$Id: whats_new_11.omh 3686 2015-05-11 16:14:25Z bradbell $
/* --------------------------------------------------------------------------
CppAD: C++ Algorithmic Differentiation: Copyright (C) 2003-15 Bradley M. Bell

CppAD is distributed under multiple licenses. This distribution is under
the terms of the
                    Eclipse Public License Version 1.0.

A copy of this license is included in the COPYING file of this distribution.
Please visit http://www.coin-or.org/CppAD/ for information on other licenses.
-------------------------------------------------------------------------- */

$begin whats_new_11$$
$spell
	cxx
	namespace
	boostvector
	mega
	inv
	bthread
	pthread
	pthreads
	Std
	retape
	cygwin
	vcproj
	pthread
	https
	abs_geq
	CondExpRel
	num
	Taylor
	Tanh
	Rosen
	Runge
	executables
	matlab
	Addr
	sizeof
	undefine
	config
	adolc
	initializer
	vec
	Bool
	op
	omp_alloc
	isnan
	gcc
	hasnan
	instantiation
	Microsoft
	svn
	hes
	openmp
	prefixdir
	pkgconfig
	Fadbad
	Wshadow
	omh
	dir
	libspeed
	cppad_ipopt_nlp.hpp
	automake
	makefile
	builddir
	gpl
	http://www.coin-or.org/download/source/CppAD/cppad-20110101.0.cpl.tgz
$$

$section Changes and Additions to CppAD During 2011$$

$head Introduction$$
This section contains a list of the changes to CppAD during 2011
(in reverse order by date).
The purpose of this section is to
assist you in learning about changes between various versions of CppAD.

$head 12-30$$
$list number$$
There was a bug when using $cref abs$$ with an $code AD< AD<double> >$$
argument, whereby the corresponding $code AD<double>$$ operation sequence
depended on the value of the argument to the $code abs$$ function.
$lnext
Change the specifications for the derivative of the $cref abs$$
function to be the $cref sign$$ function instead of a directional derivative.
$lnext
Add the $cref sign$$ function to the $codei%AD<%Base%>%$$ list of
available functions.
In addition, add the
$cref%sign%base_std_math%sign%$$ function to the list of
$cref%base type requirements%base_require%$$.
$lend

$head 12-28$$
The file $cref/time_test.hpp/time_test/Include/$$
was not being included by $code cppad/cppad.hpp$$.
This has been fixed.

$head 12-21$$
The types
$code SizeVector$$,
$code NumberVector$$,
$code ADNumber$$, and
$code ADVector$$,
were in the global namespace and this was causing warnings
about the shadowing of these declarations.
The
$cref/cppad_ipopt namespace/cppad_ipopt_nlp/cppad_ipopt namespace/$$
was created to avoid these problems.
The simplest way to make old $cref cppad_ipopt_nlp$$ code work with this
change is to use the command
$codep
	using namespace cppad_ipopt;
$$

$head 12-20$$
$list number$$
Change $code team_start$$ to $cref/team_create/team_thread.hpp/team_create/$$
and $code team_stop$$ to $cref/team_destroy/team_thread.hpp/team_destroy/$$.
$lnext
Change $code NDEBUG$$ mentions to include link to
$cref/NDEBUG/Faq/Speed/NDEBUG/$$.
$lnext
Improve $cref memory_leak$$ documentation.
$lend

$head 11-29$$
THe $cref time_test$$ routine was still executing the
test at least twice, even if that was not necessary for the specified
minimum time. This has been fixed.

$head 11-27$$
Move $code multi_thread.cpp$$ to $cref thread_test.cpp$$
and fix its $cref/running/thread_test.cpp/Running Tests/$$ instructions.

$head 11-24$$
Create $cref preprocessor$$ section with pointers to all the
preprocessor symbols that are in the CppAD API.

$head 11-21$$
Separate $cref/--with-boostvector/auto_tools/--with-testvector/$$
for $cref/boost_dir/auto_tools/boost_dir/$$.
This enables one to specify $icode boost_dir$$ for
$cref team_bthread.cpp$$ with out using boost vectors.

