Conference azur::mcc

  (H)  Allow Entity relationships -- such to navigate between entities.

       Such a fix would elminite the need to place Alarm Rules under
       the Domain Entity.  Rules should exist under MCC/Alarms/Rule.

       Relationships allows you to select an entity (say a node4) and
       'jump' to the Alarms Entity to access Rules .. with a *minimum*
       of mouse movements...Say the MB2 button  will have RULES on the
       menu which will instantly connect you to Alarm Rules.

       Likewise, this technique can help simplify Exporting & Recording too.

  /keith

(H) Ability to lock the map to prevent unauthorised changes.

(H) Allow multiple circuits to be drawn between 2 icons without placing the  
    lines on top of each other - this was a major complaint from a customer. 

(H) Have 1 icon to represent a node running multiple protocol stacks. eg. You 
    should not have to represent a PhaseV Ultrix system with both a Node and 
    SNMP icon. 

(M) Optional grid on the map to make it easier to line up icons etc.

(M) Ability to be able to rubberband a group of objects and automatically align
    them either horizontally or vertically - similiar to functionality provided
    in MSU.

(M) Allow the Find function to work across multiple domains.

(M) Allow horizontal movement within the domain hierarchy - it is very time 
    consuming having to keep going up and then down, particularly with the 
    PhaseV AM because of the large number of child entities. 

Regards

Ian Pratt

These suggestions are from BBN.

(H) Allow the user to RESET an alarm severity from an alarm rule.  
    You can clear the alarm rule now, but it just posts a clear
    notification so the entity does not get set to the latest
    problem.  It would be nice to have a feature that works from
    an alarm rule that is analogous to what the user can do from
    the Iconic Map.

(L) Line names do not track the line very well when the lines are moved.

(M) Allow the user to set pipeline width in a similar fashion to line
    width.


...kjn

(H) Complete a task with fewer required mouse clicks....example.  When select
an entity then pull down the operations list, many of the operations don't 
require additional parameters, such as TEST or REGISTER.  If the user selects
one of these, why through up an additional screen and have them START the
process.  A single entry of the desired function should be sufficient.

(M) Allow the ability for a user to add their own entries for MB2 prompting
of data fields.  i.e.  when asked to fill in list items such an reference
data, many of these fields will be used in many entities.  There should be
a way to add a user name, or telephone number or descriptive string to
a list and simply save or recall that item using the MB2 feature.

(L) The ability to modify any item in the MCC_RESOURCE.DAT file from the
iconic map.  Main screen colors, etc.

(M) More intutitive menu options and organization......How many people know
what SALIENCE means w/o having to reference a dictionary.  COLOR CHANGE may
not be as eligant and it may require a few more characters, but its meaning
is much clearer. 
    Not all selections seem to be were you might think they should be and
in some cases you need to go too many places to accomplish similar tasks.
Why use three different windows in the management area to do what is currently
being done.  Wouldn't it be better to move all three to a single display and
organize the display into logical units (i.e. maps, window placement, window
characteristics, defaults....etc.)   This might also be a good place for
heirarchical windows.
    Rather than just asking what size and where you want various windows 
in terms of pixels or relative position statements, why not do something
similar to MicroSoft Windows where a smaller window is drawn and each item
can be selected/clicked and then worked with to define its color.  Take this
one step further and allow the user to place the screens then save this
placement scheme.
    Remove or replace any ambiguous or non-obvious selection or reference
(i.e. in the customization map....PROPAGATION.....propagation of what)
After having worked with MCC for a period of time, I realize this means
propagation of color change with regard to alarms, but is it all that obvious?
    Which window is the VIEW PORT......any item which would cause a user to
reference the rather vast documentation set is time lost and leads to 
frustration.  

(M) The ability to zoom on maps.....in working with another field test (R/CMP)
they also draw maps to represent entities on a system or group of systems.
They give the user the ability zoom in or out of the maps allowing more or
less detail as needed.  Large complex networks could benefit from a wide view
and zoom when alarms occur.

(M) The ability to seperate certain window displays from the rest and have
same displayed either at other locations or on other workstations.  During
one demo, a customer asked if just the notification window could be brought
up w/o the others, or if notifications could be displayed on another
workstation.  

(L) If not given the ability to do the previous request then perhaps the 
ability to fill the background screen with one display such as the network
map and have other displays overlayed and moved w/o causing the background
to come to the foreground unintentionally.

(M) About the only feature that I liked on HP Openview that I thought we may
be lacking and this may not really be an iconic map feature but could be
concidered a user interface feature, is the ability to describe a group of
macro events, if you will, to be selected and performed with a single
operation.  IN this case, I believe they are able to group and save a series
of MIB variables.  These groups can be listed with user defined names and
place in a menu list.  A single click on the list intitates a report of the
chosen MIB variables.  I believe they can also use this list to start a
real time graph.


