[cciug] Unsent Message Returned to Sender

From: POSTMASTER (POSTMASTER@alltel.com)
Date: Thu Jan 06 2000 - 01:10:08 EST

Next message: jay@sony.co.in: "[cciug] Date problem"
Previous message: Paul D. Smith: "Re: [cciug] Any problems exporting views in CC 4.0?"
Next in thread: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Notice to Sender
================

This message was received by this installation but could not be
delivered to its intended cc:Mail recipient(s).

Original subject: cciug-digest V2 #723

Intended recipient(s) who DID NOT receive this message:

tracy.king..@alltel.com

The following cc:Mail error(s) were recorded:

*** Message recipient is unknown ***

-------- Original Message Text --------

cciug-digest Wednesday, January 5 2000 Volume 02 : Number 723

----------------------------------------------------------------------

Date: Tue, 4 Jan 2000 22:29:23 -0600
From: Brad Appleton <bradapp@enteract.com>
Subject: Re: [cciug] How do you handle JAVADOC output ?

On Wed, Jan 05, 2000 at 05:38:12AM +0200, Dan Lupu wrote:
> I would like to have a nightly run of Javadoc, to update project
> documentation.
> I would also like to have the Javadoc output under a vob, so developers can
> access the documentation no matter which platform they use.

Sounds like a "nightly build" to me. For the most part, I'm inclined
to treat it like a software build where it's documentation that is
built rather than executables (and the target derived objects are
the .html files). In this case, it sounds like you want to "install"
the docs once they are built by publishing them to a repository.

Ask yourself if your environment and project is such that it is
legitimately necessary to version control the derived objects (your
HTML output from javadoc) or if it's sufficient to simply VC the source
from which they were derived. This will help you decide if you should
install/publish your HTML output docs to a VOB or to a normal filesystem
structure (either of which can be used to serve up webpages if you like).

> I could create a vob only for Javadoc output and remove +ACY- recreate it every
> night (?+ACE-).
> Any ideas ?
>
> Dan Lupu
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- --
Brad Appleton <bradapp@enteract.com> http://www.enteract.com/~bradapp/
"And miles to go before I sleep." -- Robert Frost
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 06:31:45 +0000 (GMT)
From: Christian Goetze <cg@digisle.net>
Subject: [cciug] Showstopper ClearCase 4.0 bug.

I am experimenting with ClearCase 4.0, and I am wondering if the following
bug is specific to my (somewhat peculiar) installation of clearcase or if
this can be reproduced at other places. Please give it a shot, it's really
trivial to test (replace names with whatever works on your site - I've
tried it with different view/element combinations and I get consistant
results):

% ct -version
ClearCase version 4.0 (Fri Nov 19 09:58:04 EST 1999)
@(#) MVFS version 4.0 (Fri Oct 29 20:20:51 EDT 1999)
cleartool V4.0 (Thu Nov 11 20:16:22 EST 1999)
db_server V4.0 (Fri Nov 5 20:12:00 EST 1999)
VOB database schema version: 53

Here's the problem:

The file xxx contains:

ls -short /view/main/vob/history/history
setcs -tag main /ccase/config_specs/main

I do:

% cat xxx | cleartool
/view/main/vob/history/history@@/main/192
cleartool: Error: Operation requires a valid network path to the view
storage.
Not supported on a view restricted handle.
cleartool: Error: Cannot get config spec for view: error detected by
ClearCase subsystem.

This is a show stopper for me (and yes, this _does_ work on 3.2.1 - been
using this for more than 2 years).

Note that it only happens if I try to setcs the view I accessed via the
view extended pathname.

I tried this on two different installations at two different customer
sites. Same problem.

Here's the rest of my info (you're probably going to come back and ask for
it again, but then...)

+=================== COMPANY INFO HERE HERE HERE ====================+

   Digital Island
   45 Fremont Street #1200
   San Francisco CA 94105
   415.738.4323 (email preferred)
   cg@digisle.net

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 08:04:30 +0000 (GMT)
From: Christian Goetze <cg@digisle.net>
Subject: RE: [cciug] Showstopper ClearCase 4.0 bug.

One thing I forgot to mention was that running the two commands
separately, like this:

cleartool ls -short /view/main/vob/history/history
cleartool setcs -tag main /ccase/config_specs/main

causes no problem whatsoever.
- --
cg

> -----Original Message-----
> From: Christian Goetze
> To: cciug@Rational.Com
> Sent: 1/4/00 10:31 PM
> Subject: [cciug] Showstopper ClearCase 4.0 bug.
>
>
> I am experimenting with ClearCase 4.0, and I am wondering if the
> following bug is specific to my (somewhat peculiar) installation of
> clearcase or if this can be reproduced at other places. Please give it
> a shot, it's really trivial to test (replace names with whatever works
> on your site - I've tried it with different view/element combinations
> and I get consistant results):
>
> % ct -version
> ClearCase version 4.0 (Fri Nov 19 09:58:04 EST 1999)
> @(#) MVFS version 4.0 (Fri Oct 29 20:20:51 EDT 1999)
> cleartool V4.0 (Thu Nov 11 20:16:22 EST 1999)
> db_server V4.0 (Fri Nov 5 20:12:00 EST 1999)
> VOB database schema version: 53
>
> Here's the problem:
>
> The file xxx contains:
>
> ls -short /view/main/vob/history/history
> setcs -tag main /ccase/config_specs/main
>
> I do:
>
> % cat xxx | cleartool
> /view/main/vob/history/history@@/main/192
> cleartool: Error: Operation requires a valid network path to the view
> storage.
> Not supported on a view restricted handle.
> cleartool: Error: Cannot get config spec for view: error detected by
> ClearCase subsystem.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 09:03:31 +0000
From: Neal Welland <wellanna@cvsumailhost.marconicomms.com>
Subject: Re: [cciug] CC 3.2 patches 55,56 for HP10 and Sun5

Mark Howell wrote:

> Has anyoune had any problems with patches 55 and 56 on HP 10.20 or
> Solaris 2.6?
>

We installed patches 55 and 56 on a Solaris 2.6 box a couple of weeks ago.
No obvious problems. What are your symptoms?

Neal.

>
> --
> ========================================================================
> | Mark Howell | Email: mark.howell@motorola.com |
> | BAS TETRA Engineering | smtp: markho@baseng.comm.mot.com |
> | UK Engineering LMPS (Europe) | X400 Mail: BMH007@email.mot.com |
> | MOTOROLA | Voice: +44 -01 256 484046 |
> | Viables, Basingstoke | Fax: +44 -01 256 484561 |
> | Hampshire, UK RG22 4PD | Maildrop: EUR-UK, ZUK02 |
> ========================================================================
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- --

                               @@@@@@@@@
                              @@ ~ ~ @@
                               ( * * )
============================-oOOo-(_)-oOOo-=============================
Neal A. Welland Marconi Communications
Phone : +44 (0)24 7656 2197 New Century Park
Fax : +44 (0)24 7656 2826 PO Box 53
Email : Neal.Welland@marconicomms.com Coventry. CV3 1HJ.
                                     Oooo. England
============================-.oooO-=-( )-=============================
!!!!!!!!!!!!!!!!!!!!!!!!!! ( ) ) /
!!USE FIXED FONT TO VIEW!! \ ( (_/
!!!!!!!!!!!!!!!!!!!!!!!!!! \_)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 11:02:32 -0000
From: "Batlin, Alex" <alex.batlin@nomura.co.uk>
Subject: [cciug] view-storage-pname

I am ashamed to be asking this question, however, it's something that is
bothering me. When registering or tagging a view or a VOB, when one
specifies the tripple path, one also specifies the view-storage-pname, which
is usually the same as the host-stg-pname.

Why is this the case and can the view-storage-pname ever be different to
either host-stg-pname or global-stg-pname.

> mktag -view -tag view-tag [-tcomment tag-comment] [-replace]
> [-nstart] [-ncaexported] [-region network-region]
> [-host hostname -hpath host-stg-pname -gpath global-stg-pname]
> view-storage-pname

Regards,

Alex.

- ----------------------------------------------------------------------------
- --------
Alex Batlin | alex.batlin@nomura.co.uk
Nomura International plc, Nomura House | Phone: +44 (0)20 7521 3728
1 St. Martin's-le-Grand, London EC1A 4NP | Fax: +44 (0)20 7521 3686

- ----------------------------------------------------------------------------
- --------
The information contained in this message is intended for the named
recipients
only. If you are not an intended recipient of this message, you must not
copy,
distribute or take any action in reliance on it and you should notify the
sender
immediately. E-mail transmission cannot be guaranteed to be secure or
error-free
as information could be intercepted, corrupted, lost, destroyed, arrive
late or
incomplete, or contain viruses, and Nomura International plc excludes
liability for
(1) any errors or omissions in the contents of this message which arise as
a result
of e-mail transmission and (2) the transmission of any viruses. If
verification is
required please request a hard-copy version. Nomura International plc is
regulated
by the Securities and Futures Authority Limited and is a member of the
London
Stock Exchange.

- ----------------------------------------------------------------------------
- --------
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 13:58:17 EET
From: "Bittoo Bhalla" <bittoo_08@hotmail.com>
Subject: [cciug] Merging Problem

Hi List ,
We are using multisite ,we have vob for project say "A" at one site with
/vobs/a as mount point of vob inside this vob there are several
directories.Developers use to work on a branch of this vob .People out there
at this site has created a sync packet and sent it to me to be merged to the
main branch.They have applied a label to the changed elements .Now the
problem is if i try to merge to /main branch of vob a using command:
/vobs/a>cleartool . findmerge -ftag <relevant_view_tag> -print

I cannot see any elements for merge.The config specs of the
<relevant_view_tag> has a proper rule for selecting the changed elements
with Label which was applied there at other site.
I think they had not applied label to the directory inside which the
elements are labelled .Now if i use command :
/vobs/a>cleartool . findmerge -fversion /main/<relevant_branch>/LATEST
- -print ,it shows me a list of number of elements which need merge ,but I
want to merge only the labelled elements with label they have applied .So
please help me ,I would be more than thanful for the same.'

Sincerely,
Bittoo
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 13:12:06 +0100
From: "EXTERN Chapman Roy (WA-Consultants; K5/ESQ1)" <Roy.Chapman@de.bosch.com>
Subject: AW: [cciug] view-storage-pname

Don't be ashamed, this one has foxed me for years as well. When passing the
triplet (gpath, hpath and the path) I have found, from trying every
combination known, that it will only accept the triplet is all three are the
same. I therefore always pass all three as global pathnames (under Unix
this would be /net/hostname/etc/...) and it invariably works. Now I am told
that I should be able to pass a differnt global and local pathname
(different as in the local pathname does not include /net/hostname) but I
have tried. And as Homer Simpson once said, "Trying is the first step on
the long road to failure".

Roy Chapman
External Consultant

Robert Bosch GmbH * (00 49) 0711/811-31795 / GSM (00 49)
0173/4887139
K5/ESQ * (00 49) 0711/811-31800
Postfach 20 02 40 * mailto:roy.chapman@de.bosch.com
D-70442 Stuttgart

> ----------
> Von: Batlin, Alex[SMTP:alex.batlin@nomura.co.uk
> Gesendet: Mittwoch, 5. Januar 2000 12:02
> An: CCIUG (E-mail)
> Betreff: [cciug] view-storage-pname
>
>
> I am ashamed to be asking this question, however, it's something that is
> bothering me. When registering or tagging a view or a VOB, when one
> specifies the tripple path, one also specifies the view-storage-pname,
> which
> is usually the same as the host-stg-pname.
>
> Why is this the case and can the view-storage-pname ever be different to
> either host-stg-pname or global-stg-pname.
>
> > mktag -view -tag view-tag [-tcomment tag-comment] [-replace]
> > [-nstart] [-ncaexported] [-region network-region]
> > [-host hostname -hpath host-stg-pname -gpath
> global-stg-pname]
> > view-storage-pname
>
> Regards,
>
> Alex.
>
>
>
> --------------------------------------------------------------------------
> --
> --------
> Alex Batlin | alex.batlin@nomura.co.uk
> Nomura International plc, Nomura House | Phone: +44 (0)20 7521 3728
> 1 St. Martin's-le-Grand, London EC1A 4NP | Fax: +44 (0)20 7521 3686
>
> --------------------------------------------------------------------------
> --
> --------
> The information contained in this message is intended for the named
> recipients
> only. If you are not an intended recipient of this message, you must not
> copy,
> distribute or take any action in reliance on it and you should notify the
> sender
> immediately. E-mail transmission cannot be guaranteed to be secure or
> error-free
> as information could be intercepted, corrupted, lost, destroyed, arrive
> late or
> incomplete, or contain viruses, and Nomura International plc excludes
> liability for
> (1) any errors or omissions in the contents of this message which arise
> as
> a result
> of e-mail transmission and (2) the transmission of any viruses. If
> verification is
> required please request a hard-copy version. Nomura International plc is
> regulated
> by the Securities and Futures Authority Limited and is a member of the
> London
> Stock Exchange.
>
> --------------------------------------------------------------------------
> --
> --------
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 08:01:47 -0500 (EST)
From: Greg Dickie <greg@discreet.com>
Subject: RE: [cciug] Trigger Problem

try changing the trigger to -exec "perl
\\yosemite\ccstg_d\Triggers\INHIBIT_DUPLICATE_ELE.pl"

Thats what I had to do.
The more I try to work with NT, the more I hate it ;-(

HTH
Greg

On 04-Jan-00 Robin LaSalle wrote:
>
> Due to all the recent talk about triggers on the list, I decided to try
> implementing the "evil twin" trigger that I downloaded from the ABS website.
> Alas! I created the trigger type, attempted to create an element and was
> kindly told that it could not be executed.
>
> exact error message:
>
> cleartool: Error: Can't exec
> ""\\yosemite\ccstg_d\Triggers\INHIBIT_DUPLICATE_ELE
> .pl"":
> cleartool: Error: Trigger action "-exec
> \\yosemite\ccstg_d\Triggers\INHIBIT_DUPL
> ICATE_ELE.pl" unable to run: Exec format error
> cleartool: Error: Unable to create element "testmkelem.txt".
>
> YET! if you just type in the path to the perl script in a command window,
> the perl script executes perfectly. Also, I have many triggers that simply
> execute "false.exe" and those work fine. This is an all NT environment.
>
> Help!
>
> Robin La Salle
> Software Configuration Manager
> WavTrace, Inc.
> (425) 468-2244
> rlasalle@wavtrace.com
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- ---------------------------------------------------------------------
Greg Dickie
Just A Guy*
*from discreet (the logic is gone)
Montreal
(514) 954-7171
greg@discreet.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 08:16:43 -0500
From: David Boyce <dsb@world.std.com>
Subject: RE: [cciug] Showstopper ClearCase 4.0 bug.