$head 11-20$$
$list number$$
Move $code sum_i_inv.cpp$$ to $cref harmonic.cpp$$.
$lnext
Include the date, time, CppAD version, and
$cref/team_name/team_thread.hpp/team_name/$$ in the
$cref thread_test.cpp$$ output.
$lend

$head 11-18$$
$list number$$
The $cref thread_test.cpp$$ program was truncating $icode test_time$$
to the nearest integer.
This has been fixed.
$lnext
The $cref time_test$$ routine has been made more efficient and now
check for the case where only one execution of the test is necessary
to achieve the desired $icode test_time$$
(it used to always run at least two).
$lnext
The $code sum_i_inv_time.cpp$$ and $cref multi_newton.cpp$$
routines were calling the test an extra time at the end to check for
correctness.
The results of the last test are now cached and used for the correctness
test so that an extra call is not necessary (to make the tests run faster
when only a few repetitions are necessary).
$lend


$head 11-17$$
$list number$$
Create another speed testing routine $cref time_test$$ which is
like $cref speed_test$$ but it returns the time instead of rate
and as a $code double$$ instead of a $code size_t$$.
The use it for the timing tests in
$code sum_i_inv_time.cpp$$ and $cref multi_newton_time.cpp$$.
$lnext
Add $icode test_time$$ as a command line argument
to the multi-threading $code sum_i_inv$$
and $cref/multi_newton/thread_test.cpp/multi_newton/$$
timing tests.
$lend


$head 11-09$$
Change $code thread_team.hpp$$ to $cref team_thread.hpp$$ and do
the same for all other names that ended in $code _team$$; e.g.,
$cref team_openmp.cpp$$.

$head 11-07$$
The users choice for $cref test_vector$$ was not actually changing
the tests that the user ran. This have been fixed.

$head 11-06$$
Make all the output generated by $cref multi_newton.cpp$$ valid
matlab and octave input so it is easy to plot the results.

$head 11-04$$
Use thread specific data to simplify $cref team_openmp.cpp$$.

$head 11-01$$
Make $cref team_bthread.cpp$$ more similar to $cref team_pthread.cpp$$.

$head 10-30$$
$list number$$
Reorganize and improve the $cref multi_thread$$ section and its subsections.
$lnext
There was a bug in $cref multi_newton_work.cpp$$ that only showed up when
the number of threads was greater than or equal 4.
This has been fixed.
In addition,
$cref/CPPAD_MAX_NUM_THREADS/multi_thread/CPPAD_MAX_NUM_THREADS/$$
was increased from 2 to 4 (to enable testing for this bug).
$lnext
The accuracy expected in the $code  sum_i_inv.cpp$$ results were failing
when $code mega_sum$$ was large.
This check has been modified to include a correction for $icode mega_sum$$.
$lend

$head 10-29$$
The following changes were merged in from $code branches/thread$$:

$list number$$
Move $code openmp$$ to $code multi_thread/openmp$$.
and create $code multi_thread/bthread$$, $code multi_thread/pthread$$
with similar tests.
$lnext
Put multi-threading common code in $code multi_thread$$ directory
and threading system specific code in $codei%multi_thread/%threading%$$
for $icode threading$$ equal to $code openmp$$, $code bthread$$,
and $code pthread$$.
$lnext
Update the README file.
$lnext
Remove the $code bug/optimize.sh$$ file (no longer a bug).
$lnext
Make $code arc_tan.cpp$$
utility that can be used by multiple multi-threading tests.
$lnext
Create $cref team_thread.hpp$$ specifications,
move OpenMP management to $cref team_openmp.cpp$$,
Boost thread management to $cref team_bthread.cpp$$,
and pthread management to $cref team_pthread.cpp$$.
$lnext
All of the make files were modified so that the command
$codep
	make test
$$
would run the tests for the current directory.
$lnext
Extend the multi-threading speed tests
$code sum_i_inv.cpp$$ and $cref multi_newton.cpp$$ so they run
using Boost threads and pthreads (as well as OpenMP threads).
$lend

$head 10-14$$
Fix some compiler warnings about shadowed variables that were
detected by $code g++$$ version $code 4.6.1 20110908$$.

