Re: [Diffserv] Comments on the TCB of the Conceptual Model - msg 1 of 2

[Andrew Smith <[e-mail]>]

Can you think of a better word than "Action" then? It is really just a label for
a chapter of the document - there's no implication in the text that an Action
element actually does something to change packets that pass through it.

Also, for the benefit of those on this list who haven't waded through the Policy
WG mail-list-morass, could you please explain in words of one syllable or less,
the difference between a data model and an information model?

Andrew


John Strassner wrote:
> 
> Hi Andrew,
> 
> I don't want you to rearrange the document. I don't even
> want you to change the modeling [sic] ;-) word. Rather, what
> I was trying to understand was the taxonomy used to develop
> the model.
> 
> Here's the root of the problem. Your four categories are
> classification, metering, actions, and queuing. What first
> confused me was the "action" category. Classifiers, meters,
> markers, droppers, and queues all are "actions" that are
> taken on the packet (whereas, for example, counters are not,
> even though they are included in your action category). So I
> was trying to understand how you defined an "action".
> 
> This then led to the more general question of an overall
> taxonomy, and how such a taxonomy could be used to build an
> information model as well as a data model. Using an OO
> approach, I would have preferred to see classifiers, meters,
> markers, droppers, and queues all as subclasses of a more
> general class that was used as the basis of conditioning
> traffic. The model as currently described implies that these
> are very different things with not a lot in common.
> 
> regards,
> John
> ----- Original Message -----
> From: "Andrew Smith" <[e-mail]>
> To: <[e-mail]>
> Cc: <[e-mail]>
> Sent: Monday, July 03, 2000 5:54 PM
> Subject: Re: [Diffserv] Comments on the TCB of the
> Conceptual Model - msg 1 of 2
> 
> > John,
> >
> > The 4 categories chosen are a fairly arbitrary taxonomy
> for grouping
> > descriptions of somewhat similar things into the same
> chapter of the document:
> > things to do with pattern matching on fields within
> packets are in one chapter,
> > things to do with measuring packet arrival events against
> some sort of history
> > are in another, things to do with pulling packets out of
> queues onto an output
> > are in another, etc.. We could rearrange the document into
> a single chapter, if
> > that would help you read and understand it, but I fail to
> see why the document
> > should not choose whatever arbitrary categories it wants
> to, if it provides a
> > structure for the descriptions.
> >
> > You seem to have a very fixed idea of what "modelling"
> (sic) means - maybe you'd
> > like us to change the name of the document to avoid the M
> word.
> >
> > Andrew
> >
> >
> > -----Original Message-----
> > From: John Strassner [SMTP:[e-mail]]
> > Sent: Wednesday, June 28, 2000 11:46 PM
> > To: Andrew Smith; [e-mail]
> > Subject: Re: [Diffserv] Comments on the TCB of the
> Conceptual Model - msg 1 of 2
> >
> > I got this from several hall discussions. So if this isn't
> > true, then what is the rationale for separating elements
> > into the four categories that you've defined? Especially
> > since all of them, from a modeling point-of-view, provide
> > actions...
> >
> > regards,
> > John
> > ----- Original Message -----
> > From: "Andrew Smith" <[e-mail]>
> > To: "'John Strassner'" <[e-mail]>;
> > <[e-mail]>
> > Sent: Wednesday, June 28, 2000 2:14 PM
> > Subject: RE: [Diffserv] Comments on the TCB of the
> > Conceptual Model - msg 1 of 2
> >
> >
> > > John,
> > >
> > > I don't know where you get the idea that "number of
> inputs
> > and outputs" is
> > > the fundemental taxonomy of the components of this
> model.
> > It's not.
> > >
> > > Andrew
> > >
> > >