1*******************  VERSION 2.3.2 *********************
2More free() troubles.  Thanks to ken_yap atatatat E M A I L dot C O M
3for pointing out a segfault.  Took out malloc()/free() calls for both
4holdaddress and holdprefix.  strtok() calls assign these pointers right
5off anyway.
6
7Thanks to Herman Robers <herman.robers@pinewood.nl>
8for the Solaris and RedHat stuff, and the nice Makefile.
9
10*******************  VERSION 2.3.1 *********************
11doh!
12Bugfix in cleanup.  Pointers had been set to NULL before free()'ing.
13
14*******************  VERSION 2.3 ***********************
15
16Improvements:
17
181.  Manual page.
19
202.  "-v" option to print version information.
21
223.  modified usage().
23
24Minor bugfixes:
25
261.  Fixed problem with segfault when only one argument is given and
27    that argument is not valid.
28
292.  Fixed stupid malloc().
30
31*******************  VERSION 2.2 *********************
32
33This version incorporates ideas suggested by Iain Lea, namely:
34
351. Short form usage, such as `cidr 192.168.1.1/27`
362. Better output format.  IP addresses are aligned.
373. Example uses in usage() output.
38
39I think these are nice improvements and justify this somewhat rapid
40update.
41
42Bugs:
43
441. Doug Denny reports that inet_aton() and inet_ntoa() cause linker
45   errors when building cidr 2.0 under Solaris 2.7.
46
472. ip address 255.255.255.255 / mask 255.255.255.255 seems to hang on
48   the host count.  I say "seems" because for some masks there is a long
49   wait while it calculates the host count, but those masks are usually
50   the ones resulting in a large number of hosts, such as a /8 and above.
51   A /32 count is usually instantaneous, but it seems to get stuck when
52   used with 255.255.255.255 ip address.
53
54All previous functionality remains.  Long form is still there:
55i.e. `cidr -b 11111111000000001111111100000000 -q 255.255.255.255 -H`
56
57*******************  VERSION 2 *****************************
58
59This version incorporates ideas suggested by David A. Bandel:
60
611. There is now a "-H" option to print host addresses.
62
632. The host addresses will correctly represent valid addresses for
64   unusual netmasks, i.e., masks with non-contiguous 1-bits, such as
65   11111111111111111111111100001000
66
673. The total addresses count now reflects only a count of valid host
68   addresses within the given subnet.  In version 1, the network address
69   and broadcast address were included in the count.
70
71