$head 10-12$$
$list number$$
The MAC version of the $code pthread$$ library does not include the
$code pthread_barrier_wait$$ function; i.e., is not compliant with
the IEEE Std 1003.1, 2004 Edition for $code pthread$$.
This caused the $code pthread_simple_ad.cpp$$ to fail to compile on the MAC.
This has been fixed by not compiling the $code pthread$$ examples
unless $code pthread_barrier_wait$$ is present.
$lnext
The $cref cppad_ipopt_nlp$$ routine has been changed to
$cref optimize$$ the functions $latex r_k (u)$$ such that
$codei%retape(%k%)%$$ is false.
$lend

$head 09-06$$
$list number$$
Add the
$href%
	http://www.boost.org/doc/libs/1_47_0/doc/html/thread.html%
	boost multi-threading
%$$
examples $cref a11c_bthread.cpp$$ and $code bthread_simple_ad.cpp$$.
$lnext
Improve documentation for
$cref/thread_num/ta_parallel_setup/thread_num/$$ argument
to $code parallel_setup$$.
$lnext
More simplification of $code bthread_simple_ad.cpp$$ example.
$lend

$head 09-05$$
Simply and fix some problems with $code pthread_simple_ad.cpp$$,
including avoiding a $cref/Bug in Cygwin/team_pthread.cpp/Bug in Cygwin/$$.

$head 09-02$$
$list number$$
The OpenMP speed test program $code openmp/run.cpp$$ was not setting
the number of threads for the one thread case (so dynamic thread adjustment was
used). This has been fixed.
$lnext
The $cref thread_alloc.cpp$$ example was missing from the
Microsoft $code example/example.vcproj$$ file and a attempt was made
to link to missing OpenMP routines (this has been fixed).
In addition,
some Microsoft compiler warning have been fixed; see the examples and tests
in the Windows install instructions.
$lnext
There was an oversight,
and $code CPPAD_MAX_NUM_THREAD$$
was being set to 2 when $code _OPENMP$$ was not defined.
This has been fixed and
$cref/CPPAD_MAX_NUM_THREADS/multi_thread/CPPAD_MAX_NUM_THREADS/$$
has been documented and is now part of the CppAD API.
$lnext
The $code pthread_simple_ad.cpp$$ test failed under cygwin.
This was because the previous test $code openmp_ad.cpp$$ was
set up calls to OpenMP routines that were still in effect
when $code pthread/simple_ad$$ ran.
This has been fixed by making
$icode%num_threads% == 1%$$ a special case in
$cref/parallel_setup/ta_parallel_setup/$$.
$lend

$head 09-01$$
$list number$$
Modify the CppAD trunk using the
changes from svn revision 2060 to revision 2081 in the branch
$codep
	https://projects.coin-or.org/svn/CppAD/branches/pthread
$$
These changes are described below under the headings
$cref/08-21/whats_new_11/09-01/08-21/$$ through
$cref/08-31/whats_new_11/09-01/08-31/$$.
$lnext
There was a bug in the $cref old_atomic$$ functions
in the case where none of the elements of the argument to the function
was a $cref/variable/glossary/Variable/$$.
This has been fixed.
In addition, $cref old_tan.cpp$$ generated an assert for this
case and this has also been fixed (in addition to including an example
for this case).
$lend

$subhead 08-31$$
$list number$$
Move the $code sum_i_inv_time.cpp$$ test from
$code openmp/run.sh$$ to $code openmp/run.cpp$$.
$lnext
Change $code --with-openmp$$ to
$cref/OPENMP_FLAGS=openmp_flags/auto_tools/openmp_flags/$$ configure
command line argument.
$lend

$subhead 08-30$$
$list number$$
Create the $code openmp/run.cpp$$ program
and move the $code openmp_multi_newton.cpp$$ test from
$code openmp/run.sh$$ to $code openmp/run.cpp$$.
This uses $cref/configure/auto_tools/Configure/$$
information for building the tests.
$lnext
Document the $code --with-openmp$$ configure
command line argument.
$lnext
Move $code openmp/multi_newton.hpp$$ to
$code openmp/newton_method.hpp$$ and
$code openmp/multi_newton.cpp$$ to
$code openmp/newton_example.cpp$$.
$lend

