Conference azur::mcc

    sorry to have to ask but...
     
    Which kit, v1.1? or v1.0?

    A nit but the release notes for the X1.1 kit (17-Sep-90 announcement)
    are thicker when printed than the user manual. Why is everything double
    spaced? Sigh... (the other user's of the printer were not amused)
    
    s/rob
    

    In section 1.4 of the release notes, the example command to VMSINSTAL
    does not contain the correct product name spec.
    
    @ sys$update:vmsinstal decmcc-bms2011 <location> options N
    
    should read
    
    @ sys$update:vmsinstal mccbmsx011 <location> options N
    
    s/rob
    

RE: .1

V1.1. Actually, I thought that X1.1.0 encoded that, but that is just a
      baselevel? I'll change the basenote title to include version.

Another nit: The startup files report the version as X1.0.1

    The release notes claim that (in two places no less) that this kit does
    not check for the correct amount of disk space.
    
    However, when I try to install it with MCC_COMMON on a DFS disk the
    install fails because it does not have sufficient space (cause DFS
    disks don't return that info).
    
    Sigh...
    
    s/rob

Section 1.3 states: "Ensure that the DECmcc V1.0 Basic Management System (BMS)
External Field Test Update software is installed on your system."

That is the UT1.0.1 kit, right? Which of the DIR, BMS, TK kits need to
be installed before the V1.1 X1.1.0 kit?

The DECmcc V1.1 Use manual pointed to by Dave Moore's 17-Sep memo is the
same as that pointed to by Peter Savage's 19-Jul memo. The date on the
manual is June 30.

The READ_ME_FIRST from the 17-Sep kit references the "Recording Entity
Attribute Data" chapter of the DECmcc Use guide.

No such chapter in that document.

Frankly, I must be missing something about the way you use baselevel
numbers. Both of the memos referenced above contain X1.1.0 in the title.
Shouldn't the base level have changed between the two kits?

    Petes message points to the version 1.0 releases of the manual and
    not the 1.1 doc set in the works (which Dave is referencing). Maybe
    Steve Jong can supply pointers to the 1.1 doc set.....
    
    

    No, you've got all that exists for V1.1 at this moment.

Hi Steve,

	Is there doc for the Historian,Exporter, and the Reports. thanks

Thien Doan
EIS/Network Management Solutions

    
    I reviewed DECmcc-use manual two weeks back containing chapters
    pertaining to Historian/Export/Reports etc.  But I don't find them in
    the DECmcc-use manual copied from KEEL yesterday.
    
    How do you expect us to test these products without the proper
    documentation?  
    
    Will new documentation (DECmcc-Use) manual be added into KEEL and the
    old one be deleted?
    
    Thanks..
    
    kandha

    You have indeed got the June 30 draft of _DECmcc Use_, not the most
    recent one.
    
    We will get the most recent version into WORDY:: and KEEL:: as soon as
    possible.
    
    I'm sorry for any inconvenience this is causing you.

It looks like .6 got lost. To reprise:

 On a system that has never had MCC on it, which V1.0 kits (BIR, BMS, TK)
have to be installed prior to installing the V1.1 17-Sep-90 X1.1.0 kit?

    Well, it's wedged. sigh...
    
    I tried to do something that I guess there isn't currently support for.
    Oh well.
    
    I had a domain open in the map window. All was fine. I was working on
    a command proceedure to feed the FCL to load up entities and domains
    and save a lot of point click type point click... you get the idea.
    
    So, off in another decterm window I updated the comand file and then
    started up an FCL. I ran the commands through it ok.
    When it completed I went back to the MAP window and wanted to get it
    to update with all the new stuff I just put in to the domain.
    
    There is an OPEN DOMAIN option and a CLOSE WINDOW option (but that
    exits MCC altogether). I needed a CLOSE DOMAIN option.
    
    Well, I selected OPEN DOMAIN and gave it the same name as the current
    one and I got a POP UP
       "Map Window Message: A PM bug has been unearthed."
    
    It is in front of all other windows and clicking on the "Acknowledged"
    button makes it go away for a second and then I get another one....
    
    In the command window that I started MCC from there was an ACCESS
    VIOLATION displayed. The error was;
    
    %SYSTEM-ACCVIO, access violation, reason mask=00, virtual
    address=0059D800, PC=0
    0389DB8, PSL=03C00000
    
    I can't get rid of it so I will abort MCC and reload it.
    
    s/rob

    Try it now:  WORDY::USER$657:[PUBLIC.NMS]DECMCC_USE_V11_IFT2.PS
    (5966 blocks).

RE: 14

Rob, the trick I use to close the last domain without exiting DECmcc is
to select Create Window first. That creates a generic window. Then you
can close your domain and it won't be the last window.

Pull down File from the generic window and re-open your domain.

Re .14, .16

Richard's way will work.  Rob's way *SHOULD* work too.  Quarable.

fred

    Re .2                                                   
    
    If one uses parameters like /NOFEED and /PASSALL it looks somewhat
    better on print.
    
    TCP/IP.
    
    The license for TCP/IP is unloaded in MCC$PAKS.COM, but the file
    does not included a new one ?
    
