J. Blustein
Kurose & Ross Misprints
- Page 1, Section 1.1, line 2
principle
should be principal
.- Page 19, Figure 1.7
Mbs
should be Mbps
- in three places.
- Page 36, paragraph 3
-
In
ADSL:
asymmetric digital subscriber loop
-
(loop should be line)
- Page 44, under
End-to-End Delay
, line 8
-
pair or routers
-
should be
pair of routers
.
- Page 54, Figure 1.26
- the circles should be concentric (not overlapping)
- Page 82, Figure 2.5, the last line
-
In
Typically UPD
-
UPD should be UDP
- Page 112, Paragraph 3, the middle line:
-
Section 2.3
-
should be Section 2.2
- Page 120, line 14
-
principle
should be principal
.
- Page 158, last line of the first paragraph
-
FileName
should be fileName
- Page 181, section 3.3.2 (about UDT
-
This is not an error, but you should be aware that the carry bit
from the highest order bits is copied to the sum in the lowest order
bit. For example (using 4 bits instead of 16):
1000+1000=0001 and 1001+1001=0011.
-
See RFC 1936 for more
details.
- Page 187, line 3
receiver
should be sender
- Page 191, Figure 3.14
sendpktsndpkt
- Page 198, Figure 3.19, left-hand side
-
The window should not be advanced for just any non-corrupted ACK. The window should only
advance if the ACK is for a packet that is in the
current window, i.e. if
(acknum(rcvpkt) == base) || (acknum(rcvpkt) >
base). If the ACK is for a packet with
sequence number less than base then the packet should be
ignored.
-
The event part (above the line) should read:
rdt_rcv(rcv_pkt)
&& notcorrupt(rcvpkt)
&& getacknum(rcvpkt)>=base
- Page 199, Figure 3.20, the top line
-
rdr_rcv(...)
-
should be
rdt_rcv(...)
- Page 199, Receipt of an ACK
- The description of cumulative acknowledgment is correct but
the description of GBN is wrong. There are different ways of implementing
GBN and the one
described in this paragraph is different from the one described
later in the chapter (on page 220, in the section entitled
`A Few Interesting Scenarios' inside section 3.5.5).
- The GBN protocol that Kurose &
Ross discuss later requires that the sender receive an
individual ACK for each packet. The version of
GBN described here (with cumulative
acknowledgments) is inconsistent with the later discussion and is
wrong.
- Page 207, first line
packet ordering
- should be
packet re-ordering
.
- Page 217, Figure 3.31, under
event
, line 3:
-
compue
should be compute
.
- Page 220, first partial paragraph
- See the note about GBN ACKs on page 199.
- Page 262, problem 3:
-
The reference to Figure
3.12
-
should be to Figure 3.11.
See Also
See also the web page of misprints maintained by the authors:
- Go to the online book's about page, and
- follow the errata link.
Acknowledgments
Most of the errors were reported by Dr. E. Grundke and Yufei Sun.
Jason Rouse pointed out the error in Fig. 1.26 (p. 54). Ahmed El-Rayes pointed out the contradiction
between the descriptions of Go-Back-N on pp. 199 and 220.
- Version:
- 23 July 2002
- CS 3171 Prof.:
- J. Blustein <jamie@cs.dal.ca>