$subhead 08-25$$
$list number$$
Replace $cref omp_alloc$$ by $cref thread_alloc$$ in
$cref multi_thread$$, the section on how to use CppAD in parallel.
$lnext
Implement $cref omp_alloc$$ as links to corresponding $cref thread_alloc$$
sections.
$lnext
Create the $code pthread_simple_ad.cpp$$ example that does AD using the
pthread library.
In addition, fix some problems in $code openmp_simple_ad.cpp$$
$lnext
Move $code openmp/example_a11c.cpp$$ to
$cref|example/a11c_openmp.cpp|a11c_openmp.cpp|$$.
$lnext
Move $code openmp/parallel_ad.cpp$$ to
$code openmp_simple_ad.cpp$$.
$lend

$subhead 08-23$$
Beginning steps in replacing $cref omp_alloc$$ by $cref thread_alloc$$:
$list number$$
Replace $cref omp_alloc$$ by $cref thread_alloc$$
in the $cref library$$.
$lnext
move $cref omp_alloc$$ to the deprecated section of the
documentation.
$lnext
Change all $cref omp_alloc$$ section names to begin with $code omp_$$,
and change all $cref thread_alloc$$ section names to begin with $code new_$$.
$lnext
Convert $cref CppAD_vector$$ from using $cref omp_alloc$$ to using
$cref thread_alloc$$ for memory allocation.
$lnext
Extend the $cref memory_leak$$ routine to also check the
$cref thread_alloc$$ allocator.
$lend

$subhead 08-21$$
Create the OpenMP and pthread examples
$cref a11c_openmp.cpp$$,
$cref a11c_pthread.cpp$$, and
$code openmp_simple_ad.cpp$$.
These OpenMP examples were originally in the $code openmp$$ directory,
and have been moved, and modified to conform, to the normal example directory.

$head 08-11$$
Modify the CppAD trunk using the
changes from svn revision 2044 to revision 2056 in the branch
$codep
	https://projects.coin-or.org/svn/CppAD/branches/base_require
$$
These changes are described below under the headings
$cref/08-04/whats_new_11/08-11/08-04/$$ through
$cref/08-10/whats_new_11/08-11/08-10/$$.

$subhead 08-10$$
$list number$$
Add the output stream optional argument $icode s$$ in
$codei%
	%f%.Forward(0, %x%, %s%)
%$$
See $cref/zero order forward mode/forward_zero/$$ and $cref PrintFor$$.
$lnext
Improve $cref omp_alloc.cpp$$ example.
$lend


$subhead 08-09$$
$list number$$
$cref/Warning:/base_require/Warning/$$
Add $cref/epsilon/base_std_math/limits/$$ to the
$icode Base$$ type requirements.
$lnext
Extend $code epsilon$$ to AD types.
$lend

$subhead 08-08$$
$list number$$
Improve the $cref base_require$$ documentation for
$cref/standard math functions/base_std_math/$$.
$lnext
$cref/Warning:/base_require/Warning/$$
Add $code abs_geq$$ to the $cref/requirements/base_require/$$
for a user defined $icode Base$$ type.
$lnext
Check that zero order forward mode results are approximately equal,
instead of exactly equal,
after an $cref optimize$$ operation.
This fixes a bug in the optimize correctness check
(The  order of operations can be changed by $code optimize$$
and hence the zero order forward mode results may not be exactly the same.)
$lend

$subhead 08-07$$
Improve the $cref base_require$$ documentation for
$cref/EqualOpSeq/base_identical/EqualOpSeq/$$,
$cref/Identical/base_identical/Identical/$$
$cref/Integer/base_require/Integer/$$, and
$cref/Ordered/base_ordered/$$
operations.

$subhead 08-06$$
Add the $cref/CondExpRel/base_cond_exp/CondExpRel/$$ paragraph to the
base requirements documentation.
This was missing and are required for
$cref CondExp$$ to work with $codei%AD<%Base%>%$$ arguments
and a non-standard $icode Base$$ type.

$subhead 08-04$$
$list number$$
$cref/Warning:/base_require/Warning/$$
Change the $cref/include/base_require/Include Order/$$ file name to
$cref/base_require.hpp/base_require/$$.
$lnext
Use $cref base_float.hpp$$ and $cref base_double.hpp$$ as additional examples
for the $cref/CondExp Base requirements/base_cond_exp/$$
requirements.
$lend

$head 08-03$$
Change $cref PrintFor$$ condition from less than or equal zero
to not greater than zero;i.e., not positive.
This makes $code nan$$ print because it results in false for all comparisons.

