Changes:
- Changed unfactoring to improve simplification (again). This was a
bug fix (first one in a long time).
- Restored "epsilon" to original value in version 10.9b. Required
for factoring large polynomials.
- Changed factor_times() to preserve divides. This improved the
"unfactor" command.
- Added polynomial factoring to the very beginning of "simplify"
command. This is necessary in case you "unfactor" a complicated
equation, then "simplify".
USER-VISIBLE CHANGES
o Underscore '_' is now allowed in syntactically valid names, and
make.names() no longer changes underscores. Very old code
that makes use of underscore for assignment may now give
confusing error messages.
o Package 'base' has been split into packages 'base', 'graphics',
'stats' and 'utils'. All four are loaded in a default
installation, but the separation allows a 'lean and mean'
version of R to be used for tasks such as building indices.
Packages ctest, eda, modreg, mva, nls, stepfun and ts have been
merged into stats, and lqs has been returned to MASS. In all
cases a stub has been left that will issue a warning and ensure
that the appropriate new home is loaded. All the time series
datasets have been moved to package stats. Sweave has been
moved to utils.
Package mle has been moved to stats4 which will become the
central place for statistical S4 classes and methods
distributed with base R. Package mle remains as a stub.
Users may notice that code in .Rprofile is run with only the
new base loaded and so functions may now not be found. For
example, ps.options(horizontal = TRUE) should be preceded by
library(graphics) or called as graphics::ps.options or,
better, set as a hook -- see ?setHook.
o There has been a concerted effort to speed up the startup of
an R session: it now takes about 2/3rds of the time of 1.8.1.
o A warning is issued at startup in a UTF-8 locale, as currently R
only supports single-byte encodings.
plus many new features and bug fixes.
Changes:
- Removed some simplification optimizations in favor of correctness
and simpler code. Minor slowdown.
- Doubled "MAX_COMPARE_TERMS".
- Simplify order poly_div then smart_div reversed.
- Kludge for size checking in poly_div removed.
- Broke "am.in". Edited.
- Added "simplify poly" option to do poly_div first.
Changes:
- Fixed HTML mode "screen_rows" and "screen_columns" to be unlimited.
- Allow square brackets ([]) as parentheses and optimized parser.
- Simplification was improved.
Changes are bug-fixes mostly, but also synchronizes bundled GD
with GD 2.0.22 and updates PCRE to version 4.5. Several NetBSD
patches were integrated too, so future pkgsrc updates would
be even more smooth.
Full list of changes since PHP 4.3.4 is available at:
http://www.php.net/ChangeLog-4.php#4.3.6http://www.php.net/ChangeLog-4.php#4.3.5
Changes since 10.9b:
- Removed more MS-DOS specific code; functionality improved.
- Implemented what I call "fraction slack". This perfects float to
fraction conversion, and now Mathomatic doesn't rely on perfect
floating point routines. "-O" option added to CFLAGS in file
"makefile" (we can do this now).
- Changed modf() to fmod() where possible.
- Accuracy increased one more digit by reducing the "epsilon" variable.
- Cleaned up "makefile".
- Took a step towards internationalization with gettext(3) by
calling _() for all English strings. Everything ported except for
the "help" command.
- Commented out some unfactoring code that wasn't helpful and took out
the "pause" command from "all.in" for easier testing.
- Rearranged "simplify" command to simplify better.
- Removed some more unfactoring code and made simplification much
better. Go figure. Perfection has been achieved.
- Added "unfactor fully" option.
- Added "make test" to fully test Mathomatic after compilation.
- Removed masking of 8th bit in all Mathomatic input.
- Updated primes program with a nice user interface.
* Version for GNOME 2.6 (FINAL) call for tarballs. Includes:
- new locales for pa, en_GB, bg and ga.
- localized online user documentation from the Sun translation team.
for de, es, fr, it, ja, ko, sv, zh_CN, zh_HK and zh_TW.
Packaged by Bruce J.A. Nourish, and slightly modified by me.
Mathomatic is an automatic algebraic manipulator that is self-testing
and strictly follows the rules of algebra.
Mathomatic is an interpreter that can:
- Solve, simplify, and compare algebraic equations.
- Combine simultaneous equations.
- Perform basic calculus operations.
- Generate the sensitivity formula for one or more variables in an
equation.
- Act like a double precision floating point programmable calculator.
- Perform complex number and polynomial arithmetic.
- Compute the Greatest Common Divisor of numbers or polynomials.
- Generate "C" code from equations.
explicitly unset NeedFunctionPrototypes, which can no longer safely be done in recent
X11 sources (as _Xconst is only defined if this is set, yet is used everywhere).
by moving the inclusion of buildlink3.mk files outside of the protected
region. This bug would be seen by users that have set PREFER_PKGSRC
or PREFER_NATIVE to non-default values.