I'd say you have a real bug here. I was able to reproduce it (you can
simplify the test case to use setcs -current BTW). Also, it works fine if
you reverse the order of the commands. So report it. When a bug has a nice
little test case like this, you (we) have a good chance of getting it fixed
pretty quickly.

Can't help mentioning the UUOC.

- -David B

At 08:04 AM 1/5/00 +0000, Christian Goetze wrote:

>One thing I forgot to mention was that running the two commands
>separately, like this:
>
>cleartool ls -short /view/main/vob/history/history
>cleartool setcs -tag main /ccase/config_specs/main
>
>causes no problem whatsoever.
>--
>cg
>
> > -----Original Message-----
> > From: Christian Goetze
> > To: cciug@Rational.Com
> > Sent: 1/4/00 10:31 PM
> > Subject: [cciug] Showstopper ClearCase 4.0 bug.
> >
> >
> > I am experimenting with ClearCase 4.0, and I am wondering if the
> > following bug is specific to my (somewhat peculiar) installation of
> > clearcase or if this can be reproduced at other places. Please give it
> > a shot, it's really trivial to test (replace names with whatever works
> > on your site - I've tried it with different view/element combinations
> > and I get consistant results):
> >
> > % ct -version
> > ClearCase version 4.0 (Fri Nov 19 09:58:04 EST 1999)
> > @(#) MVFS version 4.0 (Fri Oct 29 20:20:51 EDT 1999)
> > cleartool V4.0 (Thu Nov 11 20:16:22 EST 1999)
> > db_server V4.0 (Fri Nov 5 20:12:00 EST 1999)
> > VOB database schema version: 53
> >
> > Here's the problem:
> >
> > The file xxx contains:
> >
> > ls -short /view/main/vob/history/history
> > setcs -tag main /ccase/config_specs/main
> >
> > I do:
> >
> > % cat xxx | cleartool
> > /view/main/vob/history/history@@/main/192
> > cleartool: Error: Operation requires a valid network path to the view
> > storage.
> > Not supported on a view restricted handle.
> > cleartool: Error: Cannot get config spec for view: error detected by
> > ClearCase subsystem.
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 08:16:38 -0500
From: kleach2@csc.com
Subject: Re: [cciug] view-storage-pname

The view-storage-pname and the host-storage-pname should always be the same
path. However, the global-storage-pname may differ depending on your
installation. The hpath is telling the registry that this is the path in which
the view-storage directory, for any given View, can be found on the host, which
houses the local disk, on which the storage area resides. The gpath refers to
the location at which the view-storage area is accessed on all other
workstations within the network.

If you create a view-storage directory called test.vws under /export/viewstore
on host mustang and then mount /export/viewstore to /export/viewstore on all
other workstations, the path for the view-storage-pname, hpath and gpath will be
the same.

However, if you mount /export/viewstore from host mustang to
/net/export/viewstore the view-storage-pname and the hpath will be the same but
the gpath will be /net/export/viewstore/test.vws as this is the structure that
exists on all other workstations. It's all dependent on how you access directory
structures on the workstations Vs the servers.

Hope this helps.

alex.batlin@nomura.co.uk on 01/05/2000 06:02:32 AM

To: cciug@rational.com
cc: (bcc: Kenneth R Leach/CIV/CSC)

Subject: [cciug] view-storage-pname

Why is this the case and can the view-storage-pname ever be different to
either host-stg-pname or global-stg-pname.

Regards,

Alex.

- ----------------------------------------------------------------------------
- --------
The information contained in this message is intended for the named
recipients
only. If you are not an intended recipient of this message, you must not
copy,
distribute or take any action in reliance on it and you should notify the
sender
immediately. E-mail transmission cannot be guaranteed to be secure or
error-free
as information could be intercepted, corrupted, lost, destroyed, arrive
late or
incomplete, or contain viruses, and Nomura International plc excludes
liability for
(1) any errors or omissions in the contents of this message which arise as
a result
of e-mail transmission and (2) the transmission of any viruses. If
verification is
required please request a hard-copy version. Nomura International plc is
regulated
by the Securities and Futures Authority Limited and is a member of the
London
Stock Exchange.

- ----------------------------------------------------------------------------
- --------
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 14:22:05 +0100
From: Lennart Larsson <lennart.larsson@emw.ericsson.se>
Subject: Re: [cciug] list of standard triggers

On Mon, 3 Jan 2000, Mark Dahlman wrote:

> We have been running with a fairly raw installation of
> CC for the past several months. As I am learning more
> about CC, I am realizing that both the users and the CC
> admins need to protect themselves from undesired actions.
> Hence, we need to be more robust with regard to triggers.

> Could you please share with me what you consider to be a
> fairly standard list of triggers an organization should
> adopt?

We have three mandatory triggers. Those are on rmelem, rmbranch, and
rmver. The triggers protect
version with labels. For example, you can't do rmelem on an element with
any labels on it.

The reasons is that some labels identify importent configurations. Such
as cutomer relese,
verified alfa versions with protocol, etc. Those labels are of course
locked or obsolete.

An alternative is to forbid the use of those commands for all except vob
owner, but I hear from one responsible from a site with that aproach,
that it turned out to be a lot of work for the VOB owner.

Lennart Larsson
ClearCase responsible
Ericsson Microwave Systems AB
Sweden
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 13:23:39 -0000
From: Ford David <David.Ford@Tms-Ltd.com>
Subject: [cciug] ClearCase 4.0

Hello CCIUG,

When can we mere mortals get hold of a copy of ClearCase 4.0? Apparently we
have to request it rather than Rational automatically sending it? Is this
true.

David Ford
David Ford
Software Configuration Manager
Thomson Marconi Sonar Ltd

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 13:24:07 -0000
From: "Batlin, Alex" <alex.batlin@nomura.co.uk>
Subject: RE: [cciug] view-storage-pname

Thanks for confirming my suspicions. It still seems silly to me to have to
specify the same parameter twice, but that question really must be addressed
to Rational!

Thanks,

Alex.

- -----Original Message-----
From: kleach2@csc.com [mailto:kleach2@csc.com
Sent: 05 January 2000 13:17
To: alex.batlin@nomura.co.uk; cciug@Rational.Com
Subject: Re: [cciug] view-storage-pname

The view-storage-pname and the host-storage-pname should always be the same
path. However, the global-storage-pname may differ depending on your
installation. The hpath is telling the registry that this is the path in
which
the view-storage directory, for any given View, can be found on the host,
which
houses the local disk, on which the storage area resides. The gpath refers
to
the location at which the view-storage area is accessed on all other
workstations within the network.

If you create a view-storage directory called test.vws under
/export/viewstore
on host mustang and then mount /export/viewstore to /export/viewstore on all
other workstations, the path for the view-storage-pname, hpath and gpath
will be
the same.

However, if you mount /export/viewstore from host mustang to
/net/export/viewstore the view-storage-pname and the hpath will be the same
but
the gpath will be /net/export/viewstore/test.vws as this is the structure
that
exists on all other workstations. It's all dependent on how you access
directory
structures on the workstations Vs the servers.

Hope this helps.

alex.batlin@nomura.co.uk on 01/05/2000 06:02:32 AM

To: cciug@rational.com
cc: (bcc: Kenneth R Leach/CIV/CSC)

Subject: [cciug] view-storage-pname

Why is this the case and can the view-storage-pname ever be different to
either host-stg-pname or global-stg-pname.

Regards,

Alex.

- ----------------------------------------------------------------------------
- --------
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 14:07:03 +0000
From: "Olivier Dehon" <dehon_olivier@jpmorgan.com>
Subject: [cciug] Retrieve branching points

After reading and re-reading the manual about the query_language,
I couldn't find an easy solution to the following problem:

What query would yield the version number off which a branch was created
on an element. Something like the predecessor of the element 0 on that
branch.

This question has been asked in the past on this list, but I wonder if
someone ever since managed to find a solution that does not use a
Perl script, or maybe if this particular problem has been addressed in
V4.0.

Thanks and regards, -Olivier

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 09:15:02 -0500
From: "K.Spence" <kspence@modes.tc.faa.gov>
Subject: Re: [cciug] ClearCase 4.0

I requested my copy via the web.

<http://www.rational.com/sitewide/support/upgrades>
http://www.rational.com/sitewide/support/upgrades

Kim

- ----- Original Message -----
From: Ford David <David.Ford@Tms-Ltd.com>
To: Cciug (E-mail) <cciug@Rational.Com>
Sent: Wednesday, January 05, 2000 8:23 AM
Subject: [cciug] ClearCase 4.0

>
> Hello CCIUG,
>
> When can we mere mortals get hold of a copy of ClearCase 4.0? Apparently
we
> have to request it rather than Rational automatically sending it? Is this
> true.
>
> David Ford
> David Ford
> Software Configuration Manager
> Thomson Marconi Sonar Ltd
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 09:29:46 -0500
From: Ron Natalie <ron@sensor.com>
Subject: Re: [cciug] ClearCase 4.0

Ford David wrote:

>
> When can we mere mortals get hold of a copy of ClearCase 4.0? Apparently we
> have to request it rather than Rational automatically sending it? Is this
> true.
>
Unfortunately, it is, just another sign of the overall degrade of
tech support since Rational took over.

http://www.rational.com/cgi-bin/www/fuji_cc.cgi

Now getting manuals is next to impossible. I don't know who they sent
the danged postcard to at my site but I never got it. They don't even
have the ability to request or order Clearcase 4.0 manuals on the web
page.

Furthermore, my tech support guy has been spinning his wheels for two
weeks now trying to get our supposed permanent license renewed, we've
been short of two of our licenses since the end of december.
Complicating
matters is that we used to be able to work through our sales rep
to get these problems worked out (It used to be that ATRIA would contact
us when the licenses and support contracts were about to expire, but
they don't do that anymore) but that's hard to pin down because
evidentally there has been much turnover in the local office of late.

While supporting clearcase has never been an easy task, lately it's
turned into a real nightmare. Support was never able to identify
why the 3.2.1 lockmanager dumps core at our site, which means we are
stuck at 3.1.1 (which Rational has abandoned support for). We can't
upgrade to 4.0 because our server machines are TIER II. And of course,
then there was the continual Y2K lies from Rational. We were told
a rosy picture that Clearcase was Y2K compatible. THis seemed odd
since they didn't qualify it with a version, so a month later they
said, OK all version 3 releases are Y2K comaptible. In actuallity,
it turns out that NO 3.anything releases were Y2K compatible without
patches. You still can't get a CDROM from RATIONAL and expect it
to work without the tediuous method of downloading multiple patches
for each and every platfrom you have to use individually.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 10:11:27 -0500
From: Gladys Christopher-Knight <gladys@gscmail.gsfc.nasa.gov>
Subject: Re: [cciug] Merging Problem

Bittoo:

I think what is happening here is that you are performing the merge
to /main on the original (which masters /main), which is correct.
The view (-ftag) you are trying to use does not live on the same
machine as the original vob. It lives on the replica machine (vob a).
Unless your view storage is located on a machine (say machineX) where
both machines (original vob machine and vob a machine) can see this
area, I don't think you do the merge via -ftag. Why not do the merge
via label? Just ask the replica site to make sure that labels are
applied to all relevant files/directories (ct findmerge . -fversion
LABELNAME -print/-merge)

Gladys Christopher-Knight
_________
>Hi List ,
>We are using multisite ,we have vob for project say "A" at one site
>with /vobs/a as mount point of vob inside this vob there are several
>directories.Developers use to work on a branch of this vob .People
>out there at this site has created a sync packet and sent it to me
>to be merged to the main branch.They have applied a label to the
>changed elements .Now the problem is if i try to merge to /main
>branch of vob a using command:
>/vobs/a>cleartool . findmerge -ftag <relevant_view_tag> -print
>
>I cannot see any elements for merge.The config specs of the
><relevant_view_tag> has a proper rule for selecting the changed
>elements with Label which was applied there at other site.
>I think they had not applied label to the directory inside which the
>elements are labelled .Now if i use command :
>/vobs/a>cleartool . findmerge -fversion
>/main/<relevant_branch>/LATEST -print ,it shows me a list of number
>of elements which need merge ,but I want to merge only the labelled
>elements with label they have applied .So please help me ,I would be
>more than thanful for the same.'
>
>Sincerely,
>Bittoo
>______________________________________________________
>Get Your Private, Free Email at http://www.hotmail.com
>
>- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 17:55:36 EET
From: "Bittoo Bhalla" <bittoo_08@hotmail.com>
Subject: [none]

Hi Gladys Christopher-Knight,
Thanks for early reponse. Yes I am applying the merge on /main (master).Here
we have two view servers and one vob server which has multitool installed .I
am using the right view ,it is being created on one of these view servers,it
used to work this way earlier .I used to apply sync packet on the vob server
(with multitool installed )and then from the view server I do the merge,all
the views are created on the same view server m/c.
I, am Sorry that I forgot to mention one thing that when I give command :
cleartool . findmerge -ftag <relevant_view_tag> -print
I get the output as only three files ,which needs merge.

when I used command :ct findmerge . -fversion
LABELNAME -print/-merge) as suggested by you I get the same output message .