$head 08-02$$
$list number$$
Change $cref PrintFor$$ so it no longer aborts execution when
there is no operation sequence being recording; see
$cref/start recording/Independent/Start Recording/$$.
$lnext
Improve the $cref print_for_cout.cpp$$ example.
$lend

$head 07-31$$
Add a conditional version of the $cref PrintFor$$ command
$codei%
	PrintFor(%text%, %y%, %z%)
%$$
which only prints when $icode%z% <= 0%$$.
This is useful for error reporting during forward mode; i.e.,
reporting when the argument to the $code log$$ function is not valid.

$head 07-29$$
$list number$$
The routines $cref/set_max_num_threads/omp_max_num_threads/$$ and
$code get_max_num_threads$$ were created.
User's will need to replace calls to
$cref/max_num_threads/old_max_num_threads/$$
by calls to $code set_max_num_threads$$.
$lnext
The functions $cref omp_efficient$$ was deprecated because it
has not been shown to be useful.
$lend

$head 07-28$$
$list number$$
Change $cref omp_return_memory$$ so that if $cref omp_max_num_threads$$ is one
(the default), $cref omp_alloc$$ does not hold onto memory
(keep it available for the corresponding thread).
$lnext
Add files that were missing from the Microsoft Visual Studio
$code example$$ and $code test_more$$ subdirectory
project files.
$lnext
Fix some warnings generated by Microsoft Visual Studio 2010 build.
$lend

$head 07-27$$
Make $code tan$$ and $code tanh$$ $cref/atomic/glossary/Operation/Atomic/$$
operations; see $cref tan_forward$$ and $cref tan_reverse$$.

$head 07-25$$
Finish the $cref old_atomic$$ example $cref old_tan.cpp$$.
This is also a design and implementation of the routines necessary
to make $code tan$$ and $code tanh$$ CppAD atomic operations.

$head 07-18$$
The reverse mode formulas for $latex Z(t)$$ need to involve
the lower order Taylor coefficients for $latex Y(t)$$.
This has been fixed in $cref tan_reverse$$.

$head 07-17$$
$list number$$
Fix bug in $cref old_atomic$$ functions.
To be specific, the Taylor coefficients for $latex y$$,
of order less than $icode k$$,
were not passed into the $code old_atomic$$
$cref/forward/old_atomic/forward/$$ callback function.
$lnext
Derive the theory for including the tangent and hyperbolic tangent
as CppAD atomic operations $cref tan_forward$$ and $cref tan_reverse$$;
see the wish list item $code Tan and Tanh$$.
$lnext
Implement and test forward mode calculation of derivative for
the tangent and hyperbolic tangent functions;
see the new $cref old_atomic$$ example $cref old_tan.cpp$$.
$lend

$head 07-14$$
$list number$$
The $cref auto_tools$$ instructions for running the individual
correctness and speed tests were out of date.
This has been fixed; see
$cref/example and tests/auto_tools/make/Examples and Tests/$$.
$lnext
Move $code parallel_ad.cpp$$ from $code example$$ directory
to $code openmp$$ directory (and convert it from a function to a program).
$lnext
Simplify $code example_a11c.cpp$$ by making it just a correctness test.
$lnext
Change $code openmp/run.sh$$ so that it runs correctness tests with
the compiler debugging flags.
$lend

$head 07-13$$
$list number$$
static hash code data that was begin used by multiple threads
when recording $codei%AD<%Base%>%$$ operations
$cref omp_in_parallel$$ execution mode. This has been fixed.
$lnext
Make the $cref/sparsity/Sparse/$$ calculations safe
for use during $cref omp_in_parallel$$ execution mode.
$lnext
Add the $code parallel_ad.cpp$$ example.
$lnext
Change $code example_a11c.cpp$$ example so that is just a correctness
(not speed) test.
$lend

$head 07-11$$
$list number$$
Change the upper limit for
$cref omp_max_num_threads$$ from 32 to 48.
$lnext
Add $cref/parallel/ta_in_parallel/$$ documentation for,
$code nan$$,
$cref/Rosen34/Rosen34/Parallel Mode/$$, and
$cref/Runge45/Runge45/Parallel Mode/$$.
$lnext
Fix $cref CheckNumericType$$ and $cref CheckSimpleVector$$
so they work properly when used in parallel mode.
$lend


