1* Release 3.0
2Early work on a github release for the masses.
3
4* Release 2.5
5
6Added in various documentation fixes from Michael Ernst (BTW: has anyone
7considered getting slave labour or University Students doing technical
8writing in CS to do proofreading as an exercise for GNU documents. At
9least from the teaching point of view this would seem to me to be a good
10thing).
11
12Added in various fixes from Eugen Dedu which include re-releasing
13the perf/ performance measurement library and various other fixes.
14
15* Release 2.4
16
17Built in a new test system which may or may not work :-) but aims to
18handle separate build/install directories properly. The complete set
19of tests has not been ported to the new test system yet. That can wait
20for the next revision.
21
22Added in Paul Suggs fixes.
23
24Added in Phil Bleckers patches for 2.3, in particular:
25
26    o Check for gettimeofday in configure.in, and blindly assumed that if
27      it's not found then ftime is available.
28
29    o Changes to now.c are included to use ftime. (This allowed me to
30      cross-compile from Unix to MS-Windows using mingw32.)
31
32    o Check to see if stdin is defined around the prototype to L_buffer_dump
33      so that L_buffer.h can be included by itself in C++ files without nana.h.
34
35    o Remove the hard-coded CFLAGS from a couple of Makefile.am.
36
37
38Redid the changes for cross compiling suggested by Alex Lancaster.
39
40* Release 2.3
41Eiffel.h has been modified to provide DO .. END macros for invariant
42checking at the suggestion of Bertrand Meyer (Eiffels author).
43
44* Release 2.2
45Fixes for cross compiling supplied by Alex Lancaster applied, minor
46changes to configuration system.
47
48* Release 2.1
49
50	0. Added cycles.h which does cycle level timing using the
51	   CPU specific instructions or registers. (This is currently
52           only supported on Pentium compatible machines using the
53	   RDTSC instruction and ./configure --enable-rdtsc).
54	1. Minor fixes to the test system.
55
56* Release 2.0
57Added a few extra tools and some more documentation, in particular:
58
59	0. Nana Shortform Generator -- generates a HTML summary of your codes
60	   interface.
61	1. Nana Performance Measurement -- measures the space/time overhead
62	   of the nana library (or anything else).
63	2. Minor bug fixes/introductions.
64	3. Modified the test system to use the gnuware package
65	   to keep track of version, host, target, changes.
66
67* Release 1.14
68Minor fixes for egcs and a bit of documentation tidying up.
69
70* Release 1.13
71Fixes to documentation and problems with compiling under Solaris.
72
73* Release 1.12
74Added WITHOUT_NANA macro to allow you to compile without any
75nana features. You must use the VL((...)) version of macros
76for any macro with a variable number of arguments.
77[Hopefully this solution is better than the previous one]
78
79* Release 1.11
80Minor fixes to documentation.
81
82Added functions for handling the variable number of arguments
83to macros problem (well sort of, see VL(...) in L.h). This was
84suggested by Gordon Matzigkeit.
85
86* Release 1.10
87
88Added a flag, etc to L_buffer.* so that the writing to circular buffers
89can have wraparound disabled (thus preserving the first (oldest) k bytes
90of messages, rather than the default last (newest) k bytes).
91
92A L_time component has been added that logs a time stamp and char*
93into a circular buffer. It should be a bit faster than L_buffer since
94it doesn't use vsnprintf formatting.
95
96A majordomo mail archive has been setup at (nana@it.ntu.edu.au).
97Send "subscribe nana <your-address>" to nana-request@it.ntu.edu.au
98
99After some discussions with Richard M. Stallman (rms) Nana will be
100released as a GNU tool. Note that the release conditions remain the
101same, i.e. you can still use nana in a commercial project.
102		[Thanks to Richard for all his work over the years]
103
104* Release 1.09
105Added a call list modules (calls) which is a container for
106function calls. This can be used to print all the objects
107in a program or to check internal consistency.
108
109This version should be released on the FreeBSD 2.2.5R CDROM.
110
111* Release 1.08
112
113Some minor fixes to the check-mail target.
114Modified L_buffer_puts to allow a NULL pointer as the argument, this maps
115to "(null)" (suggested by Phil Blecker).
116
117* Release 1.07
118Added examples and documentation for the use of Eric Allman's syslog
119with nana.
120
121Fixed an embarassing bug in L_buffer.c.
122
123* Release 1.06
124A fix for a small autoconfigure problem under FreeBSD plus some documentation
125changes. The FreeBSD ports mechanism defines INSTALL_SCRIPT which
126seems to confuse the autoconf 2.12 generated configure script.
127
128* Release 1.04/1.05
129This fixes a bug that was introduced into L_buffer.c at 1.03
130to allow for machines without vsnprintf. Thanks to Neal Becker
131and Oliver Elphick.
132
133* Release 1.03
134Minor changes to documentation, fixes to tests and fixes to various
135portability problems.
136
137* Release 1.02
138Some minor changes to the testing mechanisms and libraries.
139The check target now submits the result to a mail alias where
140I can check the results.
141
142* Release 1.01
143
144L_buffer.h - improved documentation for circular buffer logging etc.
145
146eiffel.h - support for eiffel like assertions (class invariants, method
147	pre and postconditions).
148
149Qstl.h - support for Standard Template Library (STL) iterators and quantifiers.
150	(only for C++ programmers)
151
152L_buffer.h - documented and tidied the logging to a circular buffer library.
153	This keeps the last X messages in an in-core buffer.
154	(this is ideal for those "what was the system doing before it
155	broke" questions).
156
157Some hopefully useful changes to configuration, in particular:
158
159% I_DEFAULT=fast ./configure
160 		which uses single instructions for assert failure
161vs
162
163% I_DEFAULT=verbose ./configure
164		which prints out a nice message (and then dumps core).
165
166Documentation updated and improved (still a way to go).
167
168Tests have been rewritten.
169
170Updated to automake-1.1o.
171
172* Release 1.00
173
174I've removed the Ada support from the system and made up a FreeBSD
175port and package of the system. I've also implemented some of the
176suggestions made by George K.Bronnikov and Alexander Samoilov.
177
178
179
180