produced by
a company that sells an ORB (Object Request Broker) ... and that probably
isn't even a solution as that company makes a different ORB than we actually
use.

So, what can you do?

The easiest thing to do is to keep using your laptop and a dialup connection
...

I don't know whether your router has the ability to "open a hole" to a
specific IP address.  If so, this would be a possiblity ... but it also makes
your firewall less secure.

What can/will we do to fix this?  A couple of possibilities?

1. In some sense, we may have to do nothing ... firewall vendors may begin to
recognize and deal with CORBA and may provide provisions for this in the
future.

2. We are beginning to keep an eye on a related technology known as SOAP
(Simple Object Access Protocol, I think) which packages messages in a form
that they can be accessed through the HTTP protocal that already passes
through firewalls.  I've seen some evidence that there are people working on
SOAP to CORBA bridges that may allow us to circumvent this problem in the
future.  None of this, however, is going to happen in the near future and
depends on development by folks other than ourselves before we could begin to
try to use their stuff.

I hope this helps,

Good luck,

John


From bmurray at snf.stanford.edu  Fri Apr 20 21:46:26 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 20 Apr 2001 21:46:26 -0700 (PDT)
Subject: Oracle Initialization Parameters
Message-ID: <Pine.LNX.4.04.10104202142050.4242-100000@snf.stanford.edu>

John and Mike,

I have increased the values of a number of Oracle initialization parameters
from "small" to "medium".  (I also increased the maximum number of open
cursors from 50 to 100.)  I stopped and started the Coral servers and
database at 9:30pm tonight so that all is now in effect.

Bill



From shott at snf.stanford.edu  Sat Apr 21 12:57:26 2001
From: shott at snf.stanford.edu (John Shott)
Date: Sat, 21 Apr 2001 12:57:26 -0700
Subject: Stored procedures ...
Message-ID: <3AE1E626.796AD592@snf.stanford.edu>

Bill and Mike:

I think that I now have 4 stored procedures that should make it easier to do
some of our reporting stuff.  Two belong to "activity" and two belong to
"accounting".  The activity procedures are activity_summary() and
equipment_summary() (with source files activity_summary_source.sql and
equipment_summary_source.sql, respectively).  These are found in
labnet/activity/billing.  The accounting procedures are accounting_summary()
and nnun_summary() that are found in labnet/accounting/billing.

I have also put a README file and a Makefile in each directory.  Each makefile
includes the SQL username (actmgr or accmgr, respectively) ... but the
password is left blank so that it needs to be supplied (and isn't given away
in the makefile).

I have committed all of these in the repository.  While they sould replace the
files activity_summary.sql and accounting_summary.sql (that had "hardwired"
dates), I haven't yet removed those two files from the repository until I'm
100% sure that these are OK ... although I did test them by re-running
February and March activity/accounting to make sure that things match up.

So, I think that my next step is to get going on the testing of JDK 1.3.1 (on
the development environment, of course).

Bill, I sort of hate to tag yet another version before going to 1.3.1 and can
probably easily "re-commit" these changes if we do need to fall back to 1.2. 
If, however, it is safer to re-tag, let me know and I can probably do that
too.

Thanks,

John


From shott at snf.stanford.edu  Sun Apr 22 12:17:40 2001
From: shott at snf.stanford.edu (John Shott)
Date: Sun, 22 Apr 2001 12:17:40 -0700
Subject: Coral with Java 1.3.1 running on Guilden ...
Message-ID: <3AE32E54.D3483A38@snf.stanford.edu>

Bill and Mike:

I've made the changes required to run coral using Java 1.3.1 Release Candidate
and have done a "make prod-install" on Guilden.  The servers that are
currently running are the new ones.

I haven't checked in any of these changes ...

One interesting thing:

In the past, whenever I tried to run the development client from rosen, for
example, it would be the really slow and poorly behaving response that we
attributed to bad/slow x-tunneling through ssh.  While it is still somewhat
slow, all of the windows come up properly ... for example, I don't have to hit
"Esc" to force a redraw and I no longer have to guess where some of the "OK"
buttons are on display panels that don't show up at all.

I don't know whether this is an improvement in JDK 1.3.1 or whether we are
seeing an improvement in the newer ssh client (which has probably also changed
in the last month or so) ... but, for whatever reason, I can get noticeably
better response ... it is actually useable.

So, it will be interesting to see how long the servers appear to run on the
development side using Java 1.3.1 ... I've got my fingers crossed.

Thanks,

John

p.s. Just for reference, the only deprecation warning that I get was in the
file labnet/util/AgentImpl.java.  It claims that the method extract_Principal
from org.omg.CORBA.Any is deprecated ... I don't know whether this is new or
something that we've been seeing for a long time.