$subhead openmp/run.sh$$
The following changes were made to $code openmp/run.sh$$:
$list number$$
$lnext
Change to $code openmp/run.sh$$
maximum number of threads
instead of specifying the entire set of values to be tested.
$lnext
Change settings for
$code newton_example$$
so that $code n_gird$$ is a multiple of the maximum number of threads.
$lnext
Report dynamic number of thread
results as a separate result in the summary output line.
$lnext
Fix automatic removal of executables from $code openmp$$ directory
(was commented out).
$lnext
The documentation for $code openmp/run.sh$$ was moved to the
$code multi_thread$$ section.
$lend


$head 07-10$$
$list number$$
Add link to $tref Discrete$$ in $cref multi_thread$$.
$lnext
Make use of the $cref TrackNewDel$$ routines
$cref omp_in_parallel$$ execution mode an error (it never worked properly);
see $cref/TrackNewDel multi-threading/TrackNewDel/Multi-Threading/$$.
$lnext
Change $cref memory_leak$$ so that it checks for a leak in all threads.
This is what $code openmp_newton_example.cpp$$ and
$code sum_i_inv_time.cpp$$
assumed was being done.
$lend

$head 07-09$$
All the OpenMP parallel execution requirements
have been grouped in the section $cref multi_thread$$.

$head 07-07$$
Add the routine $cref parallel_ad$$ to fix bug when using
$codei%AD<%Base%>%$$ in $cref/parallel/omp_in_parallel/$$ execution mode.

$head 06-23$$
$list number$$
Fix a bug whereby the assert
$codei%
	Error detected by false result for
		! omp_in_parallel()
	at line %n% in the file
	%prefix%/include/cppad/omp_alloc.hpp
%$$
sometimes occurred.
$lnext
The routine $cref omp_max_thread$$ was deprecated,
use the routine $cref omp_max_num_threads$$ instead.
$lnext
The deprecated routines have been grouped together in the
$cref deprecated$$ section of the CppAD manual.
$lend

$head 06-21$$
$list number$$
The $code openmp/run.sh$$ routine was changed to use zero,
instead of $code automatic$$, for automatic choice of
$code openmp/run.sh$$
number of repeats
and
maximum number of threads.

$lnext
The output of each of the OpenMP examples / speed tests
(run by $code openmp/run.sh$$)
was changed to be valid matlab / octave assignment statements.

$lnext
In the case where OpenMP is enabled during compilation,
a summary for the different number of threads as added at the end of
the $code openmp/run.sh$$ output.

$lend

$head 06-18$$
$list number$$
The $cref/tape_addr_type/auto_tools/tape_addr_type/$$ option was added
to the $cref/configure/auto_tools/Configure/$$ command line.
$lnext
The function $cref/size_op_seq/seq_property/size_op_seq/$$ results uses
$code sizeof(CppAD_TAPE_ADDR_TYPE)$$
where it used to use $code sizeof(size_t)$$.
$lnext
Remove $code cppad/config.h$$ from CppAD distribution,
(put the information in $code cppad/configure.hpp$$.)
This removes the need to undefine symbols that were
defined by $code cppad/config.h$$ and that did not begin with $code CPPAD_$$.
$lnext
Change $cref/adolc/auto_tools/adolc_dir/$$ library linkage so it
works with version $code ADOL-C-2.2.0$$.
$lend

$head 05-29$$
Fix bug (introduced on $cref/05-22/whats_new_11/05-22/$$)
whereby constructor might not be called (but required) when the
$cref/base type/base_require/$$ is not plain old data.