    After installation TCP/IP functionality seems to be gone. Do I have to
    make a new complete TCP/IP installation ?
    
    "BMS2".
    
    Are all the previous used "MCCxxxxBMS2" command files with this
    installation obsolete ? (They survived the installation)
    
    Thanks.
    
    S�ren H Christiansen

    RE .6 and .13
    
    You should be able to load a V1.1 IFT onto a system with NO prior
    MCC kit. Simply install the functions you want. You do NOT have
    to install the DIR kit before the BMS kit, DIR is a subset of the
    files included in the BMS kit.
    

    I tried it again (actually something similar cause I have to back out a
    lot to do exactly the same thing) and could not reproduce it. Sorry.
    
    I do believe though that there should be a choice to CLOSE a DOMAIN
    and return you to an empty map window. Creating a new window to do it
    is an interesting and useful workaround but a hack.
    
    s/rob

    I have a command proceedure that registers node4 and bridges, creates
    domains, and domain members.
    
    Today I added some registrations and the associated create domain
    member directives and re-ran the proceedure from the FCL in line mode.
    
    For all bridge and domain entities that were previously registered I
    recieved the correct results, the "Entity Exists"  message.
    
    However, for both the domain members that previously existed and the
    new ones I was adding I got the "Create Successful" message.
    
    I would have expected some other response for the previously existing
    domain members.
    
    s/rob

    I have created a bunch of nicely nested domains via command proceedures
    and it seems to work fine.
    
    However (I know, why is there always a however...) when I populate a
    domain with members in this way it gets to pick the icon to use.
    
    Is there a way for me to change which icon is used for an existing
    member? I was using domains for demprs, delnis and terminal servers
    at the moment and would like to change them to use the more appropriate
    icons.
    
    Also, is there a way to specify the icon to use in the create domain
    member command from the FCL?
    
    thanks
    s/rob

	re .18
>    The license for TCP/IP is unloaded in MCC$PAKS.COM, but the file
>    does not included a new one ?
 
	The PAK for TCP/IP is still @ CSEVEN::cseven$dua1:[daniele.tcpip_eft].
	It works with the new BMS.
  
>    After installation TCP/IP functionality seems to be gone. Do I have to
>    make a new complete TCP/IP installation ?
 
	Yes.  Installing a new BMS or Director leaves a new dictionary, parse
	table, etc.  You need to reinstall ALL asynchronous modules, like
	TCP/IP, or the DECnet Diagnostic FM.

	The SSB kit for TCP/IP and Diagnosis FM will allow you to load the
	dictionary without rebuilding the parse table, so you only have to
	do it once.  But it's still pretty grim.  Hopefully the mcc system
	files and installation will change to accomodate what's installed over
	the director, because this will become untenable soon.

	My, there's alot going on in this note.

    RE: .21
    
    This problem turns out to be a DNS problem (actually an architecture
    problem) and not an MCC problem. Sorry for the false alarm.
    
    s/rob
    

re: .23,

>	My, there's alot going on in this note.

	Good point!

	Is there any particular reason why we are trying to cram all
    discussions about V1.1 bugs into replies to this one basenote?
    Seems to me that it would be much easier to sort through all of these
    issues if there was a seperate note for each bug/feature reported.
    It might also make sense to consider creating a new and separate
    notesfile for each major field test of the product (i.e. keep this
    notesfile around for generic DECmcc disscussions, and then create two
    additional notesfiles (one for the V1.0 SSB, and one for V1.1 FT, etc.)

	Furthermore, I propose we establish a semi-formal note naming
    convention (such as the one used successfully in the ::JOBS notesfile)
    for the note titles of the bugs which are reported, so as to help bring
    the users of this product bring the appropriate attention to the problems
    as they come up.  This note naming convention might follow the format
    of "BUG:<DECmcc version><description>".

	For example: "BUG:V1.1-0 Chokes on Mumblefratz Command"

	Given the relatively enormous size and scope of the DECmcc product,
    coupled with the relatively enormous size of the DECmcc development
    group and associated user community, it would make sense to take
    full advantage of the capabities of VAXnotes to help in the testing
    process.  After all, there are many even smaller software products
    which successfully use notesfiles for testing in a similar fashion.

				    -davo

p.s.    Can you imagine all bug reports for the latest version of VAX/VMS
    	having to be entered as replies to a single note?

    ref .21, .24
    
    
    Yes, it is DECdns problem. However, MCC is planning to intercept those
    messgaes, and give  messages that are expected.
    
    That is when you are adding a member which is already a member, you will
    get a message indicating this. Similarly, when you are deleteing a
    member which is not a member, the message  should indicate this status too.
    
    To my knowledge this should be  in EFT.
    
    
    /jay

>    I have created a bunch of nicely nested domains via command proceedures
>    and it seems to work fine.
>    
>    However (I know, why is there always a however...) when I populate a
>    domain with members in this way it gets to pick the icon to use.
>    
>    Is there a way for me to change which icon is used for an existing
>    member? I was using domains for demprs, delnis and terminal servers
>    at the moment and would like to change them to use the more appropriate
>    icons.