Hope these are useful.

Al

>(H) Allow the user to RESET an alarm severity from an alarm rule.  
>    You can clear the alarm rule now, but it just posts a clear
>    notification so the entity does not get set to the latest
>    problem.  It would be nice to have a feature that works from
>    an alarm rule that is analogous to what the user can do from
>    the Iconic Map.

The current functionality covers this to a certain extent.  The "clear" severity
is the lowest severity.  If an alarm rule fires, then clears itself, any other
notification with a higher severity on that same entity will "take over" the 
color of the entity.

BBN's suggestion is probably a result of earlier baselevels, which didn't
do de-escallation of notifications properly.  You might want to check with them
after confirming that they've seen T1.2.7 that this "reset" capability is
still H.

>(H) Allow the user to RESET an alarm severity from an alarm rule.  
>    You can clear the alarm rule now, but it just posts a clear
>    notification so the entity does not get set to the latest
>    problem.  It would be nice to have a feature that works from
>    an alarm rule that is analogous to what the user can do from
>    the Iconic Map.

* The current functionality covers this to a certain extent.  The "clear" severity
* is the lowest severity.  If an alarm rule fires, then clears itself, any other
* notification with a higher severity on that same entity will "take over" the 
* color of the entity.

The clear is overridden by a higher severity only if you have chosen 
the `most severe' option, not the `latest severity' option.  Even if I 
chose the 'latest severity' option, I might want the abilty to clear a 
critical alarm and have the severity drop back to the 'latest severity' 
problem, not counting the clear.

...kjn

    H) Improve the performance of the iconic map PM. It takes along time on
    both VMS and ULTRIX to "open" an enity for the first time by double
    clicking. Listing and opening domains is also very slow.
    
    H) Make navigation easier by allowing both navigation across domains
    and allowing wildcards in the entity name. Highlight the icon
    representing the entity that is found.
    
    M) Allow a user to go right from a notification to the entity that
    caused the alarm or event. If several entities caused it, allow the
    user to traverse the list.
    
    H) Make it easier to modify alarm rules. I would like to set a modify
    option in the rule operation pull-down menu. I don't know an easy way
    to change, for example, the polling rate on an alarm from the iconic map pm.
    
    H) Make it easy to copy alarm rules using the iconic map pm. I may want
    to create the same rule with slight modifications for several entities.
    
    M) Have an autodiscovery program that will update an existing map which
    was previously created with the autoconfiguration program. This allows
    the network manager to keep the maps up to date without recreating
    them. These "discovered" entities could just be registered in the
    domain and have to be placed on the map by the network manager who may have
    already customized the map. 
    
    H) Allow flexible naming on alarm rules and navigation. Registered
    names and synonyms should be able to be used interchangeably. We should
    never have to specify the namespace prefix to get alarms to work
    properly.
      
    Comment: Baselevel 1.2.7 added many important features especially the
    ability to use wildcards for the global entity in alarm rules and the
    autoconfiguration programs. Keep up the good work!                    

(M) Provide quick access to frequently used functions through a Function
    Bar at the top of the window. That bar should contain Icon Buttons for
    frequently used functions e.g. show characteristics, enable, disable, view
    rules/members, look up...
    For an example look at DECwrite V2.0 or at various pc products such as MS
    Word or Excel.

FWIW, .8 is right on target.  We could also use some more "status" information
in addition to just the domain name to keep .8's "smart-icons" company.

(H) Many folks have discussed user-defineable attribute groups.  The IMPM
    needs user-defineable mega-attribute groups, where "mega" means that
    the groups can span levels of the entity hierarchy.  So, for example, the
    user can click on an SNMP global entity, which would activate a bunch of
    "smart-icons" specific to SNMP a-la suggestion in .8, some of those 
    smart-icons being a show of the important information about that SNMP
    entity, e.g. the most important attributes from all of its interfaces
    along with other important information, regardless of where in the
    entity hierarchy it resides.

    Here is my vote for the "theme" of IMPM V1.3:

	DIVORCE THE USER INTERFACE FROM THE ENTITY MODEL
	(well, as much as possible in a short period of time)

    What's the estimate of SHOW operations to all others?  5:1 I'd guess?
    What's one of the biggest pains in IMPM?  LOOK-INTO an entity/child entity
    hunting for the correct subentity/partition combination to give you the 
    information you're looking for.  You give folks the ability to SHOW relevant
    information from multiple instances of multiple child entity levels in one 
    or two mouse button clicks, and you've gone a long way towards "fixing" the
    interface.