$head 05-28$$
$list number$$
Add the $cref omp_efficient$$ routine to the $cref omp_alloc$$ system.
$lnext
Improve the $code omp_alloc$$ tracing so it prints the same pointer
as returned to the user (not an offset version of that pointer).
$lend

$head 05-26$$
Fix Visual Studio project files that were broken during the change on 05-22.
In addition, in the file $code cppad/omp_alloc.hpp$$,
suppress the following Microsoft Visual Studio warning
$codep
	warning C4345: behavior change: an object of POD type constructed with
	an initializer of the form () will be default-initialized
$$


$head 05-22$$
$list number$$
The old memory tracking routines $cref TrackNewDel$$ have been deprecated.
Their use should be replaced using the $cref omp_alloc$$
a memory allocator which is designed to work well in a multi-threading OpenMP
environment; see $cref/purpose/omp_alloc/Purpose/$$.
$lnext
The replacement of $code TrackNewDel$$ by $code omp_alloc$$
has been throughout the CppAD source code, including the examples
that used $code TrackNewDel$$; namely,
$cref mul_level_adolc.cpp$$, $cref mul_level_adolc_ode.cpp$$.
$lnext
The CppAD vector template class
and the $cref/vectorBool/CppAD_vector/vectorBool/$$ class
were modified to use the $code omp_alloc$$
$cref/memory/CppAD_vector/Memory and Parallel Mode/$$ manager.
This should improves its speed of memory allocation
$cref omp_in_parallel$$ sections of a program.
$lnext
The $cref speed_test$$ argument
$cref/size_vec/speed_test/size_vec/$$ call was by value,
instead of by reference (as documented).
This has been fixed and the call is now by reference.
$lnext
The $cref/capacity/CppAD_vector/capacity/$$ function has been added
to the CppAD vector class.
$lnext
The simple vector
$cref/
	element constructor and destructor/
	SimpleVector/
	Element Constructor and Destructor
/$$
description has been changed to explicitly specify that the
default constructor is used to initialize elements of the array.
$lnext
The $cref/size_op_seq/seq_property/size_op_seq/$$ documentation
has been improved to mention that the allocated memory may be larger.
$lend

$head 05-11$$
$list number$$
Avoid ambiguity in the definition of the
$cref/complex isnan/base_complex.hpp/isnan/$$ function.
$lnext
Errors during $code make test$$ were not
being detected.
This has been fixed.
$lend

$head 05-03$$
$list number$$
If $code NDEBUG$$ is not defined,
the $cref/hasnan/nan/$$ function is used to make sure that the
results of any $cref Forward$$ operation does not contain a
nan (not a number).
If so, an error message is generated and the program terminates.
This error message and termination can be caught; see $cref ErrorHandler$$.

$lnext
In the event that the
$cref cppad_ipopt_nlp$$ objective function, the constraints,
or their derivatives are infinite,
an error message is generated and the program terminates
(proved that $code NDEBUG$$ is not defined and the default error
handler has not been replaced).
$lend

$head 04-29$$
$list number$$
The Microsoft Visual Studio 2003 project files
for the Windows examples and tests no longer worked because
the current version of CppAD uses local types in template instantiation; see
$href%
	http://msdn.microsoft.com/en-us/library/bh44f2cb(v=vs.71).aspx%
	Compiler Error C2918
%$$.
These project files were converted to Visual Studio 2008 where they do work
(if you use a later version, Visual Studio should automatically
convert them for you).

$lnext
The old speed test directory was moved to $code speed_cppad$$
before the new $cref speed$$ test organization was created on 2006-12-11
(revision 715 of the repository).
The old speed tests have not been used for years and so have been deleted.
$lend

$head 04-20$$
The $code openmp/run.sh$$ script what changed
to take an argument that specifies which tests is run
(it no longer runs all the tests).
Also improve the openmp test program output formatting.

$head 04-19$$
The $icode use_ad$$ option was added to the
$code openmp_newton_example.cpp$$ test case.

$head 03-19$$
The subversion write protected directory
$code bin/.svn$$ was mistakenly part of the
$cref/compressed tar file/download/Compressed Archives/$$.
It has been removed.

$head 03-11$$
The vector of sets argument
$cref/r/old_atomic/rev_hes_sparse/r/$$
to the $code old_atomic$$ function $icode rev_hes_sparse$$
must have size greater than or equal to $icode n$$.
There was a check that its size was greater than or equal $icode q$$.
This was incorrect and has been fixed.

$head 03-05$$
Add the $cref/conjugate gradient/conj_grad.cpp/$$ example.

$head 02-22$$
Add the $cref/atomic/speed_main/option_list/atomic/$$ option to the
speed test program and use
$cref old_mat_mul.hpp$$ during the
$cref cppad_mat_mul.cpp$$ speed test when the atomic option is specified.