BUILDLINK_PACKAGES should be ordered so that for any package in the
list, that package doesn't depend on any packages to the left of it
in the list. This ordering property is used to check for builtin
packages in the correct order. The problem was that including a
buildlink3.mk file for <pkg> correctly ensured that <pkg> was removed
from BUILDLINK_PACKAGES and appended to the end. However, since the
inclusion of any other buildlink3.mk files within that buildlink3.mk
was in a region that was protected against multiple inclusion, those
dependencies weren't also moved to the end of BUILDLINK_PACKAGES.
Many, many, many additions, bug fixes, and improvements since the last
packaged version. Too many to even summarize here. Please refer to the
change log in the distfile or on the homepage.
Changes sinec 5.1.12:
* [ES] use multiplication sign in formulae of regression report
* [ES] allow for DL modules to define functions of 4 and 5 parameters
* [ES] fixed compilation under FreeBSD-5.1 and OS/X
* [ES] on 64-bit platforms, when specific compiler/optimization
combinations were used, some dialogs could crash randomly
* [ES] on 64-bit platforms, when specific compiler/optimization
combinations were used, X run-time warnings could be issued
about unrecognized XKeySym's
* [ES] set filling defaulted to even-odd filling rule in the X11
driver
* [ES] fixed potential memory corruption
* [ES] fixed page size sync in the free mode
* [TO] fixed a race condition in the grace_np library
* [ES] enabled annotated values for HILO sets
* [ES] added IMIN(x) and IMAX(x) functions returning index of minimal or
maximal value of array x
* [ES] updates for T1lib-5.0 API
* [WB] documented commands for calculating running properties
* [MD] enabled font kerning in the MIF driver
* [ES] Grace might fail starting in the GUI mode when a certain
sequence of command line arguments was passed
* One more slight adjustment to the fix for bug 135064
(http://bugzilla.gnome.org/show_bug.cgi?id=135064)
Entering numbers that start with the numeric point character ('.")
are now visibly displayed.
* Adjustment to the fix for bug 135064
(http://bugzilla.gnome.org/show_bug.cgi?id=135064)
Fix to remove a potential spurious initial thousands separator
character after a change sign operation.
Fixed calc man page examples to move -f to the end of the line.
Thanks goes to Michael Somos <somos at grail dot cba dot csuohio
dot edu> for pointing this out.
Linux and gcc now compiled with -Wall -W -Wno-comment.
Fixed a post increment that was reported by R. Trinler <trinler at
web dot de> and fixed by Ernest Bowen <ernie at turing dot une dot
edu dot au>.
Fixed pi.cal to not depend on the buggy pre-2.11.9 post increment
behavior.
Added config("cygwin") to determine if calc was compiled under Cygwin.
The config("cygwin") is a read-only configuration value that is 1
when calc was compiled under Cygwin and 0 otherwise. Regression
tests 949 and 950 are skipped when config("cygwin") is true.
The Makefile variable HAVE_NO_IMPLICIT is empty by default so that
the Makefile will test if the compiler has a -Wno-implicit flag.
Added HAVE_UNUSED Makefile variable. If HAVE_UNUSED is empty,
then the Makefile will run the have_unused program to determine
if the unused attribute is supported. If HAVE_UNUSED is set to
-DHAVE_NO_UNUSED, then the unused attribute will not be used.
The Makefile builds have_unused.h which defines, if the unused
attribute is supported:
#define HAVE_UNUSED /* yes */
#define UNUSED __attribute__((unused)) /* yes */
or defines, if the unused is not supported (or if the Makefile
variable is HAVE_UNUSED= -DHAVE_NO_UNUSED):
#undef HAVE_UNUSED /* no */
#define UNUSED /* no */
Fixed numerous warnings about comparison between signed and unsigned
value warnings and unused parameter warnings in version.c, zrand.c,
string.c, shs1.c, shs.c, qtrans.c, qmath.c, qfunc.c, md5.c, matfunc.c,
hist.c, file.c, const.c, blkcpy.c, seed.c, opcodes.c, func.c, qio.c,
zrandom.c, custom/c_argv.c, custom/c_devnull.c, custom/c_help.c,
custom/c_sysinfo.c, addop.c and calc.c.
Fixed some typos in this file.
By default, compile with -O3 -g3. The Makefile comments on how some
distributions might need to use -O2 -g or -O -g.
features:
* Rich set of number classes
* Elementary, logical, transcendental functions
* C++ as implementation language brings efficiency, type safety,
and algebraic syntax
* Memory efficiency
* Speed efficiency
* Interoperability
* Fixed bug 135064. Changing the sign of a large number (when "Show
Thousands Separator" is on and this is not the C locale), no longer
corrupts the display.
My special thanks to Miloslav Trmac for fixing this problem in a
really elegant manner. This is an important change to the way that
numbers are displayed.
