Re: ISP Cases Draft: Which sections have too much information?

[itojun@iijlab.net]

>From the WG meeting today there was consensus that there
>was too much information in the draft.  Can we get specific
>feedback on which sections should be paired back?
>http://www.ietf.org/internet-drafts/draft-mickles-v6ops-isp-cases-02.txt

	i understand your motive to document the operational details, but
	i don't think the detail all down to the copper level will help
	readers - readers will need a proper level of abstraction.

	for instance:

page 20, figure 3.3.2
page 22, figure 3.3.3
page 23, figure 3.3.4
	do you really need to go into ATM/AAL5?  i don't think it make any
	difference even if we remove all ATM/AAL5 details (there's no real
	reference to ATM/AAL5 in the text).  there's no reference to PPPoE,
	or PPPoA - there's no difference in the description.  the most
	important points are
	(1) there are 3 IP devices - customer hosts, customer router, and
	    ISP edge router
	(2) customer router and ISP edge router will establish a point-to-point
	    connectivity, either by ATM PVC, PPPoE, or PPPoA.
	(3) in some cases, ISP edge router uses L2TP to aggregate connections
	(4) customer router may implement NAT (sigh)

	if you abstract it to proper level, subsections in 3.3.2 will become
	a single diagram.

<--customer------>                <--ISP---------->

+-----+  +-------+                +----+------+
|Hosts|--+Router +----------------+    ISP    |  C  
+-----+  +-------+                |    Edge   +=>O
                                  |   Router  |  R
                                  +-----------+  E

	same goes to section 3.6 (public access wireless LAN).  i don't think
	there's any difference between 802.11b-based solution and 802.11x-based
	solution, with respect to IPv6 transition.

itojun