$head 02-19$$
There was a bug when $cref omp_max_thread$$ was set to one,
and $code NDEBUG$$ was not defined,
the thread corresponding to parameters was one, but the only valid thread
number was zero (only one thread)
and an CPPAD stopped with an assertion error.
This has been fixed.

$head 02-17$$
There was a mistake in $code openmp/run.sh$$ where it attempted
to remove a non-existent file in the case where
$code openmp/run.sh$$
openmp_flag was not $code ""$$.
This has been fixed.

$head 02-15$$
A matrix multiply speed test has been added.
So far, this has only implemented for the
$cref/cppad/cppad_mat_mul.cpp/$$ and $cref/double/double_mat_mul.cpp/$$ cases.
(For the time being this test is not available for the other
speed comparison cases.)

$head 02-09$$
A variable in $code old_atomic.hpp$$ was declare of type $icode Base$$
when it should have been declared of type $code size_t$$.
It caused the $cref old_atomic$$ feature to fail with some base types.
This has been fixed.
$pre

$$
The $cref old_mat_mul.hpp$$ example has been improved by
caching the $latex x$$ variable information and using it during
$cref/reverse Hessian sparsity/old_atomic/rev_hes_sparse/$$ calculations.
$pre

$$
Some of the $cref old_atomic$$ documentation was extended to
include more explanation.

$head 02-06$$
The use can now define complex $cref/atomic/old_atomic/$$ operations
and store them in a CppAD $cref ADFun$$ object.
This item has been remove from the
$cref/wish list/WishList/$$.
$pre

$$
The documentation for $cref RevSparseHes$$ had a dimension error.
This has been fixed.
$pre

$$
A faster set operations item was added to the wish list.
This has since been satisfied by
$cref/cppad_sparse_list/cmake/cppad_sparse_list/$$ choice during
the install process.

$head 02-02$$
The documentation for $cref ForSparseJac$$ had some formatting errors.
The errors have been fix and the documentation has been improved.

$head 02-01$$
The subversion
install instructions were brought up to date.
They have since been replaced by just separate
$cref/subversion download/download/Subversion/$$
instructions.


$head 01-19$$
The directory where the $cref pkgconfig$$ file $code cppad.pc$$
is stored has been moved from
$icode%prefixdir%/lib/pkgconfig/cppad.pc%$$
to
$icode%prefixdir%/share/pkgconfig/cppad.pc%$$;
see
$href%http://lists.fedoraproject.org/pipermail/devel/2011-January/147915.html%
devel@lists.fedoraproject.org%$$.

$head 01-16$$
The following have been fixed:
$list number$$
The install of the documentation failed
when it was done from a directory other than the top source directory.
$lnext
The GPL distribution had the output of the
$cref/configure/auto_tools/Configure/$$ command in it.
$lnext
Since the change on 01-09, the file
$code omh/whats_new_11.omh$$ has been required to build the
documentation (and it has been missing from the distribution).
$lnext
Fadbad was generating warnings due to the $code -Wshadow$$ flag
with the $code g++$$ compiler.
The Fadbad $cref/speed/speed_fadbad/$$ tests have a special
flag with this warning removed from the
$cref/cxx_flags/auto_tools/cxx_flags/$$.
$lend

$head 01-09$$
There were some problems running $code make test$$ in the releases
$codei%
http://www.coin-or.org/download/source/CppAD/cppad-20110101.0.%license%.tgz
%$$
where $icode license$$ is $code gpl$$ or $code cpl$$.
$list number$$
The version of automake used to build the corresponding
$code makefile.in$$ files did not define $code abs_top_builddir$$.
$lnext
The include file $code cppad_ipopt_nlp.hpp$$ was always installed, even if
$cref/ipopt_dir/auto_tools/ipopt_dir/$$ was not defined on the
$code configure$$ command line.
$lnext
The speed test library $code libspeed.a$$ was being installed
(it is only intended for testing).
$lend
These problems are fixed in the trunk
and these fixes will be copied to the corresponding stable and
release versions; i.e.,
$codei%
http://www.coin-or.org/download/source/CppAD/cppad-20110101.1.%license%.tgz
%$$
will not have this problem.

$end