Where as when I give command with -fverion /main/<relevant_branch>/LATEST
then I get huge listing of files which needs merge .
Any suggestions ?

Again A bundle of thanks to you....
Sincerely,
Bittoo

- ----------------------------------------------------------------------
I think what is happening here is that you are performing the merge
to /main on the original (which masters /main), which is correct.
The view (-ftag) you are trying to use does not live on the same
machine as the original vob. It lives on the replica machine (vob a).
Unless your view storage is located on a machine (say machineX) where
both machines (original vob machine and vob a machine) can see this
area, I don't think you do the merge via -ftag. Why not do the merge
via label? Just ask the replica site to make sure that labels are
applied to all relevant files/directories (ct findmerge . -fversion
LABELNAME -print/-merge)

Gladys Christopher-Knight
_________
>Hi List ,
>We are using multisite ,we have vob for project say "A" at one site with
>/vobs/a as mount point of vob inside this vob there are several
>directories.Developers use to work on a branch of this vob .People out
>there at this site has created a sync packet and sent it to me to be merged
>to the main branch.They have applied a label to the changed elements .Now
>the problem is if i try to merge to /main branch of vob a using command:
>/vobs/a>cleartool . findmerge -ftag <relevant_view_tag> -print
>
>I cannot see any elements for merge.The config specs of the
><relevant_view_tag> has a proper rule for selecting the changed elements
>with Label which was applied there at other site.
>I think they had not applied label to the directory inside which the
>elements are labelled .Now if i use command :
>/vobs/a>cleartool . findmerge -fversion /main/<relevant_branch>/LATEST
>-print ,it shows me a list of number of elements which need merge ,but I
>want to merge only the labelled elements with label they have applied .So
>please help me ,I would be more than thanful for the same.'
>
>Sincerely,
>Bittoo

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 08:55:47 -0700
From: "Grasso, Elyse" <egrasso@access-health.com>
Subject: RE: [cciug] view-storage-pname

My installations have different values for the global and host pathnames. I
have found that it actually works better to have the view-storage-pname
match the global-storage-name.

If sysadmin in their infinite wisdom decide to move somebody's home
directory to a different drive, the global path doesn't change ( all home
directories are mounted as /home/whatever on all machines per local policy)
and I can use cleartool register -replace and cleartool mktag -replace to
fix things. If the view-storage-pname matches the old local path, cleartool
tends to get cranky.

There have been a few cases of views that were usable for a couple of months
after the host-pname changed without the registry being updated.

- -----Original Message-----
From: kleach2@csc.com [mailto:kleach2@csc.com
Sent: Wednesday, January 05, 2000 6:17 AM
To: alex.batlin@nomura.co.uk; cciug@Rational.Com
Subject: Re: [cciug] view-storage-pname

The view-storage-pname and the host-storage-pname should always be the same
path. However, the global-storage-pname may differ depending on your
installation. The hpath is telling the registry that this is the path in
which
the view-storage directory, for any given View, can be found on the host,
which
houses the local disk, on which the storage area resides. The gpath refers
to
the location at which the view-storage area is accessed on all other
workstations within the network.

If you create a view-storage directory called test.vws under
/export/viewstore
on host mustang and then mount /export/viewstore to /export/viewstore on all
other workstations, the path for the view-storage-pname, hpath and gpath
will be
the same.

However, if you mount /export/viewstore from host mustang to
/net/export/viewstore the view-storage-pname and the hpath will be the same
but
the gpath will be /net/export/viewstore/test.vws as this is the structure
that
exists on all other workstations. It's all dependent on how you access
directory
structures on the workstations Vs the servers.

Hope this helps.

alex.batlin@nomura.co.uk on 01/05/2000 06:02:32 AM

To: cciug@rational.com
cc: (bcc: Kenneth R Leach/CIV/CSC)

Subject: [cciug] view-storage-pname

Why is this the case and can the view-storage-pname ever be different to
either host-stg-pname or global-stg-pname.

Regards,

Alex.

- ----------------------------------------------------------------------------
- --------
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 11:01:30 -0500
From: "Darby, David" <David.Darby@fmr.com>
Subject: RE: [cciug] Retrieve branching points

Olivier,

Are either of these what you are looking for?

ct desc -short -ancestor filename@@/main/branch/LATEST
filename@@/main/LATEST

ct desc -pred -short filename@@/main/branch/0

- -David

- -----Original Message-----
From: Olivier Dehon [mailto:dehon_olivier@jpmorgan.com
Sent: Wednesday, January 05, 2000 9:07 AM
To: cciug@Rational.Com
Subject: [cciug] Retrieve branching points

After reading and re-reading the manual about the query_language,
I couldn't find an easy solution to the following problem:

What query would yield the version number off which a branch was created
on an element. Something like the predecessor of the element 0 on that
branch.

Thanks and regards, -Olivier

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 16:19:33 +0000
From: Stephen Toop <steve@patriate.com>
Subject: Re: [cciug] list of standard triggers + [cciug] Trigger Problem

Hi, cciugers and all that have responded so far to these threads.

Having watched this thread over the last few days I feel that I must make
you aware of a product that
we at Patriate have developed to provide ease of process management and
trigger control. That product
is Access*Rite, and yes it is a commercial product. (Sorry in advance for
advertising it here).

If you wish to know more then read on. If not, thank you for getting this
far and sorry to waste your valuable trigger
coding time.

Access*Rite has been developed after a number of years of experience
working with ClearCase.

Some of our staff have been working with ClearCase for over 8 years now and
it is they who have said :-
'But we've seen and solved these issues before'.

The key to Access*Rite is that it is easy to use, does not require users to
understand how to implement the ClearCase trigger system and supports Unix
and NT, including interop setups.

Basically Access*Rite consists of a defined set of triggers which are
installed into a VOB and which control ALL the ClearCase operations
that are triggerable. No further triggers are required as any additional
process or control operations are managed through Access*Rite.

Any existing trigger definitions can co-exist with Access*Rite, or
(recommended) the scripts and functions can be added to the Access*Rite
external scripts sections.

As a brief overview I will illustrate some of the key features :

1. Trigger script order management.
2. User Groups
3. Performance.
4. Pattern Management.

If you are still here, thank you and I shall now explain further :

1. Trigger script order management.
One of the key issues when defining multiple triggers on a single ClearCase
operation is defining the order in which they fire ( I know that
there is an undocumented order, but Atria/PureAtria/Rational have always
claimed that it may change). Also there may well be the requirement to
allow other trigger scripts to run even if one of the multiples fail.
Access*Rite manages both of these conditions. Any external script or Perl
function can be added to the list for processing on any ClearCase
operation. The scripts be defined to run either before or after any of the
built-in Access*Rite process control features. Each script is given a
priority level, a continue on error flag, list of exempt users and/or
groups, branch names(wildcarded if required), pathnames and vob names and
also element type.

2. User Groups.
Another area of management headaches is that of defining a set of users
which might be exempt from any trigger activity. Using the base ClearCase
the -nuser switch is required with a list of individuals who are exempt.
Some manage this by building the list into each script they write, or use
attributes on branches etc( see Performance later). Access*Rite allows you
to define any number of groups of users and apply these groups to any part
of the Access*Rite process or external trigger scripts. Updating the user
list is then simply a matter of modifying the group definition list and all
operations using that group are updated ( aka LATEST).

3. Performance.
A key design concept of Access*Rite was to ensure that performance was not
adversely affected by the trigger system. To this end
all the Access*Rite cleartool commands have been optimised. Existing users
have not complained about the performance levels.

4. Pattern management.
Quite often trigger scripts need to work across multiple branches, paths,
labels, attributes etc. Base ClearCase requires upfront definition of the
restriction or inclusion list for each and every type name. Sometimes the
triggers are redefined to manage this list, or again each script developed
will have its own list of types. Both need constant management. This is
cumbersome if new type names are arriving on a frequent basis. Access*Rite,
allows trigger operations and process control to be defined for patterns of
types, thus ensuring that existing and new types matching the pattern are
automatically included in the process control.

These are just 4 of the many features of Access*Rite and I thank you for
your time in reading this far. If you would like to know more about the
product please don't hesitate to contact me or visit our web site.

As I said at the beginning, sorry for the commercial advertising here, but
after 8 years of ClearCase I feel I've earned the right to post at least
one advertisement.

In case you were wondering there are 16 mentions of the product :
Access*Rite in this mail.

Happy triggering. I'll just press another button on the GUI to add all
dev_* branches to control list.

Regards
Steve.

Steve Toop
Patriate Limited
Eagle House, The Ring
Bracknell, Berkshire,UK RG12 1HB
Tel: +44 (0)1344 646041 Mobile:+44 (0)860 364311
Main Tel: +44 (0)1344 382111
steve@patriate.com WWW: Patriate.com

"Providing Professional, Technical, Expertise for the Rational ClearCase
product family"

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 11:32:56 -0500
From: "Darby, David" <David.Darby@fmr.com>
Subject: RE: [cciug] view-storage-pname

Alex,

The -hpath & -gpath names are usually and should be different. The -hpath
argument is the local pathname that a process running on the "-host" machine
uses. The -gpath is used by all processes that are not on the "-host"
machine (usually specified with /net/hostname prefix or an equivalent) to
access the view-storage area. For performance reasons, the hpath should not
be a network wide pathname because it is meant to eliminate NFS when
accessing it locally.

I questioned the need for this final argument myself for a long time.
However, I discovered that it is possible to create a view while not
actually being on the intended view server host. So, the view-storage-pname
argument for the mkview command is the same as the -hpath argument when you
are creating a view local to the machine you are logged into. And, the
view-storage-pname argument for the mkview command is the same as the -gpath
argument when you are creating a view that is not on the machine you are
logged into.

Here's and example. Consider the case where you have some large view
server (powerful), but you are logged into your desktop box (wimpy). Your
mkview command would look like:

        ON WINMPY
        ---------
        ct mkview -tag myview -host powerful \
        -hpath /local/viewstorage/myview.vws \
        -gpath /net/powerful/local/viewstorage/myview.vws \
        /net/powerful/local/viewstorage/myview.vws

        ON POWERFUL
        -----------
        ct mkview -tag myview -host powerful \
        -hpath /local/viewstorage/myview.vws \
        -gpath /net/powerful/local/viewstorage/myview.vws \
        /local/viewstorage/myview.vws

In the first case, the final argument matches the "-gpath" argument
instead of the "-hpath" argument. I admit this is not a commonly used
feature of mkview, but this is why the final argument is required. I
suppose you could use this "feature" to allow users to create views on a
powerful view server - but not allow them to actually log into and work
directly on the view server.

Discovering this feature leads to the next question - why is the
view-storage-pname required and not defaulted to the "-gath" value? The
answer is, you could have some insane global mounting process where the
- -gpath you need for the view is not the same as the global path on your
local box... Also, what if you don't specify the optional triplet
arguments...

Thanks,
- -David

- -----Original Message-----
From: Batlin, Alex [mailto:alex.batlin@nomura.co.uk
Sent: Wednesday, January 05, 2000 8:24 AM
To: 'kleach2@csc.com'; CCIUG (E-mail)
Subject: RE: [cciug] view-storage-pname

Thanks for confirming my suspicions. It still seems silly to me to have to
specify the same parameter twice, but that question really must be addressed
to Rational!

Thanks,

Alex.

- -----Original Message-----
From: kleach2@csc.com [mailto:kleach2@csc.com
Sent: 05 January 2000 13:17
To: alex.batlin@nomura.co.uk; cciug@Rational.Com
Subject: Re: [cciug] view-storage-pname

The view-storage-pname and the host-storage-pname should always be the same
path. However, the global-storage-pname may differ depending on your
installation. The hpath is telling the registry that this is the path in
which
the view-storage directory, for any given View, can be found on the host,
which
houses the local disk, on which the storage area resides. The gpath refers
to
the location at which the view-storage area is accessed on all other
workstations within the network.

If you create a view-storage directory called test.vws under
/export/viewstore
on host mustang and then mount /export/viewstore to /export/viewstore on all
other workstations, the path for the view-storage-pname, hpath and gpath
will be
the same.