(H) Tag this one onto the user-definable mega-attribute-group: let the user
    define the form for displaying the attributes.  With this capability, an
    AM developer and/or management expert could ship pre-packaged 
    smart-icons/forms that make the product much more useable out of the box.

Make sure you check the mcc_futures conference for any ideas.

    Hi,
    
    	here are some requests from Peugeot, using DECmcc for about one
    year. The majors improvements requested are :
    
    - to have a kind of physical global entity (not a domain ...) grouping
    on the iconic map all the global entities defined for a physical entity
    (for example station + node4 + SNMP). One of the goal is when you get
    an alarm on this physical global entity, you expand it and then you
    work on the "classic" global entity. (H)
    
    - to have minimum graphical tools. It means, as under DECwrite, to have
    grids, scales, aligning, grouping and moving facilities in a domain.
    (H)
    
    - to have a lock switch on the map disabling all modify action on
    maps. (I think it is a very easy feature, it should be in official
    version 1.2.) (H)
    
    - to have a link between an entity instance and its icon somewhere in
    the MIR. This is to be able in FCL procedures to say REGISTER <entity>
    , WITH ICON = <icon identifier>. This to be able to create maps with
    the right icon for each entity instance. This will also give the
    ability to really COPY an entity from one domain to another one,
    keeping the right icon. (M)
    
    - to have a severity level, or a graphic symbol (as the bell), associated 
    to state "entity under alarming". In version 1.2 it is impossible to know
    immediatly, without requesting rule status, if an entity is polled by a 
    rule. (H)
    
    - to have an easy way to choose the character string for an entity. I
    know it is possible under 1.2 but it must be easy from the iconic map.
    (L)
    
    - to have an interdomain find. (L)
    
    Thanks for your great work.
    Fran�ois EBLE.
    EIS Strasbourg France.    
    

Hi there,

(H) Multiple lines (different entities) between icons.

(H) User Defined Partitions. This is a **great** way of allowing the user to
    define a single click for those always done operations. As mentioned in a
    previous note this operation should be able to provide any attributes the 
    user wants from the entire entity/child structure. MCC should allow
    several 'buttons' to be user defined. This is kind of like radio station
    presets (6 am and 6 fm) except it would be 3 node4 and 3 snmp and 3 bridge
    'buttons' the user can define. Wildcards should be supported.

(H) Text. Something should be done with this. I see improvements in 1.2 with
    the alt id but I still have Circuit subentity text which I had to move off
    the map into the 'Black Sea' (read: ugly). How about an option when 
    placing it on the map to make it non-visible, or the user can define his 
    own text.

(H) For the autoconfig buffs, they need more map editing abilities. The ability
    to select an entity(s) and 'push' them into or 'pull' them out of domains
    is clearly needed. Also maintaining lines as subentities from domain to
    domain should be supported (see MSU for all of this).

(M) A single click on icons containing ... should mean "select all children".
    If you develop user defined partitions as described above this one could
    probably be put on the shelf.

(M) Provide map zooming.

(M) Provide for looking 'across' domains rather than just up and down.

(M) Notify events should by default only be shown once in the 'box' if it is
    the same event, same entity, just different domains.

kind regards,
brad...

    
    Hello,
    
    Some of these suggestions have all ready been made but I'll second them
    for the sake of emphasis.
    
    1) I know this is obvious and probably in progress, but MOTIF is
    needed. (H)
    
    2) Allow change of attribute partition within the operations display
    window, as with "change entity".  (M)
    
    3) Add "all attributes" as a qualifier to the show command for the 
    operations pulldown. (L)
    
    4) Allow the ability to manage a multi-protocol entity from one icon.
    (I understand that this involves much more that the user interface.)
    (M)
    
    5) Add an operations pop-up menu to MB3. (M)
    
    6) Provide a rubberband function to select multiple entities. (L)
    
    7) Add "Deselect all" to the edit pull-down. (H)
    
    8) Allow FIND command to work across domains. (H)
    
    9) Improve performance of IMPM startup. (H)
    
    Thanks,
    Scott

Some of these suggestions have already been made above.


(H)	Performance (especially startup and navigation) and use of 
	resources (especially memory) should be significantly improved.

(H)	Reduce the number of verbs in the operations menu pulldown. With
	the addition of the Historian and Exporter directives to every entity,
	the pulldown menu has too many verbs. A suggestion is to group the
	Record and Export directives into separate [-> sub-menues.  

(H)	Allow for the input and output of large (subject to ILV/DT limits)
	attribute values.

(M)	When double-clicking on an entity class, Display instances as they 
	are received instead of waiting to receive all before displaying 
	them. 

(H)	List box: When there is only one attribute partition for a given
		  verb, the attribute partition should be automatically
		  selected when the verb is. This is annoying with verbs 
		  like Set, Add, and Remove when the only settable attributes
		  come from characteristics partition.   