From shott at snf.stanford.edu  Wed Apr 25 19:08:14 2001
From: shott at snf.stanford.edu (John Shott)
Date: Wed, 25 Apr 2001 19:08:14 -0700
Subject: XML and XML Schema ...
Message-ID: <3AE7830E.3E4A1B5D@snf.stanford.edu>

Bill and Mike:

While I am (as is often the case) at the limits of my understanding ...
I'm beginning to read about proposals and current implementations of XML
Schemas that, as near as I can tell, seem to deal with the question of
defining datatypes (including all sorts of contstraints) in XML documents.  As
I sit here reading this, I'm thinking to myself ... this is just what Mike
needs for setting up and digesting the input that will ultimately allow new
members to submit and/or change their member information, for example.  As
Bill and I were discussing this afternoon, it appears as if this really allows
a lot of power and flexibility to build process libraries, runsheets, etc. out
of XML documents.

I've been looking over a document at:

http://www.w3.org/TR/2001/PR-xmlschema-2-20010330/datatypes.html

For example, the definition of a simple datatype like a string includes the
following "constraining facets", to use their terminology, that include:


         string has the following constraining facets: 

              length 
              minLength 
              maxLength 
              pattern 
              enumeration 
              whiteSpace 

Note: Pattern in a regular expression, and enumeration is a list of acceptable
values.

While I may be not quite understanding what I'm reading, I get the impression
that by incorporating some of this schema stuff ... we could generate an XML
document that would represent a "member" that would have the smarts to detect
an invalid field, etc.

Also, the Apache folks (and presumably IBM and other XML "heavies" already
seem to have tools that support this stuff.  For example:

http://xml.apache.org/xerces-j/schema.html describes the Apache XML package
that seems to work with a lot of this stuff (xerces-j is the Apache Java XML
parser ...

Later,

John


From bmurray at snf.stanford.edu  Thu Apr 26 07:01:47 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Thu, 26 Apr 2001 07:01:47 -0700 (PDT)
Subject: XML and XML Schema ...
In-Reply-To: <3AE7830E.3E4A1B5D@snf.stanford.edu>
Message-ID: <Pine.LNX.4.04.10104260656570.21079-100000@snf.stanford.edu>

John,

This sounds like the answer.  It gives us everything we need including
the regular expressions for validation of data.  As soon as I complete
the rewrite of the area stuff, I'll check this out so that Mike can
remained focused on getting the resource client gui completed.

This also gives a jump on the process library analyis and design.

Thanks,
Bill

On Wed, 25 Apr 2001, John Shott wrote:

> Bill and Mike:
> 
> While I am (as is often the case) at the limits of my understanding ...
> I'm beginning to read about proposals and current implementations of XML
> Schemas that, as near as I can tell, seem to deal with the question of
> defining datatypes (including all sorts of contstraints) in XML documents.  As
> I sit here reading this, I'm thinking to myself ... this is just what Mike
> needs for setting up and digesting the input that will ultimately allow new
> members to submit and/or change their member information, for example.  As
> Bill and I were discussing this afternoon, it appears as if this really allows
> a lot of power and flexibility to build process libraries, runsheets, etc. out
> of XML documents.
> 
> I've been looking over a document at:
> 
> http://www.w3.org/TR/2001/PR-xmlschema-2-20010330/datatypes.html
> 
> For example, the definition of a simple datatype like a string includes the
> following "constraining facets", to use their terminology, that include:
> 
> 
>          string has the following constraining facets: 
> 
>               length 
>               minLength 
>               maxLength 
>               pattern 
>               enumeration 
>               whiteSpace 
> 
> Note: Pattern in a regular expression, and enumeration is a list of acceptable
> values.
> 
> While I may be not quite understanding what I'm reading, I get the impression
> that by incorporating some of this schema stuff ... we could generate an XML
> document that would represent a "member" that would have the smarts to detect
> an invalid field, etc.
> 
> Also, the Apache folks (and presumably IBM and other XML "heavies" already
> seem to have tools that support this stuff.  For example:
> 
> http://xml.apache.org/xerces-j/schema.html describes the Apache XML package
> that seems to work with a lot of this stuff (xerces-j is the Apache Java XML
> parser ...
> 
> Later,
> 
> John
> 



From LWells at sirostech.com  Thu Apr 26 13:41:32 2001
From: LWells at sirostech.com (Linda Wells)
Date: Thu, 26 Apr 2001 13:41:32 -0700
Subject: FW: FW: Coral Access
Message-ID: <E1C9813E4449D411A70000A0C9EEA2F5099135@MAIL>



-----Original Message-----
From: mfitzpatrick at mail.sirostech.com
[mailto:mfitzpatrick at mail.sirostech.com]
Sent: Thursday, April 26, 2001 10:25 AM
To: LWells at sirostech.com
Subject: RE: FW: Coral Access


Of course we're using a firewall.  Anyone not using a firewall these days is
crazy.  They will need to give very exact specifications on what their
program requires in order to change the firewall to enable this.