However, if you mount /export/viewstore from host mustang to
/net/export/viewstore the view-storage-pname and the hpath will be the same
but
the gpath will be /net/export/viewstore/test.vws as this is the structure
that
exists on all other workstations. It's all dependent on how you access
directory
structures on the workstations Vs the servers.

Hope this helps.

alex.batlin@nomura.co.uk on 01/05/2000 06:02:32 AM

To: cciug@rational.com
cc: (bcc: Kenneth R Leach/CIV/CSC)

Subject: [cciug] view-storage-pname

Why is this the case and can the view-storage-pname ever be different to
either host-stg-pname or global-stg-pname.

Regards,

Alex.

- ----------------------------------------------------------------------------
- --------
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 11:38:30 -0500 (EST)
From: stevenm <stevenm@comversens.com>
Subject: [cciug] Question regarding vob locks

We recently upgraded to ClearCase 3.2.1 and it was my understanding that
vob locks were a thing of the past with this release of ClearCase.... or
at least I was told that. Is this incorrect? The impression I am getting
is that the source in the vobs can be backed up without the vobs having
to be locked, but the d.o.'s need the vobs to be locked in order for them
to be backed up? Is this true?

Thanks in advance,
Steve MacDonald
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 17:43:13 +0100
From: vobadmsw@swb.siemens.de (VOB Administrator)
Subject: [cciug] Job opportunity

Configuration Manager for Embedded Systems Development Group

- -- Background:

   In the A&D (Automation and Drive Technology) Division of Siemens
   we develop and manufacture electronic control systems and drives
   for machine tools and production machines.
   RTS is a 100% subsidiary of Siemens and operates as an independent
   competence centre which provides project management and software
   design and development services for these products.

- -- Job description:

   Software integration and software configuration management for a
   development project of approx. 1 million LOC supporting up to
   40 separately marketed products.
   Responsibilities include system generation, support for the tool
   and CM infrastructure, and release management.
   The software development environment uses tools from Rational
   (ClearCase, etc) under Solaris in a multisite environment.

- -- Conditions of Employment:

Permament employment

- -- Experience Required:

   Unix, including shell programming; configuration management
   software. Knowledge of c / c++ and TCL is desirable.
   Working knowlege of both German and English is essential.

- -- Location:

Germany: Schwieberdingen (near Stuttgart)

- -- Company:

RTS Echtzeitsoftware GmbH

- -- Contact:

Send mailto:iain.kipps@swb.siemens.de
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 08:58:18 -0800
From: "Roger Wells" <RWells@illuminet.com>
Subject: [cciug] Templates in archive libraries with older cfront C++ compiler using ptrepository

We have several tens of thousands of lines of code that we have been maintaining for several years using wrapper scripts around RCS, working in C++ on HP-UX. We still use the older cfront-based compiler, which instantiates templates at link-time by stor
ing C-langage code fragments for the instantiations in a directory, by default called ptrepository. Much of our code is compiled to create archive libraries (lib*.a) and this archive library code uses templates extensively. We are converting to using Cl
earCase. We were well into the conversion when we discovered that special provisions have to be made for handling these ptrepository-based template instantiations. According to the (v. 3.2 UNIX) ClearCase User's Manual, there are three methods provided
for handling cfont templates. However, of the three, only one can be used with templates in archive libraries; this is called the "Forced Instantiation Model". This involves a (in my opinion) fairly cumbersome system of creating dum!
!
my source files with references to all the template classes or functions. These dummy sources must be maintained along with the actual source file; whenever a new template is added to the source files, the corresponding reference must be added to the du
mmy files.

This is discussed in p. 268 and pp. 278-282 of the ClearCase User's Manual for UNIX, v.3.2. I first wanted to confirm that we are reading the manual correctly and that the "Forced Instantiation Model" is our only ClearCase option (other than moving to t
he newer HP aCC compiler.) Second, if anyone has experience in using this "Forced Instantiation Model" and could advise on any techniques to simply creation of all the dummy source files. As I say, we are dealing with tens of thousands of lines of code
with extensive use of templates. I have created a simple lex script which seems to pull out all the template references from the source code. Does anyone have experience as to whether this is an effective way to start.

- - Roger Wells
  Sr. Project Engineer
  Network Applications, ILLUMINET
  rwells@illuminet.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 17:09:55 +0000
From: "Olivier Dehon" <dehon_olivier@jpmorgan.com>
Subject: RE: [cciug] Retrieve branching points

David,

Well, that's handy for one element, but what I wanted to achieve was a "ct find"
query that would return the elements that respond to the following criterion:

The element has a branch foo, and the predecessor of
eltname@@/.../foo/0 has label BAR

That's why I mentioned the query language. Maybe I'm missing something...

Thanks, -Olivier

David.Darby@fmr.com on 01/05/2000 04:01:30 PM

To: dehon_olivier@jpmorgan.com, cciug@rational.com
cc:
Subject: RE: [cciug] Retrieve branching points

Olivier,

Are either of these what you are looking for?

ct desc -short -ancestor filename@@/main/branch/LATEST
filename@@/main/LATEST

ct desc -pred -short filename@@/main/branch/0

- -David

- -----Original Message-----
From: Olivier Dehon [mailto:dehon_olivier@jpmorgan.com
Sent: Wednesday, January 05, 2000 9:07 AM
To: cciug@Rational.Com
Subject: [cciug] Retrieve branching points

After reading and re-reading the manual about the query_language,
I couldn't find an easy solution to the following problem:

What query would yield the version number off which a branch was created
on an element. Something like the predecessor of the element 0 on that
branch.

Thanks and regards, -Olivier

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 12:30:28 -0500 (EST)
From: VOB administrator <vobadm.to@fp.cibc.com>
Subject: [cciug] SunWSPro C++ 5.0 templates

CCIUGers;

Welcome to the year 19100. :-)

We are trying to move our software from Sun C++ 4.2 to Sun C++ 5.0, which
is more ANSI standard C++. We make some use of templates, and Sun has
changed their template model and directories completely with this release.

In the 3.2.1 Clearcase manual it talks a little about the Templates.DB/
subdirectory that was used by the C++ 4.2 compiler, and discusses whether
or not to use .SIBLINGS_IGNORED_FOR_REUSE and other issues.

What about 5.0? Does the Clearcase 4.0 docs reference their new
SunWS_cache/ cache model?

Or has anyone else made this work with build avoidance / winkins etc.

Thanks!
Scott Gregory as vobadm

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 13:38:08 -0500
From: Lezz Giles <lgiles@avici.com>
Subject: [cciug] Any problems exporting views in CC 4.0?

I have a call in to Rational support on this, but just in case anybody
else has
a suggestion or has seen the same problem...

We export a number of views to Linux boxes. This worked fine under CC
3.2, but
since we upgraded to 4.0 we've had problems seeing the mounted views on
the
Linux boxes - specifically, when we try to look inside the mounted
filesystem,
any access to the directories or files hangs. We can mount normal NFS
filesystems
with no problems, and we can mount the exported views on other Solaris
boxes
with no problems.

So does anybody else export views to Linux from CC 4.0? Is it working
fine or are
you seeing problems?

Thanks,
Lezz Giles

- --
Leslie Giles Software Release Manager Avici Systems Inc
lgiles@avici.com Phone: 978.964.2030 Fax: 978.964.2100

"Do not fear the merge"

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 12:41:41 -0500
From: Dale LaBossiere <labo@Rational.Com>
Subject: Re: [cciug] Question regarding vob locks

stevenm wrote:

> We recently upgraded to ClearCase 3.2.1 and it was my understanding that
> vob locks were a thing of the past with this release of ClearCase.... or
> at least I was told that. Is this incorrect? The impression I am getting
> is that the source in the vobs can be backed up without the vobs having
> to be locked, but the d.o.'s need the vobs to be locked in order for them
> to be backed up? Is this true?

VOB locks are still present and they are still required for backing up VOBs.
In V3.0 we introduced a "vob snapshot" backup mechanism. This mechanism
greatly reduces the duration of the VOB lock. Of course there are tradeoffs
involved. This topic is fully discussed in the ClearCase Administrator's
manual. There are a variety of other approaches that can be used to lessen
the lock duration (e.g., multisite, mirrored disks, filesystems that support
"fast" backup mechanisms).

- --
Dale LaBossiere mailto:labo@Rational.com
Rational Software http://www.Rational.com/

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 09:55:44 -0800
From: Jerry Groesser <jgroesse@msi-world.com>
Subject: [cciug] Job Opportunity - Bellevue Washington - (2) Open Reqs.

We currently recruiting for (2) CM Engineers for a very promising (pre-IPO)
company in Bellevue Wa. (20 minutes from Seattle).

Your primary responsibility will be building, validating and packaging
releases of our products in a multi-platform software development
environment as well as documenting CM activities as assigned by the
project/program managers. You will work with the product development teams
to help them set up and manage the proper source and build configurations.
You will maintain and enhance the build, test and configuration management
environments. And you must be able to debug and resolve problems
independently.

You will work in the engineering group to automate and optimize daily
product builds for QA and release builds for production, triage build
failures and work with development engineers to resolve problems.
Ultimately, this position will take the lead for designing and implementing
a comprehensive multisite configuration management system.

You will be part of a team that is responsible for defining performance
metrics and subsequent analysis in the effort to move the next SEI/CMM
level.

Qualifications
        1 - 3 years CM experience, (1) of which must be with clearcase
        1 - 3 years Multisite experience - Highly Desired
        1 - 3 years software development experience - Highly Desired
        1 - 3 years UNIX experience - (HP, SUN)
        Must have strong scripting skills (Perl desired)
        Hands on experience using either C++ or JAVA.
        UNIX System Admin experience - Highly Desired

        Strong planning skills.
        Excellent written and oral communications skills
        Ability to work independently in a team centered environment
        Experience with process development and implementation.

        Desired experience in one or more the following disciplines
                DDTS, ReqPro, TestExpert
                Experience in telecommunications industry - Highly desired
                Web development

Please send your resumes to via email or Fax to :

Attn:
Jerry Groesser
Director, Configuration Mgt./Ops
MSI

jgroesser@metapath.com

Fax : 425.519.2001

You may find out more about our company at : http://msi-world.com/ .
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 14:29:34 -0500 (EST)
From: Greg Dickie <greg@discreet.com>
Subject: RE: [cciug] Question regarding vob locks

Ah I don't think so, at least I have never seen anything that would indicate
that.

Greg

On 05-Jan-00 stevenm wrote:
>
> We recently upgraded to ClearCase 3.2.1 and it was my understanding that
> vob locks were a thing of the past with this release of ClearCase.... or
> at least I was told that. Is this incorrect? The impression I am getting
> is that the source in the vobs can be backed up without the vobs having
> to be locked, but the d.o.'s need the vobs to be locked in order for them
> to be backed up? Is this true?
>
>
> Thanks in advance,
> Steve MacDonald
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- ---------------------------------------------------------------------
Greg Dickie
Just A Guy*
*from discreet (the logic is gone)
Montreal
(514) 954-7171
greg@discreet.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 14:32:12 -0500 (EST)
From: Greg Dickie <greg@discreet.com>
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

Oh man I hate being tier 2, I want to play with 4.0. Stop taunting me Lezz! ;-)

Greg

On 05-Jan-00 Lezz Giles wrote:
>
> I have a call in to Rational support on this, but just in case anybody
> else has
> a suggestion or has seen the same problem...
>
> We export a number of views to Linux boxes. This worked fine under CC
> 3.2, but
> since we upgraded to 4.0 we've had problems seeing the mounted views on
> the
> Linux boxes - specifically, when we try to look inside the mounted
> filesystem,
> any access to the directories or files hangs. We can mount normal NFS
> filesystems
> with no problems, and we can mount the exported views on other Solaris
> boxes
> with no problems.
>
> So does anybody else export views to Linux from CC 4.0? Is it working
> fine or are
> you seeing problems?
>
> Thanks,
> Lezz Giles
>
> --
> Leslie Giles Software Release Manager Avici Systems
> Inc
> lgiles@avici.com Phone: 978.964.2030 Fax:
> 978.964.2100
>
> "Do not fear the merge"
>
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- ---------------------------------------------------------------------
Greg Dickie
Just A Guy*
*from discreet (the logic is gone)
Montreal
(514) 954-7171
greg@discreet.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 12:04:50 -0800
From: Prasad Herur <prasad@dotwireless.com>
Subject: [cciug] Idle statistics for Distributed build

Hi all,
I was trying to get distributed build working.
I am able to rsh to all the hosts with out password now.
Can anyone shed light on how I can get Idle statistics for all the servers.
All the servers are Solaris boxes.

Below is the error message I get.
===========================================
Build host status:
    Host XXXX refused: can't get idle statistics.
    Host YYYY refused: can't get idle statistics.
    Host ZZZZ refused: can't get idle statistics.
clearmake: Warning: No builds could be started because no build hosts are
acceptable, waiting 15 seconds...
============================================

Any help is greatly appreciated.

Prasad Herur
Dot Wireless
San Diego

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:05:17 -0500
From: "O'brien, Mark" <marko@UU.NET>
Subject: [cciug] FW: ClearCase Visual C++ problem...

Hello,
I am a developer who has recently introduced ClearCase into my
development
environment and I'm having troubling setting my include directories in
DevStudio. I created a snapshot view, went into my view, opened my project
file which started Visual C++, then I went to the "Tools->Options" menu and
selected the "Directories" tab. From there, I am required to set the search
paths for the all include directories so Visual C++ can compile my project.
The problem is, Visual C++ insists on using the absolute path which does not
work in a team environment since the absolute path is different for each
team member using Visual C++ and ClearCase because ClearCase requires every
team member to have a unique view name, which equates to a unique absolute
path. Here's an example...