(M)     Allow continuous display (via scheduled time) for (multi-instance)
	table mode displays.

(L)	Allow different attribute partitions to be selected from an operations
	window.

(L)	Immediately execute the operation when selected from the operations
	menu when there are no arguments instead of opening another window.

(M) 	Simplify the procedure to modify the poll rate of the real-time 
	graph. Also relax the restriction of minimum scheduled time 
	(30 seconds ?)

(L)	Relax the restrictions of 10 seconds for minimum scheduled time.


Thanks,
Phil

    There are a couple of other issues for V1.2, or a future release of
    DECmcc.
    
    1.  Currently, there is a problem with the IMPM in that you CANNOT
        scroll to the right, nor can you position the Navigation Window
        "box" far enough to the right to see the whole picture.  I have
        only seen this when viewing an AutoTopology map.
    
    2.  I have always had a problem moving some of the bigger icons.
        Sometimes, though not always, I move the icon over 10 pixels or
        so and it jumps back to its origial position.  I would like to
        be able to move entities (lines, icons, etc.) where ever I choose
        without the IMPM snapping them to another location.  It would be
    	great to have a "SNAP TO GRID" that you could turn ON or OFF with
    	an option selection.  Then you could adjust the grid to whatever
    	measure you desired and everything would always line up.
    
    3.	When selecting a domain to open, I don't always have all of my
    	domains committed to memory.  It would be nice to be able to look
    	into a domain, then if it wasn't the right one, I could come back
    	to the FILTER/SELECTION window (with all of my domains still 
    	present) so I could select another.   At the present, this window
    	goes away every time.  How about an "OK" button and an "APPLY".
    
    4.	It would be great to be able to "freeze" an icon or other entity
    	on the map so you could put other icons, texts, etc. on it without
    	having them one of them hidden when you select the first icon.
    
    5.	A zoom capability is really needed.
    
    6.	I would really like to manipulate multiple icons at once. 
    	Selection would be made via "rubber-banding" or by holding down
    	the shift key and clicking the icons desired.  Then these icons,
    	as well as lines and text, could be moved from one domain map to 
    	another, newly created domain map, where they could be placed 
    	"as is".  This would be a great way to build subset domains.
    
    7.	I would like to be able to change line width without having to 
    	delete the current line and re-draw it with a thinner line.
    
    Thanks,
    dh

    Here are a few more ideas directly from one of my customers:
    
    1.	Include "easy-to-use" print utility as a default part of the 
    	Iconic Map for printing maps and window contents.
    
    2.	Provide a "set host" or "telnet" capability via MB2 menu select
    	option to allow you to pop up a login window for a selected host.
    
    3.	Provide "easy-to-use", iconic-based report builder that will allow 
    	the user to select fields (via point & click), drag them into place
    	in an output window, and then have the data displayed below these
    	field titles once the report is invoked.  The report could then be
    	saved for later use.
    
    4.	Definitely must be able to freeze entities on a map so they will
    	not move when you click on them.  Simply clicking on some entities
    	causes them to jump an inch or so to the left or right.  It can
    	then be a real pain to put them back since they seem to have a 
    	mind of their own when you are trying to place them.
    
    5.	Allow selection of info from windows for cutting and pasting into
    	other windows. (i.e., characteristics info)
    
    6.	Provide a window based method of entering multiple nodes with the
    	same reference information.
    
    Thanks,
    Dan

(H) Allow other applications to request the IMPM to carry out operations.

    Specifically support for;

    Adding an entity to a map window. The data passed would be 

    - Domain name
    - Entity
    - Registration args if applicable
    - Create args if applicable
    - Set Reference Arguments if applicable
    - Create Domain Member Arguments if applicable
    - x,y co-ordinates on the map
    - severity of the entity (mapped to colour change to reflect equipment
      status (e.g. un-commissioned)

    Removing an entity from the map. The data passed would be

    - Domain name
    - Entity name
    - remove indicator ( delete this domain member only, delete member from
      all domains, delete member from all domains and deregister entity).


  This feature would allow for automated update of both the MIR and the map
files to reflect the current managed objects. This is important where large 
numbers of entities are required to be added periodically from another
application e.g. an inventory management system.

Thanks - Francis. 
   

    (M) Add a switch to enable or disable a BELL for notifications from a
    certain level of severity.
    
    Dominique.
    

Hi all,

a customer of mine has the following suggestions (which I support :-)):

- Add a Delete Button to the Show Notification Detailed window. This would ease
reading and coping with notifications a lot.

- Add a reference section to SNMP Interface subentities. This would alllow the
network manager to enter information about leased lines used etc.


Thanx in advance and keep up the good work




Josch