Mike

Original Message:
-----------------
From: Linda Wells LWells at sirostech.com
Date: Thu, 26 Apr 2001 09:51:48 -0700
To: Administrator at sirostech.com
Subject: FW: Coral Access


Hi Mike,

I'm forwarding this to you hoping that you can help.

Linda

-----Original Message-----
From: John Shott [mailto:shott at snf.stanford.edu]
Sent: Thursday, April 26, 2001 9:48 AM
To: Linda Wells
Subject: Re: Coral Access


Linda:

Do you mean when you are trying to use Remote Coral?

The "Cannot access equipment information!" is the error that many people
who
are trying to access Coral from behind a firewall are seeing.  Do you
know if
you are behind a firewall at Siros?

Thanks,

John

--------------------------------------------------------------------
Mail2Web - Check your email from the web at
http://www.mail2web.com/ .
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://snf.stanford.edu/pipermail/coral/attachments/20010426/1ecc2e6b/attachment.html>

From bmurray at snf.stanford.edu  Fri Apr 27 09:37:51 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 27 Apr 2001 09:37:51 -0700 (PDT)
Subject: FW: FW: Coral Access
In-Reply-To: <E1C9813E4449D411A70000A0C9EEA2F5099135@MAIL>
Message-ID: <Pine.LNX.4.04.10104270927230.3843-100000@snf.stanford.edu>

Linda,

As your rather diplomatic system administrator has pointed out, everyone
uses firewalls today.  Although most system administrators at Stanford shun
firewalls, the SNF has most of its machines behind a firewall also.  However
like most companies, we place a few machines outside the firewall for 
applications of questionable security.  Even if we were willing and/or
able to give exact specifications on how to  open your firewall, would a
competent system administrator do it?  If I were a system admin, I probably
wouldn't.  I would place a machine outside the firewall for applications
like this one.

Bill

On Thu, 26 Apr 2001, Linda Wells wrote:

> 
> 
> -----Original Message-----
> From: mfitzpatrick at mail.sirostech.com
> [mailto:mfitzpatrick at mail.sirostech.com]
> Sent: Thursday, April 26, 2001 10:25 AM
> To: LWells at sirostech.com
> Subject: RE: FW: Coral Access
> 
> 
> Of course we're using a firewall.  Anyone not using a firewall these days is
> crazy.  They will need to give very exact specifications on what their
> program requires in order to change the firewall to enable this.
> 
> Mike
> 
> Original Message:
> -----------------
> From: Linda Wells LWells at sirostech.com
> Date: Thu, 26 Apr 2001 09:51:48 -0700
> To: Administrator at sirostech.com
> Subject: FW: Coral Access
> 
> 
> Hi Mike,
> 
> I'm forwarding this to you hoping that you can help.
> 
> Linda
> 
> -----Original Message-----
> From: John Shott [mailto:shott at snf.stanford.edu]
> Sent: Thursday, April 26, 2001 9:48 AM
> To: Linda Wells
> Subject: Re: Coral Access
> 
> 
> Linda:
> 
> Do you mean when you are trying to use Remote Coral?
> 
> The "Cannot access equipment information!" is the error that many people
> who
> are trying to access Coral from behind a firewall are seeing.  Do you
> know if
> you are behind a firewall at Siros?
> 
> Thanks,
> 
> John
> 
> --------------------------------------------------------------------
> Mail2Web - Check your email from the web at
> http://www.mail2web.com/ .
> 



From dwarnold at eksigent.com  Fri Apr 27 11:12:50 2001
From: dwarnold at eksigent.com (Don W. Arnold)
Date: Fri, 27 Apr 2001 11:12:50 -0700
Subject: problem with remote coral
Message-ID: <000401c0cf45$ac4e2770$0201a8c0@eksigent.com>

Hi,

I am trying to get remote Coral to run at our location in Livermore and I am
having a little difficulty. I downloaded Java Web Start and the Coral
application to my computer here at work (NT2000) and my computer at home
(Windows 98). The Java Web Start demo applications worked at work and home.
My system at home works just fine with Coral as well but the office computer
comes back with an error message that it "Cannot access equipment
information!" Any suggestions?