In the Tools->Options->Directories tab I place the path to the include
directory where all of the public headers are located.

E:\snapshot\jswift_view\code\include

At that point, I can compile everything fine. Then, a team member of mine
comes along and checks everything out and tries to build and it fails
because the compiler cannot find the header file. After Rob examines the
project's include directory settings he sees that the path above is invalid
for his and should be...

E:\snapshot\rob_view\code\include

My question is how do we make both of these product work harmoniously in a
team environment? Simply adding every developers personal view path in
Visual C++ is NOT acceptable. Obviously we cannot all use a standard view
name such as...

E:\snapshot\project_view\code\include

...since only one of these can be created.

Anyway, I checked the documentation and I can't find anything on this
problem. I suspect it's something simple that I'm just missing.

Can someone please help?
js

John Swift
jswift@uu.net
UUNET Technologies, an MCI WorldCom Company.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 15:19:50 -0500
From: Ron Natalie <ron@sensor.com>
Subject: Re: [cciug] FW: ClearCase Visual C++ problem...

"O'brien, Mark" wrote:
>
>
> E:\snapshot\jswift_view\code\include

Go to the explorer and share E:\snapshot\jswift_view.
Then mount it as a drive letter and make all your VC++ paths
relative to the top of the view (\code\include). We do this
so that the snapshot and dynamic views are interchangable.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:32:22 -0500
From: John_Falasco@hyperion.com
Subject: [cciug] stopping network operatins from command line

On NT, Clearcase supports Snapshot views so that users of portable machines can
take their work home. When such a user boots up at home, Clearcase attempts to
mount vobs, stat views, etc, unless one remembers to UnCheck the option "Enable
Network Operations" in the Homebase Applet.

Almost everybody forgets to do this occasionally, and suffers the extra-long
bootup time as a consequence when they are disconnected from the lan. Is there
any way to disable this option from the command line rather than within the
applet? That would allow the creation of an icon on the desktop to enable and
disable network operations (in which case, the services would be set to run as
Manual instead of Automatic).

Thanks,
- -John

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:34:41 -0500
From: "Darby, David" <David.Darby@fmr.com>
Subject: RE: [cciug] Retrieve branching points

Dehon,

Ct find queries can not directly report what you are looking for.
However, some simple ksh code should achieve what you want (note: not
extensively tested):

for elem in $(cleartool find /vobtag -all -branch "brtype(branch)" -print) ;
do
Pred=$(cleartool desc -short -pred $elem/0)
[[ $(ct desc -fmt "%l" ${elem%@@*}@@$Pred) = *LABEL* ]] && print $elem
done

Thanks,
- -David

- -----Original Message-----
From: Olivier Dehon [mailto:dehon_olivier@jpmorgan.com
Sent: Wednesday, January 05, 2000 12:10 PM
To: David.Darby@fmr.com
Cc: Cciug@Rational.Com
Subject: RE: [cciug] Retrieve branching points

David,

Well, that's handy for one element, but what I wanted to achieve was a "ct
find"
query that would return the elements that respond to the following
criterion:

The element has a branch foo, and the predecessor of
eltname@@/.../foo/0 has label BAR

That's why I mentioned the query language. Maybe I'm missing something...

Thanks, -Olivier

David.Darby@fmr.com on 01/05/2000 04:01:30 PM

To: dehon_olivier@jpmorgan.com, cciug@rational.com
cc:
Subject: RE: [cciug] Retrieve branching points

Olivier,

Are either of these what you are looking for?

ct desc -short -ancestor filename@@/main/branch/LATEST
filename@@/main/LATEST

ct desc -pred -short filename@@/main/branch/0

- -David

- -----Original Message-----
From: Olivier Dehon [mailto:dehon_olivier@jpmorgan.com
Sent: Wednesday, January 05, 2000 9:07 AM
To: cciug@Rational.Com
Subject: [cciug] Retrieve branching points

After reading and re-reading the manual about the query_language,
I couldn't find an easy solution to the following problem:

What query would yield the version number off which a branch was created
on an element. Something like the predecessor of the element 0 on that
branch.

Thanks and regards, -Olivier

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 15:42:15 -0500 (EST)
From: Greg Dickie <greg@discreet.com>
Subject: RE: [cciug] stopping network operatins from command line

Someone had posted a really neat solution to this a while ago.Basically they
always set the srvices as "start on demand" and then explicitly did a "net
start albd" and "net start the_other_one"in the network logon script. Not on
the network=no logon script= no started CC services. We are looking at doing
this soon.

Greg

On 05-Jan-00 John_Falasco@hyperion.com wrote:
>
>
>
> On NT, Clearcase supports Snapshot views so that users of portable machines
> can
> take their work home. When such a user boots up at home, Clearcase attempts
> to
> mount vobs, stat views, etc, unless one remembers to UnCheck the option
> "Enable
> Network Operations" in the Homebase Applet.
>
> Almost everybody forgets to do this occasionally, and suffers the extra-long
> bootup time as a consequence when they are disconnected from the lan. Is
> there
> any way to disable this option from the command line rather than within the
> applet? That would allow the creation of an icon on the desktop to enable
> and
> disable network operations (in which case, the services would be set to run
> as
> Manual instead of Automatic).
>
> Thanks,
> -John
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- ---------------------------------------------------------------------
Greg Dickie
Just A Guy*
*from discreet (the logic is gone)
Montreal
(514) 954-7171
greg@discreet.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:42:34 -0500 (EST)
From: "David L. Paterline" <zcdlp@cnfd.pgh.wec.com>
Subject: RE: [cciug] Retrieve branching points

If I understand correctly what you want, I think this will work (in my example,
the branch name is 'Y2K'):

        ct find . -name '*' -ele 'brtype(Y2K)' \
        -exec 'cleartool desc -pred -short $CLEARCASE_PN@@/main/Y2K/0'

Olivier Dehon wrote:
> To: David.Darby@fmr.com
> Date: Wed, 5 Jan 2000 17:09:55 +0000
> Subject: RE: [cciug] Retrieve branching points
>
>
> David,
>
> Well, that's handy for one element, but what I wanted to achieve was a "ct
find"
> query that would return the elements that respond to the following criterion:
>
> The element has a branch foo, and the predecessor of
> eltname@@/.../foo/0 has label BAR
>
> That's why I mentioned the query language. Maybe I'm missing something...
>
> Thanks, -Olivier
>
> David.Darby@fmr.com on 01/05/2000 04:01:30 PM
>
> To: dehon_olivier@jpmorgan.com, cciug@rational.com
> cc:
> Subject: RE: [cciug] Retrieve branching points
>
> Olivier,
>
> Are either of these what you are looking for?
>
> ct desc -short -ancestor filename@@/main/branch/LATEST
> filename@@/main/LATEST
>
> ct desc -pred -short filename@@/main/branch/0
>
> -David
>
>
>
>
> -----Original Message-----
> From: Olivier Dehon [mailto:dehon_olivier@jpmorgan.com
> Sent: Wednesday, January 05, 2000 9:07 AM
> To: cciug@Rational.Com
> Subject: [cciug] Retrieve branching points
>
>
>
> After reading and re-reading the manual about the query_language,
> I couldn't find an easy solution to the following problem:
>
> What query would yield the version number off which a branch was created
> on an element. Something like the predecessor of the element 0 on that
> branch.
>
> This question has been asked in the past on this list, but I wonder if
> someone ever since managed to find a solution that does not use a
> Perl script, or maybe if this particular problem has been addressed in
> V4.0.
>
> Thanks and regards, -Olivier
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
>
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
- --
David L. Paterline Westinghouse Electric Company LLC
Advanced Technical Engineer Commercial Nuclear Fuel Division
dave.paterline@cnfd.pgh.wec.com Westinghouse Energy Center
PH: 412-374-2286 4350 Northern Pike Blvd.
FX: 412-374-2284 Monroeville PA 15146

- ------------- End Forwarded Message -------------

- --
David L. Paterline Westinghouse Electric Company LLC
Advanced Technical Engineer Commercial Nuclear Fuel Division
dave.paterline@cnfd.pgh.wec.com Westinghouse Energy Center
PH: 412-374-2286 4350 Northern Pike Blvd.
FX: 412-374-2284 Monroeville PA 15146

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 12:59:06 -0500 (EST)
From: "David L. Paterline" <zcdlp@cnfd.pgh.wec.com>
Subject: RE: [cciug] Retrieve branching points

If I understand correctly what you want, I think this will work (in my example,
the branch name is 'Y2K'):

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:51:06 -0500
From: "Kamdar, Tushar" <TKamdar@Engage.com>
Subject: [cciug] ClearCase support for Redhat Linux 6.1 ??

Hi All,

Does anyone know if Rational supports, RedHat Linux 6.1 for ClearCase OR
planning to support in it's future versions of ClearCase ?

Thanks,
Tushar
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:47:08 -0500
From: "Bickford, Fred" <bickford@Rational.Com>
Subject: RE: [cciug] stopping network operatins from command line

John,

You can change how the services react and make them manual for all and have
a login script depending on user start the service and map views and mount
vobs,
or make the user execute the script after login.....

Check out : http://clearcase.rational.com/faqs/technote_1429.html

as it has the registry keys for the 3 clearcase services...

Hope this helps...

Fred

Fred Bickford IV
Rational Software

"New opinions are always suspected, and usually opposed,
without any other reason but because they are not already common".
- -- John Locke

- -----Original Message-----
From: John_Falasco@hyperion.com [mailto:John_Falasco@hyperion.com
Sent: Wednesday, January 05, 2000 3:32 PM
To: cciug@Rational.Com
Subject: [cciug] stopping network operatins from command line

On NT, Clearcase supports Snapshot views so that users of portable machines
can
take their work home. When such a user boots up at home, Clearcase attempts
to
mount vobs, stat views, etc, unless one remembers to UnCheck the option
"Enable
Network Operations" in the Homebase Applet.

Almost everybody forgets to do this occasionally, and suffers the extra-long
bootup time as a consequence when they are disconnected from the lan. Is
there
any way to disable this option from the command line rather than within the
applet? That would allow the creation of an icon on the desktop to enable
and
disable network operations (in which case, the services would be set to run
as
Manual instead of Automatic).

Thanks,
- -John

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 16:07:44 -0500
From: John_Falasco@hyperion.com
Subject: RE: [cciug] stopping network operatins from command line

Thanks - yes, we can put two icons on the desktop: one that runs a batch file
that does "net stop albd, net stop cccredmgr, net stop lockmgr", the other that
runs it's counterpart. I just thought it might be cleaner to run something like
"net stop networkServices" or whatever. Putting it in the login script is a good
idea though - I hadn't thought of that.

Thanks,
John

Greg Dickie <greg@discreet.com> on 01/05/2000 03:42:15 PM

To: John Falasco@Hyperion
cc: cciug@Rational.Com
Subject: RE: [cciug] stopping network operatins from command line

Greg

- ---------------------------------------------------------------------
Greg Dickie
Just A Guy*
*from discreet (the logic is gone)
Montreal
(514) 954-7171
greg@discreet.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: 05 Jan 2000 16:12:42 -0500
From: "Paul D. Smith" <pausmith@nortelnetworks.com>
Subject: Re: [cciug] ClearCase support for Redhat Linux 6.1 ??

%% "Kamdar, Tushar" <TKamdar@Engage.com> writes:

  kt> Does anyone know if Rational supports, RedHat Linux 6.1 for
  kt> ClearCase OR planning to support in it's future versions of
  kt> ClearCase ?

ClearCase 4.0 will support RedHat Linux (6.0 officially, I think, but no
big deal) _partially_.

You will be able to make snapshot views on Linux, and you'll be able to
run VOB servers on Linux, and you can even run view servers on Linux,
but you won't able to set to dynamic views on Linux (no MVFS for
Linux--yet?).

ClearCase 4.0 for Solaris, HP-UX, and Windows NT is available now; CC
4.0 for other platforms (including Linux) will be available ~March I
think.

See the Rational web site for announcements.

HTH.

- --
- -------------------------------------------------------------------------------
Paul D. Smith <psmith@baynetworks.com> Network Management Development
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
- -------------------------------------------------------------------------------
These are my opinions---Nortel Networks takes no responsibility for them.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 13:17:23 -0800
From: "Riedel, Martina" <Martina.Riedel@icn.siemens.com>
Subject: RE: [cciug] FW: ClearCase Visual C++ problem...

You really shouldn't use Tools->Options / Directory. that is for
installation paths and every user will have to duplicate your settings in
their installation. For your project you want to put all those settings in
Project -> settings.
You want to use either relative settings (..\include) or \code\include,
which will work from any view.
Martina
Don't Postpone Joy - Have Fun
Martina Riedel Siemens Information and Communication
Networks
phone: 561-997-3774 Martina.Riedel@ICN.Siemens.com
<mailto:Martina.Riedel@ICN.Siemens.com>

                -----Original Message-----
                From: O'brien, Mark [mailto:marko@UU.NET
                Sent: Wednesday, January 05, 2000 3:05 PM
                To: 'cciug@rational.com'
                Cc: Swift, John
                Subject: [cciug] FW: ClearCase Visual C++ problem...

