Re: [Fwd: [Forces-protocol] Presentation of theoptionsforLFB-levelmulticast]

Weiming,

On Sun, 2004-11-07 at 21:02, Weiming Wang wrote:
> Jamal,
> 
> ----- Original Message -----
> From: "Jamal Hadi Salim" <hadi@znyx.com>
> >
> > Weiming,
> >
> > As i am reading your slides I cant to seem to see what the difference is
> > with your scheme (vs what i would call "other"). Maybe we are just
> > saying the same thing.
> Firstly, I should say the index based path is fairly the same as you mentioned.
> Only thing I just want to specifically mention is even in the index based path
> scheme, the field ID also plays important role, and the field ID is different
> from index (subscript).

Is this the same as content based description in slide 23?
Note that content based is under discussion

> >
> > Lets use your example to refer to set g1 to value of vg1:
> >
> > You say:
> > AttrID2[i=1].f2[j=1].g1=vg1
> >
> > Other scheme:
> > Assuming table refered by AttrID2 has ID=2 and f2 has ID=2 and g1 has
> > ID=1, then configuring g1 as you described is:
> > [2,1,2,1,1]=vg1
> >
> > Is your difference in how this is presented? i.e you seem to say:
> > [2,[1]],[2,[1]],1 ? Did i get this right?
> [Weiming]This expression is actually not important. The important is it should
> contain f2 ID as well as indexes.

Weiming, Isnt this f2 the same as #2 that i assumed above when i said
"Assuming table refered by AttrID2 has ID=2 and f2 has ID=2 "?
Maybe you are refering to a mixture of content and Ids? 

> > It seems to me that we are both saying that a vector, such as a table
> > entry will have two or more "coordinates" (table,index) but a scalar
> > such as what you refer to as AttrID1 will have a single entry.
> [Weiming]Yes, this is another point I just want to say. By this, I just want to
> say that index-based searching is not enough. We have to use field ID (content)
> based searching.
> 

Content based access is tabled under discussion in slide 23.

> >
> > Please respond as soon as you can because i am not sure what to say
> > and i have asked around here and havent anyone who could point the
> > difference.
> Because it seems quite discussion topic and not very proper to be put in our
> protocol presentation, therefore I just want to ask Robert to help to show it.
> 

I still dont see what where we have differences. If Robert can see that
difference i think it would be worth presenting it. 

cheers,
jamal

_______________________________________________
Forces-protocol mailing list
Forces-protocol@ietf.org
https://www1.ietf.org/mailman/listinfo/forces-protocol