	Currently the only way is to edit the map file, find the entity,
	and change icon file pointed to the new file name.

>    
>    Also, is there a way to specify the icon to use in the create domain
>    member command from the FCL?
    
	No, this is because the FCL and DOMAIN FM do not know anything 
	about the Iconic Map PM, hence they don't know anything about 
	the icons used.    

    RE .25
    
    The introduction note does give a nice naming convention for
    all notes, and it was used for a while, but has since gone the
    way of the wind. I believe that it would be useful to see it
    reinstated, but to do that the users of this conference have to
    be willing.
    

I installed V1.1 onto a system that had 1.01 already running on it.  The 
installation appeared to go ok until the BMS IVP.  The BMS IVP blows up on the 
following command      
 "MANAGE/ENTERPRISE TEST MCC 0 ALARMS, TO FILE = MCCBMS$OUT.TXT".  There are no
error messages except to say that the Installation has failed.
If I do a set verify amd rerun the IVP the following error is displayed by MCC
%MCC-E-NOTFOUND,  unsupported combination of verb, entity, partition
I haven't investigated any further!

ALso when I try to run the DECwindows interface I get the following

%LIB-E-ACTIMAGE, error activating image $177$DUA0:[SYS0.SYSCOMMON.][SYSLIB]MCC_D
ECW_PM.EXE;
-RMS-E-FNF, file not found
%MCC-E-FINDUSYM, failure finding universal symbol !AS in shareable image !AS

Could it be that the verson 1.01 has something to do with any of this?

Thanks
Cindy

Ps - I have verified that I copied over the correct savesets per 
announcement in this conference

RE: 28

Since I started this note I should be the one to say: 

	Please don't add any replies to this note announcing new problems
        or asking new questions. Start a new note for each (after looking
        to see if there is one already, of course).

        Please follow the title naming conventions described in note 1.0.
        I have started doing so, and will continue.

    RE:.29
    
    It appears as though ALARMS did not enroll OR you have a symbol which
    is being substituted.  You should try the commands:
    
    $ MANAGE/ENTERPRISE DISPLAY
    
    and
    
    $ MANAGE/ENTERPRISE TEST MCC 0 ALARMS
    
    If the first command shows any symbols defined, it probably has
    something to do with the problem.  If it does not, then the second
    command should succeed.  If the second command fails, and the first
    does not show any symbols, then Alarms did not enroll for some unknown
    reason. We need to know if it is the latter situation, since I have
    never seen a Management Module fail to enroll its dispatch table
    entries during installation.  Also, make sure that the logical
    mcc_alarms_fm is not defined.  BTW, the file MCCBMS$OUT.TXT is not
    supposed to be deleted, and should contain the errors which caused
    the IVP to fail for alarms.
    
    -Matt.
    (p.s., if both commands do not show anything unusual, then the problem
    seems to have gone away.  Probably re-installing will work.)

Thanks for the suggestions.  Here are the results.  "pears to me I still have
a problem

Xlr8<Privs>> manage/enterprise disp
DECmcc (X1.1.0)

                         Default Entity =
                      Default Qualifier =
                                 Prompt = MCC>
                            Indentation = 42

No symbols exist in the symbol table.

Xlr8<Privs>>  MANAGE/ENTERPRISE TEST MCC 0 ALARMS
DECmcc (X1.1.0)

%MCC-E-NOTFOUND,  unsupported combination of verb, entity, partition
Xlr8<Privs>> sho log mcc_alarms_fm
%SHOW-S-NOTRAN, no translation for logical name MCC_ALARMS_FM

When I couldn't find MCCBMS$OUT.TXT where I thought it would be I did a
Xlr8<Privs>> dir sys$sysdevice:[*...]MCCBMS$OUT.TXT
%DIRECT-W-NOFILES, no files found

Any more ideas.  I'll reinstall for starters

Thanks
Cindy

    re:.32
    
    Okay, so now we know that the mcc_alarms_fm.exe did not get installed
    with your X1.1 kit.  Now we just have to find out why.  I looked at
    the kit, it is definitely in the MCCBMSX011.C saveset.  You didn't
    happen to see any error messages like "%VMI-E-FILENOTFND, Could not find
    file MCC_ALARMS_FM.EXE", or something like that during installation, did
    you?
    
    -Matt.

  Re: .29

  The way to invoke the Iconic Map is:

	MANAGE/ENTERPRISE/INTERFACE=DECWINDOWS

  This will point to the correct image.  The DECMCC USE book is based on an
  early baselevel last spring and is incorrectly specifying /PRESENTATION=.
  The /PRESENTATION= is still valid, but requires the name of the image.  Using
  /INTERFACE=DECWINDOWS will avoid that problem.  In the May 3 kit the image
  name was MCC$DECW_PM.EXE.  The July 19 kit and the V1.1 IFT kit have the
  correct image name MCC_ICONIC_MAP_PM.EXE.  A release note should have been
  provided until the documentation is updated.  Sorry for any inconvenience.

							- Dave

    RE:.32
    
    I just saw a node where the exact same thing happened.  The latest
    DECmcc licenses were not installed.  I bet you didn't install the
    correct licenses.
    
    -Matt.