                Hello,
                        I am a developer who has recently introduced
ClearCase into my
                development
                environment and I'm having troubling setting my include
directories in
                DevStudio. I created a snapshot view, went into my view,
opened my project
                file which started Visual C++, then I went to the
"Tools->Options" menu and
                selected the "Directories" tab. From there, I am required to
set the search
                paths for the all include directories so Visual C++ can
compile my project.
                The problem is, Visual C++ insists on using the absolute
path which does not
                work in a team environment since the absolute path is
different for each
                team member using Visual C++ and ClearCase because ClearCase
requires every
                team member to have a unique view name, which equates to a
unique absolute
                path. Here's an example...

In the Tools->Options->Directories tab I place the path to
the include
directory where all of the public headers are located.

E:\snapshot\jswift_view\code\include

                At that point, I can compile everything fine. Then, a team
member of mine
                comes along and checks everything out and tries to build and
it fails
                because the compiler cannot find the header file. After Rob
examines the
                project's include directory settings he sees that the path
above is invalid
                for his and should be...

E:\snapshot\rob_view\code\include

                My question is how do we make both of these product work
harmoniously in a
                team environment? Simply adding every developers personal
view path in
                Visual C++ is NOT acceptable. Obviously we cannot all use a
standard view
                name such as...

E:\snapshot\project_view\code\include

...since only one of these can be created.

Anyway, I checked the documentation and I can't find
anything on this
problem. I suspect it's something simple that I'm just
missing.

Can someone please help?
js

                John Swift
                jswift@uu.net
                UUNET Technologies, an MCI WorldCom Company.

                 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - -
                 with
body
                 can also
unsubscribe
                 customers-only
website

http://clearcase.rational.com/cciug/mailing_list.html
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 16:01:27 -0500
From: Mark Keil <Mark.Keil@cportcorp.com>
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

Have you tried

- -----Original Message-----
From: Lezz Giles [mailto:lgiles@avici.com
Sent: Wednesday, January 05, 2000 1:38 PM
To: cciug@Rational.Com
Subject: [cciug] Any problems exporting views in CC 4.0?

I have a call in to Rational support on this, but just in case anybody
else has
a suggestion or has seen the same problem...

So does anybody else export views to Linux from CC 4.0? Is it working
fine or are
you seeing problems?

Thanks,
Lezz Giles
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:37:15 -0600 (CST)
From: The Great Vobadmin <vobadmin@amdocs.com>
Subject: Re: [cciug] ClearCase support for Redhat Linux 6.1 ??

Rumor has it that ClearCase ships for Red Hat in March 2000.

Jonathan

/////////////////////////////////////////////////////////////////
Jonathan Craft, Software Process Engineer (217) 239-8861
jcraft@itds.com (217) 351-2640 (fax)

A M D O C S

"Please remain calm... I may be mad, but I am a professional."
- -- Mad Software Process Engineer

/////////////////////////////////////////////////////////////////
>
>
> Hi All,
>
> Does anyone know if Rational supports, RedHat Linux 6.1 for ClearCase OR
> planning to support in it's future versions of ClearCase ?
>
> Thanks,
> Tushar
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 16:38:14 -0500
From: Mervin_Manangan@instinet.com
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

This is not be an answer since I haven't tried CC 4.0 yet (much as well export
views from it!) but I did had that similar problem (NFS deadlocks - hanging of
the system) when I did export a view to a Linux client from ClearCase 3.2.1.
We're running an HPUX10.20 VOB/view server and pretty much its working well
without the non-ClearCase access from the Linux box. But once I did mount the
exported view/VOB filesystem from the server to the Linux client, NFS and MVFS I
guess bugged down and the server was adversely affected.

Has anybody successfully did a non-ClearCase access setup from Linux to an
HPUX10.20 VOB/view server before in CC 3.2.1? How did it go?

Thanks,
Mervin

Mark Keil <Mark.Keil@cportcorp.com> on 01/05/2000 04:01:27 PM

To: Lezz Giles <lgiles@avici.com>, cciug@Rational.Com
cc: (bcc: Mervin Manangan/CTS/US/INSTINET)
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

Have you tried

- -----Original Message-----
From: Lezz Giles [mailto:lgiles@avici.com
Sent: Wednesday, January 05, 2000 1:38 PM
To: cciug@Rational.Com
Subject: [cciug] Any problems exporting views in CC 4.0?

I have a call in to Rational support on this, but just in case anybody
else has
a suggestion or has seen the same problem...

So does anybody else export views to Linux from CC 4.0? Is it working
fine or are
you seeing problems?

Thanks,
Lezz Giles
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 16:35:29 -0500
From: Mark Keil <Mark.Keil@cportcorp.com>
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

(sorry about the first incomplete post, somehow it got sent, sigh)

Have you tried exporting from a 3.2.1 view server?
Does the 4.0 vob server/3.2.1 view server combo work,
or does the fact that the vob server is at 4.0 affect
the mix? (I wouldn't expect it to, but who knows?)

(Since 4.0 requires the vob servers to be loaded first,
I assume that your vob server is at 4.0 and your view server
is at 4.0)

I wonder if rational tests the export view case in
all the combinations as part of release testing?
The supported archecture to Linux is fairly common
these days so I would hope that they test that combo.

- -Mark

- -----Original Message-----
From: Lezz Giles [mailto:lgiles@avici.com
Sent: Wednesday, January 05, 2000 1:38 PM
To: cciug@Rational.Com
Subject: [cciug] Any problems exporting views in CC 4.0?

I have a call in to Rational support on this, but just in case anybody
else has
a suggestion or has seen the same problem...

So does anybody else export views to Linux from CC 4.0? Is it working
fine or are
you seeing problems?

Thanks,
Lezz Giles
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 16:44:24 -0500
From: "Lungu, James" <James.Lungu@usa.xerox.com>
Subject: RE: [cciug] ClearCase support for Redhat Linux 6.1 ??

Check it out right here:

http://www.rational.com/sitewide/support/upgrades/cdrom/index.jtmpl

- -----Original Message-----
From: The Great Vobadmin [mailto:vobadmin@amdocs.com
Sent: Wednesday, January 05, 2000 4:37 PM
To: TKamdar@Engage.com
Cc: cciug@Rational.Com
Subject: Re: [cciug] ClearCase support for Redhat Linux 6.1 ??

Rumor has it that ClearCase ships for Red Hat in March 2000.

Jonathan

/////////////////////////////////////////////////////////////////
Jonathan Craft, Software Process Engineer (217) 239-8861
jcraft@itds.com (217) 351-2640 (fax)

A M D O C S

"Please remain calm... I may be mad, but I am a professional."
- -- Mad Software Process Engineer

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 16:51:36 -0500
From: David Boyce <dsb@world.std.com>
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

Just for the record, there are a couple of Rational tech notes on this
topic - http://clearcase.rational.com/faqs/technote_726.html and
http://clearcase.rational.com/faqs/technote_1694.html, maybe others.

- -David Boyce

At 04:38 PM 1/5/00 -0500, Mervin_Manangan@instinet.com wrote:

>This is not be an answer since I haven't tried CC 4.0 yet (much as well
>export
>views from it!) but I did had that similar problem (NFS deadlocks - hanging of
>the system) when I did export a view to a Linux client from ClearCase 3.2.1.
>We're running an HPUX10.20 VOB/view server and pretty much its working well
>without the non-ClearCase access from the Linux box. But once I did mount the
>exported view/VOB filesystem from the server to the Linux client, NFS and
>MVFS I
>guess bugged down and the server was adversely affected.
>
>Has anybody successfully did a non-ClearCase access setup from Linux to an
>HPUX10.20 VOB/view server before in CC 3.2.1? How did it go?
>
>Thanks,
>Mervin
>
>
>
>
>
>Mark Keil <Mark.Keil@cportcorp.com> on 01/05/2000 04:01:27 PM
>
>To: Lezz Giles <lgiles@avici.com>, cciug@Rational.Com
>cc: (bcc: Mervin Manangan/CTS/US/INSTINET)
>Subject: RE: [cciug] Any problems exporting views in CC 4.0?
>
>
>
>
>
>Have you tried
>
>-----Original Message-----
>From: Lezz Giles [mailto:lgiles@avici.com
>Sent: Wednesday, January 05, 2000 1:38 PM
>To: cciug@Rational.Com
>Subject: [cciug] Any problems exporting views in CC 4.0?
>
>
>I have a call in to Rational support on this, but just in case anybody
>else has
>a suggestion or has seen the same problem...
>
>We export a number of views to Linux boxes. This worked fine under CC
>3.2, but
>since we upgraded to 4.0 we've had problems seeing the mounted views on
>the
>Linux boxes - specifically, when we try to look inside the mounted
>filesystem,
>any access to the directories or files hangs. We can mount normal NFS
>filesystems
>with no problems, and we can mount the exported views on other Solaris
>boxes
>with no problems.
>
>So does anybody else export views to Linux from CC 4.0? Is it working
>fine or are
>you seeing problems?
>
>Thanks,
>Lezz Giles
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

>
>
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 13:57:48 -0800
From: Kenneth Creager <K.Creager@sd.us.am.ericsson.se>
Subject: Re: AW: [cciug] view-storage-pname

I have been able to get this to work when creating a view:

gpath: /net/host/path
hpath: /path
path: /net/host/path

Why the first and last paths are the same I don't know??

Ken Creager

At 1/5/00 01:12 PM +0100, EXTERN Chapman Roy (WA-Consultants; K5/ESQ1) wrote:
>
>Don't be ashamed, this one has foxed me for years as well. When passing the
>triplet (gpath, hpath and the path) I have found, from trying every
>combination known, that it will only accept the triplet is all three are the
>same. I therefore always pass all three as global pathnames (under Unix
>this would be /net/hostname/etc/...) and it invariably works. Now I am told
>that I should be able to pass a differnt global and local pathname
>(different as in the local pathname does not include /net/hostname) but I
>have tried. And as Homer Simpson once said, "Trying is the first step on
>the long road to failure".
>
>Roy Chapman
>External Consultant
>
>Robert Bosch GmbH * (00 49) 0711/811-31795 / GSM (00 49)
>0173/4887139
>K5/ESQ * (00 49) 0711/811-31800
>Postfach 20 02 40 * mailto:roy.chapman@de.bosch.com
>D-70442 Stuttgart
>
>
>> ----------
>> Von: Batlin, Alex[SMTP:alex.batlin@nomura.co.uk
>> Gesendet: Mittwoch, 5. Januar 2000 12:02
>> An: CCIUG (E-mail)
>> Betreff: [cciug] view-storage-pname
>>
>>
>> I am ashamed to be asking this question, however, it's something that is
>> bothering me. When registering or tagging a view or a VOB, when one
>> specifies the tripple path, one also specifies the view-storage-pname,
>> which
>> is usually the same as the host-stg-pname.
>>
>> Why is this the case and can the view-storage-pname ever be different to
>> either host-stg-pname or global-stg-pname.
>>
>> > mktag -view -tag view-tag [-tcomment tag-comment] [-replace]
>> > [-nstart] [-ncaexported] [-region network-region]
>> > [-host hostname -hpath host-stg-pname -gpath
>> global-stg-pname]
>> > view-storage-pname
>>
>> Regards,
>>
>> Alex.
>>
>>
>>
>> --------------------------------------------------------------------------
>> --
>> --------
>> Alex Batlin | alex.batlin@nomura.co.uk
>> Nomura International plc, Nomura House | Phone: +44 (0)20 7521 3728
>> 1 St. Martin's-le-Grand, London EC1A 4NP | Fax: +44 (0)20 7521 3686
>>
>> --------------------------------------------------------------------------
>> --
>> --------
>> The information contained in this message is intended for the named
>> recipients
>> only. If you are not an intended recipient of this message, you must not
>> copy,
>> distribute or take any action in reliance on it and you should notify the
>> sender
>> immediately. E-mail transmission cannot be guaranteed to be secure or
>> error-free
>> as information could be intercepted, corrupted, lost, destroyed, arrive
>> late or
>> incomplete, or contain viruses, and Nomura International plc excludes
>> liability for
>> (1) any errors or omissions in the contents of this message which arise
>> as
>> a result
>> of e-mail transmission and (2) the transmission of any viruses. If
>> verification is
>> required please request a hard-copy version. Nomura International plc is
>> regulated
>> by the Securities and Futures Authority Limited and is a member of the
>> London
>> Stock Exchange.
>>
>> --------------------------------------------------------------------------
>> --
>> --------
>> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>>
>>
>>
>> http://clearcase.rational.com/cciug/mailing_list.html
>>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 14:05:22 PST
From: "Howard Zhou" <ehz002@hotmail.com>
Subject: [cciug] Identify a Hard Link

Hi, CCIUGers

Is there any way to identify a hard link? "ct describe" doesn't seem to do
the job.