Thanks,
Don



From bmurray at snf.stanford.edu  Fri Apr 27 11:11:48 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 27 Apr 2001 11:11:48 -0700 (PDT)
Subject: afs client issues
In-Reply-To: <4.3.1.2.20010425125554.00b6d8b0@kam.pobox.stanford.edu>
Message-ID: <Pine.LNX.4.04.10104271106170.3843-100000@snf.stanford.edu>

Kristen,

John and I had given up on AFS last year because we had problems with the
NT AFS client.  We really liked the Mac AFS client.  Since we are truely
a multi-platform environment, AFS has  a lot of appeal.  Escpecially since
we're already paying for 1Gb of space, and ITSS does all the backups and
restores.  We have an immediate problem that would be solved if the NT
AFS client worked properly.  So John is going to downloaded the latest
AFS client and try it again.  If we have any problems, could we get the
name and email address of your Windows developer?

Thanks,
Bill

PS If I know John, he's probably downloading the client right now so we'll
   know if we have problems by lunch time.

On Wed, 25 Apr 2001, Kristen Murray wrote:

> Bill,
> IF you have time (and I know you probably don't), could you describe the 
> problems you have had with the Windows AFS client?  We have a dedicated 
> Windows developer now who wants to hear about all rumors of problems, so 
> that he can investigate.
> 
> Thanks,
> Kristen
> 



From bmurray at snf.stanford.edu  Fri Apr 27 11:14:40 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 27 Apr 2001 11:14:40 -0700 (PDT)
Subject: problem with remote coral
In-Reply-To: <000401c0cf45$ac4e2770$0201a8c0@eksigent.com>
Message-ID: <Pine.LNX.4.04.10104271113160.3843-100000@snf.stanford.edu>

Don,

Coral will not run behind a firewall.  If you have a firewall at work, you
should ask your system administrator about access to a machine outside the
firewall.

Bill 


On Fri, 27 Apr 2001, Don W. Arnold wrote:

> Hi,
> 
> I am trying to get remote Coral to run at our location in Livermore and I am
> having a little difficulty. I downloaded Java Web Start and the Coral
> application to my computer here at work (NT2000) and my computer at home
> (Windows 98). The Java Web Start demo applications worked at work and home.
> My system at home works just fine with Coral as well but the office computer
> comes back with an error message that it "Cannot access equipment
> information!" Any suggestions?
> 
> Thanks,
> Don
> 



From bmurray at snf.stanford.edu  Fri Apr 27 13:04:53 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 27 Apr 2001 13:04:53 -0700 (PDT)
Subject: Demo @ JavaOne
Message-ID: <Pine.LNX.4.04.10104271303020.3843-100000@snf.stanford.edu>

John,

I'm not sure our application qualifies for these demos to be highlighted
by Sun during the Keynote program at JavaOne.  Take a look at this web
page when you get a chance:

http://servlet.java.sun.com/javaone/conf/keynotes/16032-sf2001.jsp

Thanks,
Bill




From bmurray at snf.stanford.edu  Fri Apr 27 13:09:06 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Fri, 27 Apr 2001 13:09:06 -0700 (PDT)
Subject: JavaOne Session Abstracts
Message-ID: <Pine.LNX.4.04.10104271308140.3843-100000@snf.stanford.edu>

John,

JavaOne session abstracts are now available on the JavaOne web site at

http://servlet.java.sun.com/javaone/index.jsp

Bill



From bmurray at snf.stanford.edu  Sat Apr 28 14:15:36 2001
From: bmurray at snf.stanford.edu (Bill Murray)
Date: Sat, 28 Apr 2001 14:15:36 -0700 (PDT)
Subject: Status
Message-ID: <Pine.LNX.4.04.10104281408370.17630-100000@snf.stanford.edu>

John and Mike,

I have completed and tested all the changes required to remove the area 
and laboratory interfaces.  Everything works great.  However, I'm still
able to reproduce the JDK 1.3.1 error where certain machines are not
displayed in the equipment hierarchy.  For this reason, I have not installed
my changes in production or checked them into the repository yet.  

During my testing, I have been able to gather a little more information on
the problem.  When I return to the office on Monday, I will spend some
time tracking this bug down.  Now that I have removed the final pieces of
student code, it may be easier to track the problem down.

See you on Monday,
Bill



From rory at siwaveinc.com  Mon Apr 30 16:39:00 2001
From: rory at siwaveinc.com (Rory McKenna)
Date: Mon, 30 Apr 2001 16:39:00 -0700 (PDT)
Subject: Steve Vargo's remote access to coral
Message-ID: <Pine.LNX.4.30.0104301638430.11794-100000@allspice>

Hi,

  I'm trying to help Steve with his computer problem.
Whenever he tries to access Coral remotely from
the Java Web Start, he enters in the username and
password, and it returns with "authentication failed".

Can he somehow verify his username and password with
someone over there ?  That is what the application suggested.

Thanks,

Rory McKenna
IT consultant for SiWave, Inc.