On Fri, Jul 11, 2008 at 02:01:46PM +0200, ropers wrote:
| However, there's a part of the document (
| http://www.3com.com/other/pdfs/infra/corpinfo/en_US/501302.pdf ) that
| I haven't yet corrected, and can't/won't correct on my own without
| asking for your opinion. I'm talking about the section on IPv6: While
| the rest of the document is quite good except for the formatting
| errors, the IPv6 section appears to have been added as an afterthought
| and edited by someone who didn't know or care, with the expectation
| that nobody would notice. This section is the numbered pages 43
| through 51 (pages 45 through 53 of the PDF). I wonder if you all could
| help me with the following:
|
| - On page 43, Figure 36 is described as depicting "a full hexadecimal
| to binary IPv6 address". However, the image only shows seven, not
| eight 16-bit integers (with leftmost zeroes omitted each time).
This is obviously an error, I'd even consider the 0 .. 128 wrong with
the leftmost zeroes omitted. The author clearly has counting issues
(and would be better of using 'more real' addresses anyway).
| - Formatting niggle (not a bug): On page 43/44, it would be better to
| structure/emphasize or bullet-point the paragraphs starting "The
| preferred form/The compressed form/The third form".
At any rate, 'the third form' is considered a bad idea around these
parts. Do v4 or do v6, don't combine the two in one notation.
| - On page 44, in table 4, the bottom row goes:
| > "FF01:0:0:0:0:0:0:43 the unspecified address ::"
| The address FF01:0:0:0:0:0:0:43 that's given in this row is the same
| as the one given in the second row, where it is described as "a
| multicast address". I don't think FF01:0:0:0:0:0:0:43 is correct in
| the bottom row, and I don't see how it would correspond to ::.
The unspecified address is 0:0:0:0:0:0:0:0, the compressed form is
correct (::).
| - On the same page, it says:
| > "This form is represented as X:X:X:X:X:X:X:X:D.D.D.D. Where the Xs
represent the hexadecimal values of the six high order 16-bit pieces of the
address."
| I think that X:X:X:X:X:X:X:X:D.D.D.D. has two Xs too many. That should
| be six, not eight Xs, am I right?
You are right .. should be 6 Xs, but again .. bad form ;)
| - On the same page, in table 5, the address 0:0:0:0:FFFF:129.144.52.38
| is given in the bottom row. I think that's one 0 too few. It should be
| 0:0:0:0:0:FFFF:129.144.52.38, am I right?
You are right .. should be 5 0s, but again .. bad form ;)
| - On page 47, in Figure 38, the same IPv6 address 3FFE:2900:1::/48 is
| given twice, both with NLA1 and SLAII. That can't be right, can it?
I think Figure 38 is mostly confusing, especially with the text
immediately following it. There's multiple errors in that part which
makes it even more confusing.
| - On page 49, an IPv6 address example list is given. It is explained
| about that list that
| > "the underlined portion of each address identifies the network prefix which
is calculated through the prefix length notation. In other words, since /48 is
the prefix length notation, then the first four integers will be the prefix
length and the rest will be the interface ID. Because each integer equals 16,
the calculation is 48 divided by 16 equals 4."
| However, the underlined parts of the addresses shown in the list on
| that page are 64 bit long, not 48, and 48 divided by 16 does not equal
| 4 (but 64 divided by 16 does).
Yuck.
| - On page 50, Figure 39 is titled
| > "ISP #2's More Specific Route into the Internet"
| Wouldn't "Internet routing examples" be a more appropriate title,
| given that an ISP #2 or ISP #1 is not mentioned elsewhere in the
| document?
|
| - On the same page, in Figure 39, the same IPv6 address
| ABF2:45AF:2574:9980:7654:FCD4:FF26:0072 is shown as assigned to two
| different PC interfaces (cf. right hand side, under router B). That
| can't be right, can it?
Uhm, I don't understand what page 50 is doing there at all. It says :
"Figure 39 demonstrates the versatility of IPv6 addressing." and then
shows a completely useless (and WRONG !) image that does not tell us
anything. Yes, we can assign IP addresses to devices. Boo-frickety-hoo
.. I think we've already established that. How does this demonstrate
the versatility of IPv6 addressing ? "Look, you can assign one IP
address to two PC's" .. yes, if you want to break your network you can
do that sort of crap, but why would you ?
| I would love to hear your feedback/corrections. In fairness, it may be
| quite a while until I get ahold of that old 3com email exchange I
| mentioned, and until I get around to getting on 3com's
| nerves^W^W^W^W^W^W contacting 3com.
|
| Many thanks for your help and your patience.
I must admist, I've only read the parts you pointed out, but so far
I'm very much impressed with how they managed to totally confuse the
reader with incorrect statements and wrong examples. The section on
IPv6 is probably best removed or mostly rewritten.
Cheers,
Paul 'WEiRD' de Weerd
PS: OK, so I'm reading a bit more in this document (the v6 section)
and it is so full of utter crap that it isn't even funny anymore.
Favorite quote : "Because each integer equals 16 [holy crap!], the
calculation is 48 divided by 16 equals 4 [uncanny crayons, batman!]."
--
>++++++++[<++++++++++>-]<+++++++.>+++[<------>-]<.>+++[<+
+++++++++++>-]<.>++[<------------>-]<+.--------------.[-]
http://www.weirdnet.nl/
