Re: Question regarding the scope of the WG

[Bill Woodcock <woody@pch.net>]

I agree that the goals you stated, if taken at that level of detail, would
be a daunting task for a simple draft.  However, I think that operators
would, by and large, prefer to leave the "how" details to tools developers
like yourself.  We're much more interest in the "what" than the "how".  So
at the risk of seeming glib, let me address these in one-liners:

    > »·······configuration database

That's inside the machine, and we don't care how you do it.

    > »·······configuration generator

If you mean a tool which we (operators) use to create configurations,
which are then loaded into boxes, it's imperative that such _not be
necessary_.  If we need them, we can hack them together in perl.  If it's
too complicated for us to hack together in perl, it's too complicated,
period, and needs to be fixed.

    > »·······tool to push the config to the network
    
Unnecessary.  Serial cables, ftp, tftp, bootp, and scp all already exist.

    > »·······trend analysis tools 

Out of scope.

    > »·······billing collection
    
Shouldn't be different from any other kind of statistical collection.

    > »·······tools to evaluate network health 
    
Out of scope.

What we're really interested in here, I believe, is how operators need to
be able to communicate with their equipment.  What we do with the data
once we've got it is out of scope, and how you represent and handle the
data internally to the box is out of scope.


                                -Bill