Howard
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:08:08 -0700
From: "WYLDER,JOE (HP-Roseville,ex1)" <joe_wylder@hp.com>
Subject: RE: [cciug] stopping network operatins from command line

Try something like this:

C:\>type ON.reg
REGEDIT4

[HKEY_CURRENT_USER\Software\Atria\ClearCase\CurrentVersion]
"Connected"=dword:00000001

C:\>type OFF.reg

REGEDIT4

[HKEY_CURRENT_USER\Software\Atria\ClearCase\CurrentVersion]
"Connected"=dword:00000000

C:\>regedit /s ON.reg

C:\>regedit /s OFF.reg

- -----Original Message-----
From: John_Falasco@hyperion.com [mailto:John_Falasco@hyperion.com
Sent: Wednesday, January 05, 2000 12:32 PM
To: cciug@Rational.Com
Subject: [cciug] stopping network operatins from command line

On NT, Clearcase supports Snapshot views so that users of portable machines
can
take their work home. When such a user boots up at home, Clearcase attempts
to
mount vobs, stat views, etc, unless one remembers to UnCheck the option
"Enable
Network Operations" in the Homebase Applet.

Almost everybody forgets to do this occasionally, and suffers the extra-long
bootup time as a consequence when they are disconnected from the lan. Is
there
any way to disable this option from the command line rather than within the
applet? That would allow the creation of an icon on the desktop to enable
and
disable network operations (in which case, the services would be set to run
as
Manual instead of Automatic).

Thanks,
- -John

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 23:13:08 +0100
From: =?iso-8859-1?Q?=D6lwing_Kenneth_-_keol?= <K.Olwing@abalon.se>
Subject: RE: [cciug] FW: ClearCase Visual C++ problem...

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

- ------_=_NextPart_001_01BF57CA.0C5741E0
Content-Type: text/plain;
charset="iso-8859-1"

Whoa, there. No need to share it, just do a 'subst X:
e:\snapshot\jswift_view', or whatever drive letter suits you. Always use
relative paths in project settings, not DevStudio global settings and you'll
be fine.

HTH,

ken1

> >
> >
> > E:\snapshot\jswift_view\code\include
>
> Go to the explorer and share E:\snapshot\jswift_view.
> Then mount it as a drive letter and make all your VC++ paths
> relative to the top of the view (\code\include). We do this
> so that the snapshot and dynamic views are interchangable.

- ------_=_NextPart_001_01BF57CA.0C5741E0
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Diso-8859-1">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>RE: [cciug] FW: ClearCase Visual C++ problem...</TITLE>
</HEAD>
<BODY>

Whoa, there. No need to share it, just do a 'subst X: =
e:\snapshot\jswift_view', or whatever drive letter suits you. Always =
use relative paths in project settings, not DevStudio global settings =
and you'll be fine.

HTH,

ken1

> > 
 > > 
 > =
>         =
E:\snapshot\jswift_view\code\include
 > 
 > Go to the explorer and share =
E:\snapshot\jswift_view.
 > Then mount it as a drive letter and make all =
your VC++ paths
 > relative to the top of the view =
(\code\include).  We do this
 > so that the snapshot and dynamic views are =
interchangable.

</BODY>
</HTML>
- ------_=_NextPart_001_01BF57CA.0C5741E0--
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 16:13:28 -0600
From: "Somer Pyron" <spyron@nortelnetworks.com>
Subject: [cciug] question

I'm trying to find out who moderates this group. I'm with the helpdesk,
and have had a lot of trouble getting in touch with whomever makes these
automatic posts. Please contact me at spyron@nortelnetworks.com as soon
as you can. Thanks.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 17:10:52 -0500
From: "Kamdar, Tushar" <TKamdar@Engage.com>
Subject: RE: [cciug] ClearCase support for Redhat Linux 6.1 ??

Thanks! all for your response.

Tushar

- -----Original Message-----
From: Lungu, James [mailto:James.Lungu@usa.xerox.com
Sent: Wednesday, January 05, 2000 4:44 PM
To: 'The Great Vobadmin'; TKamdar@Engage.com
Cc: cciug@Rational.Com
Subject: RE: [cciug] ClearCase support for Redhat Linux 6.1 ??

Check it out right here:

http://www.rational.com/sitewide/support/upgrades/cdrom/index.jtmpl

Rumor has it that ClearCase ships for Red Hat in March 2000.

Jonathan

/////////////////////////////////////////////////////////////////
Jonathan Craft, Software Process Engineer (217) 239-8861
jcraft@itds.com (217) 351-2640 (fax)

A M D O C S

"Please remain calm... I may be mad, but I am a professional."
- -- Mad Software Process Engineer

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 17:18:14 -0500
From: Suzanne Simpson <ssimpson@norcross.rms.slb.com>
Subject: [cciug] Relocate Command

All,

Has any one performed the relocate before where you lost all view private
files on an individual view. I performed a relocate and broke up one vob
into 3 additional vobs and lost all the view private files from one of the
developers view.

Any input would greatly be appreciated.

I spoke with Rational and they are currently researching the problem. They
indicated that it may be a bug in the relocate command. I checked my lost
and found from the individual view but nothing was found in the
lost/found. Any idea's!!!!!

Thanks,

Suzanne Simpson
Schlumberger

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 14:21:57 -0800
From: Robin LaSalle <rlasalle@wav1.wavtrace.com>
Subject: RE: [cciug] Relocate Command

It says very specifically in the Relocate man page that all view private
files and DOs are lost when the command is used.

Use recoverview -sync to move the files to the views lost+found directory.

Robin La Salle
Software Configuration Manager
WavTrace, Inc.
(425) 468-2244
rlasalle@wavtrace.com

> -----Original Message-----
> From: Suzanne Simpson [mailto:ssimpson@norcross.rms.slb.com
> Sent: Wednesday, January 05, 2000 2:18 PM
> To: cciug@Rational.Com
> Subject: [cciug] Relocate Command
>
>
>
> All,
>
> Has any one performed the relocate before where you lost all
> view private
> files on an individual view. I performed a relocate and
> broke up one vob
> into 3 additional vobs and lost all the view private files
> from one of the
> developers view.
>
> Any input would greatly be appreciated.
>
> I spoke with Rational and they are currently researching the
> problem. They
> indicated that it may be a bug in the relocate command. I
> checked my lost
> and found from the individual view but nothing was found in the
> lost/found. Any idea's!!!!!
>
> Thanks,
>
> Suzanne Simpson
> Schlumberger
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
> can also
> unsubscribe
>
> http://clearcase.rational.com/cciug/mailing_list.html
>
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 17:28:56 -0500
From: Ron Natalie <ron@sensor.com>
Subject: Re: [cciug] Identify a Hard Link

Howard Zhou wrote:
>
> Hi, CCIUGers
>
> Is there any way to identify a hard link? "ct describe" doesn't seem to do
> the job.
>

A directory entry in the VOB (and UNIX for that matter) is a pointer to
where the file essentials lives (in UNIX it's called an inode). When
you make a hard link, you have two entries that point toi the same file
instance. Neither one is preferred over the other. Once you use the
ln command to create the new directory entry, all the links to the file
are the peers.

Now with a plain UNIX file system, you can look at the link count (that's
the number displayed to between the permissions and the owner name. If
it is greater than 1, then more than one directory entry refers to this
file.

Pardon me if you knew all this. You are right, there doesn't seem to
be anything in the tools that lets you do something similar to determine
if a clearcase element is linked. The only thing that I can thin of
is to run /usr/atria/etc/mvfsstorage on the files you think may be linked
together. If the container comes up identifical, they are.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 17:32:12 -0500
From: Ron Natalie <ron@sensor.com>
Subject: Re: [cciug] FW: ClearCase Visual C++ problem...

> �lwing Kenneth - keol wrote:
>
> Whoa, there. No need to share it, just do a 'subst X: e:\snapshot\jswift_view',

Gosh you're right. Learn something new everyday (you can tell I'm a UNIX
hack, dragged kicking and screaming into micrsoft land). It looks like
this is how clearcase connects the dynamic views to drive letters as well
(they all show up in subst with no args). Learned two things today.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 17:37:13 -0500
From: John_Falasco@hyperion.com
Subject: RE: [cciug] stopping network operatins from command line

ooh, interesting. I'll give it a try - thanks!

"WYLDER,JOE (HP-Roseville,ex1)" <joe_wylder@hp.com> on 01/05/2000 05:08:08 PM

To: "'cciug@Rational.com'" <cciug@Rational.com>
cc: John Falasco@Hyperion
Subject: RE: [cciug] stopping network operatins from command line

Try something like this:

C:\>type ON.reg
REGEDIT4

[HKEY_CURRENT_USER\Software\Atria\ClearCase\CurrentVersion]
"Connected"=dword:00000001

C:\>type OFF.reg

REGEDIT4

[HKEY_CURRENT_USER\Software\Atria\ClearCase\CurrentVersion]
"Connected"=dword:00000000

C:\>regedit /s ON.reg

C:\>regedit /s OFF.reg

On NT, Clearcase supports Snapshot views so that users of portable machines
can
take their work home. When such a user boots up at home, Clearcase attempts
to
mount vobs, stat views, etc, unless one remembers to UnCheck the option
"Enable
Network Operations" in the Homebase Applet.

Almost everybody forgets to do this occasionally, and suffers the extra-long
bootup time as a consequence when they are disconnected from the lan. Is
there
any way to disable this option from the command line rather than within the
applet? That would allow the creation of an icon on the desktop to enable
and
disable network operations (in which case, the services would be set to run
as
Manual instead of Automatic).

Thanks,
- -John

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 16:52:03 -0600
From: Jalli Naga-WLNJ02 <Naga_Jalli-WLNJ02@email.mot.com>
Subject: [cciug] Running CCase on Solaris7

I want to install Clearcase 3.2.1 on Solaris 2.7 in Sun UE-3500 system.
Does any of you are running ClearCase on Solaris 2.7. If you are, Can you
please
let me know if I have to install specific clearcase patches.

Thank You all,

Naga Jalli
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 14:53:30 -0800
From: Kathleen Johnson <kathleen@genesyslab.com>
Subject: [cciug] Great job available in San Francisco for CM Manager

We are a great publicly held company located in downtown San Francisco. We
build state of the art muti-platform commercial software that enables
multi-media connectivity. We currently have an opening for the Manager of
Software Support Systems. This position will be heavily involved in
activities related to our use of Clear Case. In addition, the position is
responsible for providing leadership, vision and direction to our
multi-platform installation activities and for automating and modernizing
our build and packaging activities.

The Manager of Software Support Systems is responsible to insure that the
product is efficiently built, installed, packaged, and distributed to
internal and external users, in an error free manner. This position
directly supervisors and leads the individuals who perform product builds
and releases on Windows NT and UNIX systems.

The Manager of Software Support Systems works with other members of the
software development team including design, QA and product management to
understand the requirements for the building and packaging of products as
well as the installation process.

Ideal candidates will have the following experience:

- - ClearCase administration
- - Strong UNIX/NT
- - 1+ years experience in release engineering
- - Experience with Install Shield
- - Knowledge of Java installation tools
- - Excellent organization, communication, and interpersonal skills

This is a great position for the Clear Case Administrator who is looking to
branch out into other operational areas. If interested, please forward your
resume to me.

Kathleen Johnson
Director, Engineering Operations
Genesys Telecommunications, Inc.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 14:54:04 -0800
From: murali@epiphany.com
Subject: RE: [cciug] Running CCase on Solaris7

YEs, We are running CC 3.2.1 on Solaris 2.7.

You need the following patches!

3.2.1-18 & 20

Hope this helps!

murali

- -----Original Message-----
From: Jalli Naga-WLNJ02 [mailto:Naga_Jalli-WLNJ02@email.mot.com
Sent: Wednesday, January 05, 2000 2:52 PM
To: 'cciug@rational.com'
Subject: [cciug] Running CCase on Solaris7

Thank You all,

Naga Jalli
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 17:56:03 -0500
From: "Darby, David" <David.Darby@fmr.com>
Subject: RE: [cciug] Identify a Hard Link

Ron/Howard,

A little trick I've used in the past is to do a "ct mklabel -recurse" from
the vob mount point. Any elements that are hard linked give an error
message that the label already exists on an element when the label is
applied the second time.

- -David

- -----Original Message-----
From: Ron Natalie [mailto:ron@sensor.com
Sent: Wednesday, January 05, 2000 5:29 PM
To: Howard Zhou
Cc: cciug@Rational.Com
Subject: Re: [cciug] Identify a Hard Link

Howard Zhou wrote:
>
> Hi, CCIUGers
>
> Is there any way to identify a hard link? "ct describe" doesn't seem to do
> the job.
>

------------------------------

Date: Wed, 5 Jan 2000 18:06:28 -0500
From: "Kristy Gleason" <kgleason@Rational.Com>
Subject: Re: [cciug] Running CCase on Solaris7

Hi Naga,

Yes, you will need to install the latest mandatory patches that are
available from our customer-only web site

http://clearcase.rational.com/techsupport/clearcase/tech/patchlink/new/htmls
treams/clearcase_p3.2.1-sun5.html

Please note that the url line-wraps, and ends in -sun5.html.

You will be prompted for a username and password -- use your license host
host id for the username your license string as the password, then download
the 2 mandatory patches that you will see listed on that page.

- -- Kristy

- ----- Original Message -----
From: Jalli Naga-WLNJ02 <Naga_Jalli-WLNJ02@email.mot.com>
To: 'cciug@rational.com' <cciug@Rational.Com>
Sent: Wednesday, January 05, 2000 5:52 PM
Subject: [cciug] Running CCase on Solaris7

>
>
> I want to install Clearcase 3.2.1 on Solaris 2.7 in Sun UE-3500 system.
> Does any of you are running ClearCase on Solaris 2.7. If you are, Can you
> please
> let me know if I have to install specific clearcase patches.
>
> Thank You all,
>
> Naga Jalli
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Thu, 6 Jan 2000 01:08:25 +0200
From: Barr Ronald <Ronald.Barr@kone.com>
Subject: RE: [cciug] ClearCase 4.0

Rational Shipped it to us without prompting. It may be a scheduling issue.
Also, is your copy of ClearCase on a current maintenance agreement? They
will probably only ship copies to current maintenance accounts.

Ron

- ---------------------------------------------------------------
Ronald Barr Voice: (309) 743-5687
KONE Inc. Fax: (309) 743-5467
One KONE Court E-Mail: ronald.barr@kone.com
<mailto:ronald.barr@kone.com>
Moline, IL USA 61265
- ---------------------------------------------------------------

- -----Original Message-----
From: Ford David [mailto:David.Ford@Tms-Ltd.com
Sent: Wednesday, January 05, 2000 7:24 AM
To: Cciug (E-mail)
Subject: [cciug] ClearCase 4.0

Hello CCIUG,

When can we mere mortals get hold of a copy of ClearCase 4.0? Apparently we
have to request it rather than Rational automatically sending it? Is this
true.

David Ford
David Ford
Software Configuration Manager
Thomson Marconi Sonar Ltd

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 15:34:53 -0800
From: murali@epiphany.com
Subject: [cciug] Clearcase 3.2.1 for Windows 2K

Have any one is using Clearcase 3.2.1 for Windows 2K as a snap shot views?

Thanks
murali

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 17:45:41 -0600
From: "Somer Pyron" <spyron@nortelnetworks.com>
Subject: Re: [cciug] question

Thanks for all the info, I think I found what I was looking for. Sorry to
bother!

"Pyron, Somer S.L. [EXCHANGE:RICH1:9485]" wrote:

> I'm trying to find out who moderates this group. I'm with the helpdesk,
> and have had a lot of trouble getting in touch with whomever makes these
> automatic posts. Please contact me at spyron@nortelnetworks.com as soon
> as you can. Thanks.
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 17:46:06 -0600
From: "Somer Pyron" <spyron@nortelnetworks.com>
Subject: Re: [cciug] question

Thanks for all the info, I think I found what I was looking for. Sorry to
bother!

"Pyron, Somer S.L. [EXCHANGE:RICH1:9485]" wrote:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 05 Jan 2000 22:14:11 -0500
From: Nathaniel Mishkin <mishkin@Rational.Com>
Subject: RE: [cciug] view-storage-pname

At 11:32 AM 1/5/2000 , Darby, David wrote:
>I questioned the need for this final argument myself for a long time.
>However, I discovered that it is possible to create a view while not
>actually being on the intended view server host. So, the view-storage-pname
>argument for the mkview command is the same as the -hpath argument when you
>are creating a view local to the machine you are logged into. And, the
>view-storage-pname argument for the mkview command is the same as the -gpath
>argument when you are creating a view that is not on the machine you are
>logged into.

This (and your example which I elided) get it just about right. The
general answer is that the "triple" (host/hpath/gpath), if specified,
determines what's going to end up in the ClearCase view object and tag
registries for the view. The view-storage-pname argument specifies what
pathname should be used by the mkview command to actually create the
directory (i.e., it's what's passed to the mkdir() system call).

The reason for all this generality is that some sites have machines
configured in ways intended to achieve higher performance. The canonical
example (which motivated the design) is a multi-homed server machine
attached to two subnets where each subnet has its own set of client
systems. The two IP addresses for the server have distinct associated host
names (e.g., H1 and H2). Clients on each subnet want to get packets to the
server most efficiently--without going through a router--and hence want to
use the server's IP address that's on the same subnet as the client.

Stretching things a bit further for simplicity of this example (although
more complicated things come up in real life), machines on one of the two
subnets might be configured so that only one of the host names can be
resolved (looked up). The upshot is that to make a view from the other
subnet that's intended to be accessed from the first subnet, you'd have to do:

     mkview -tag myview
            -hpath /usr/views/myview
            -gpath /net/H1/usr/views/myviews
            /net/H2/usr/views/myviews

Notice how the -gpath pname arg and the view-storage-pname are
different. (In real life, what tends to happen is that you make two
regions, one for each subnet and do "mkview" followed by "mktag".)

As to why the view-storage-pname arg isn't defaulted to the -gpath pname
arg...I guess that's because we think of the view-storage-pname arg as
being the "real" argument and the -gpath arg as being optional. (The
triple as a whole is derived from the view-storage-pname arg, if the triple
is omitted.)
- --
Nat Mishkin
Rational Software

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 22:17:31 -0500 (EST)
From: Danny Zerkel-D26 <dzerkel@cas.org>
Subject: Re: [cciug] SunWSPro C++ 5.0 templates

Scott,

The ClearCase 4.0 docs do not cover Sun C++ 5.0.

Here is what I came up with:

1) To compile C++ code that uses templates:

%.o: %.c++
 -@rm -f $@
 @rm -rf $@_cache; mkdir $@_cache
 $(C++C) $(C++FLAGS) -c $(EXTRAFLAGS) $< -o $@_cache/$@
 @mv -f $@_cache/$@ $@
 @test -f $@_cache/SunWS_cache/CC_state || rm -fr $@_cache
 @rm -f $@_cache/SunWS_cache/@_WARNING

This is necessary because the Sun C++ 5.0 -ptr option doesn't work.
The rm's keep the sibling DOs to a minimum.

2) To link C++ code with templates:

$(APP): $(APP_DEPS) $(OBJS)
        -@rm -f $@
        @rm -rf SunWS_cache
        /vobs/tools/bin/mk_CC_state $(wildcard *.o_cache)
        $(LINK++) $(LDFLAGS) -o $@ $(OBJS) $(LIB_PATHS)
        @rm -rf SunWS_cache

The mk_CC_state program is a Perl script which combines the CC_states
of all the seperate objects into one CC_state file in the standard
SunWS_cache location, plus link (ln) the specified template .o files
from the seperate object directories to the standard one. In the case
of duplicate template .o files, I use the most recent modification date.
After the linking loader completes, the SunWS_cache directory is removed
to clean up the sibling DO of the temporary CC_state file.

3) I use:

.SIBLINGS_AFFECT_REUSE:

This seems to generate a very clean config rec, with no unnecessary
rebuilds and wink-ins work correctly.

Note, the $(wildcard *.o_cache) thingy requires GNU compatibility.

Danny

> Date: Wed, 5 Jan 2000 12:30:28 -0500 (EST)
> From: VOB administrator <vobadm.to@fp.cibc.com>
> Subject: [cciug] SunWSPro C++ 5.0 templates
>
> CCIUGers;
>
> Welcome to the year 19100. :-)
>
> We are trying to move our software from Sun C++ 4.2 to Sun C++ 5.0, which
> is more ANSI standard C++. We make some use of templates, and Sun has
> changed their template model and directories completely with this release.
>
> In the 3.2.1 Clearcase manual it talks a little about the Templates.DB/
> subdirectory that was used by the C++ 4.2 compiler, and discusses whether
> or not to use .SIBLINGS_IGNORED_FOR_REUSE and other issues.
>
> What about 5.0? Does the Clearcase 4.0 docs reference their new
> SunWS_cache/ cache model?
>
> Or has anyone else made this work with build avoidance / winkins etc.
>
> Thanks!
> Scott Gregory as vobadm
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- ---------------------------------------------------------------
Danny J. Zerkel - dzerkel@cas.org | Chemical Abstracts Service
Compuware Professional Services | Columbus, OH
- ---------------------------------------------------------------
"It is the explicit, legal opinion of my various employers that
nothing I say in this message should be construed as my various
employers' opinions unless explicitly stated as such."

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Thu, 6 Jan 2000 09:28:12 +0530
From: asaha@hss.hns.com
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

We were having similar problems, only our Clearcase vob and views (3.2.1) was
from a
Solaris 2.6 E3500 box. We switched to using samba and things are working better
now,
but there are still occasional problems.

Regards,
Abheek

Mervin_Manangan@instinet.com on 01/06/2000 03:08:14 AM

To: Mark Keil <Mark.Keil@cportcorp.com>
cc: Lezz Giles <lgiles@avici.com>, cciug@Rational.Com (bcc: Abheek Saha/HSS)

Subject: RE: [cciug] Any problems exporting views in CC 4.0?

Has anybody successfully did a non-ClearCase access setup from Linux to an
HPUX10.20 VOB/view server before in CC 3.2.1? How did it go?

Thanks,
Mervin

Mark Keil <Mark.Keil@cportcorp.com> on 01/05/2000 04:01:27 PM

To: Lezz Giles <lgiles@avici.com>, cciug@Rational.Com
cc: (bcc: Mervin Manangan/CTS/US/INSTINET)
Subject: RE: [cciug] Any problems exporting views in CC 4.0?

Have you tried

- -----Original Message-----
From: Lezz Giles [mailto:lgiles@avici.com
Sent: Wednesday, January 05, 2000 1:38 PM
To: cciug@Rational.Com
Subject: [cciug] Any problems exporting views in CC 4.0?

I have a call in to Rational support on this, but just in case anybody
else has
a suggestion or has seen the same problem...

So does anybody else export views to Linux from CC 4.0? Is it working
fine or are
you seeing problems?

Thanks,
Lezz Giles
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

Date: Wed, 5 Jan 2000 21:59:34 -0800 (PST)
From: David Highley <dhighley@highley-recommended.com>
Subject: Re: [cciug] Any problems exporting views in CC 4.0?

We have exported ClearCase VOB and views from a Solaris 2.6 system and
NFS mounted them on a RedHat 6.1 Linux client. We have been doing
builds and all is working. We even came up with a one line ksh alias
for the Linux side that does an rsh to the server to execute ClearCase
commands so that the developers have pretty much transparent access to
ClearCase.

Have not tried this with an HPUX box yet. The reverse is not good. We
have tried to export a file system from Linux and mount it on a Solaris
system. The mount completes, but we are not able to write to the
mounted file system. According to the Linux news threads, NFS serving is
broken for 2.x kernels, but should be fixed in kernel 2.2.14.

>
>
>
>
> We were having similar problems, only our Clearcase vob and views (3.2.1) was
> from a
> Solaris 2.6 E3500 box. We switched to using samba and things are working better
> now,
> but there are still occasional problems.
>
> Regards,
> Abheek
>
>
>
>
>
> Mervin_Manangan@instinet.com on 01/06/2000 03:08:14 AM
>
> To: Mark Keil <Mark.Keil@cportcorp.com>
> cc: Lezz Giles <lgiles@avici.com>, cciug@Rational.Com (bcc: Abheek Saha/HSS)
>
> Subject: RE: [cciug] Any problems exporting views in CC 4.0?
>
>
>
>
>
>
> This is not be an answer since I haven't tried CC 4.0 yet (much as well export
> views from it!) but I did had that similar problem (NFS deadlocks - hanging of
> the system) when I did export a view to a Linux client from ClearCase 3.2.1.
> We're running an HPUX10.20 VOB/view server and pretty much its working well
> without the non-ClearCase access from the Linux box. But once I did mount the
> exported view/VOB filesystem from the server to the Linux client, NFS and MVFS I
> guess bugged down and the server was adversely affected.
>
> Has anybody successfully did a non-ClearCase access setup from Linux to an
> HPUX10.20 VOB/view server before in CC 3.2.1? How did it go?
>
> Thanks,
> Mervin
>
>
>
>
>
> Mark Keil <Mark.Keil@cportcorp.com> on 01/05/2000 04:01:27 PM
>
> To: Lezz Giles <lgiles@avici.com>, cciug@Rational.Com
> cc: (bcc: Mervin Manangan/CTS/US/INSTINET)
> Subject: RE: [cciug] Any problems exporting views in CC 4.0?
>
>
>
>
>
> Have you tried
>
> -----Original Message-----
> From: Lezz Giles [mailto:lgiles@avici.com
> Sent: Wednesday, January 05, 2000 1:38 PM
> To: cciug@Rational.Com
> Subject: [cciug] Any problems exporting views in CC 4.0?
>
>
> I have a call in to Rational support on this, but just in case anybody
> else has
> a suggestion or has seen the same problem...
>
> We export a number of views to Linux boxes. This worked fine under CC
> 3.2, but
> since we upgraded to 4.0 we've had problems seeing the mounted views on
> the
> Linux boxes - specifically, when we try to look inside the mounted
> filesystem,
> any access to the directories or files hangs. We can mount normal NFS
> filesystems
> with no problems, and we can mount the exported views on other Solaris
> boxes
> with no problems.
>
> So does anybody else export views to Linux from CC 4.0? Is it working
> fine or are
> you seeing problems?
>
> Thanks,
> Lezz Giles
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
>
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
>
>
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>

- --

Regards,

David Highley
Highley Recommended, Inc.
2927 SW 339th Street
Federal Way, WA 98023-7732

Phone: (206) 669-0081
FAX: (253) 838-8509
Email: dhighley@highley-recommended.com
WEB: http://www.highley-recommended.com
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

------------------------------

End of cciug-digest V2 #723
***************************

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Next message: jay@sony.co.in: "[cciug] Date problem"
Previous message: Paul D. Smith: "Re: [cciug] Any problems exporting views in CC 4.0?"
Next in thread: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Maybe reply: POSTMASTER: "[cciug] Unsent Message Returned to Sender"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

This archive was generated by hypermail 2b29 : Sun May 06 2001 - 00:22:19 EDT