The fix involved resetting all the internal calculations in gcalctool
to be done in the C locale, and then localized (radix character and
thousands separator), just before the number is displayed.
I'd appreciate it if everybody could give this a good thrashing in as
many locales as possible to make sure that it's fully working.
* Fixed bug 135328. Added Shift-slash as a recognized alternate for the
divide operation (needed by the be-latin1 keyboard layout).
Changes since 4.3.44:
* Fixed bug 134540. gcalctool now correctly handles the digits
0123456789 which are shifted about the accented characters on a
French keyboard.
* Fixed bug 134484. Doing "3 x! x! x!" no longer causes gcalctool to
lock up.
Overview of changes in gcalctool 4.3.44
* Fixed bug 133764.
- "Gradients" was still appearing when in Scientific mode (instead of
"Gradians".
- In the x^2 button, the 2 should not be italicized.
- In the x^2 and x^y buttons, the x should be italicized.
- The View Thousands Separator menu item should not remain active during
an error condition.
Overview of changes in gcalctool 4.3.43
* Fixed bug 133590. The About box now lists the documenters of gcalctool.
Overview of changes in gcalctool 4.3.42
* Fixed bug 132582. Adjusted the gcalctool titlebar to use a hyphen
(rather than square brackets) to show what mode the user is currently
using.
Overview of changes in gcalctool 4.3.41
* Fixed bug 132570. Adjusted UI to use spacing to delineate groups rather
than frames (per the HIG).
Overview of changes in gcalctool 4.3.40
* Fixed bug 131571. User defined constants are now read/written out (as
gconf resources) with no thousands separator and with a radix of ".".
Overview of changes in gcalctool 4.3.39
* Fixed bug 131594. The "Insert ASCII View..." menu item and the
"Insert" button no longer use the GTK_STOCK_CONVERT stock icon
(as per the HIG).
configure tries to determine if -lm is needed by compiling
int main() {
(void)sin(0.0);
; return 0; }
and see if linking fails. But gcc3.3 optimizes that to just "return 0;"
Changes since version 1.11.7:
1.12.1 2003-12-01
Corrected omission of file "lib/udunits.inc.in" from
distribution.
Ported to Mac OSX (I hope).
1.12.0 2003-08-29
Corrected definition of "apothecary_pound".
Corrected documentation of "common_year" in the units
database.
Modified the syntax of a couple of statements in
file "src/utlib.c" to accomodate the, apparently,
non-conforming C/SX C compiler on NEC SX-4 SUPER-UX
systems, which can't dereference a pointer returned by a
function.
Added "mi" to database as alias for "mile".
Corrected documentation on origin-shift in unit
specifications.
Ported to systems with sizeof(double) <
sizeof(DOUBLE PRECISION).
Added determination of Fortran type equivalent to C pointer
to configuration script.
Added utIsInit() and utisopen() to the API.
Corrected Hertz symbol: changed from "hz" to "Hz".
Corrected reference to udunits(3) library in udunitsperl(1)
manual-page.
Added support for Absoft Fortran.
2.4.3 --> 2.4.4:
----------------
Bug fixes:
NetCDF error messages should now be correct. No more "unknown errors"!
2.4 --> 2.4.1:
--------------
Bug fixes:
- Scientific.MPI did not contain the _C_API object needed by C extension
modules that call MPI.
- The arguments to the receiveString method are now optional, as documented.
2.3.3 --> 2.4:
--------------
Bug fixes:
- Memory function calculation in Scientific.Signals.Models
returned a wrong value for the first point.
- Some indexing problems in Scientific.IO.NetCDF were fixed;
none of them caused wrong results, just unjustified exceptions.
2.3.2 --> 2.3.3:
----------------
Bug fixes:
- The method divide() assumed the wrong coefficient order.
New features:
- Module Scientific.Signals.Models
- Class Scientific.Statistics.Histogram.WeightedHistogram
2.3.1 --> 2.3.2:
----------------
Incompatible changes:
- Specification of routines in Scientific.BSP changed.
2.2 --> 2.3.1
-------------
New features:
- New modules Scientific.BSP and Scientific.BSP.IO
Changes since 4.3.35:
* Fixed bug 131059. Fix to allow the thousands separator to be used
correctly with the creation of the initial constant values in the
Russian locale.
* Fixed bug 130078. Changed the y^x button label to x^y (really x
superscript y). Adjusted the tooltip message to "Raise displayed
value to the power of y".
* Fixed bug 130282 again. Needed to remove the "|| tsep[0] == '\0'" check
from the get_tsep() routine in order to get the thousands separator to
correctly work in all locales. Note that there is no thousands separator
now for the C locale.
Changes sinec 4.2.83:
o Many bug fixes.
o Further work on enhancement request 125873. As the user is typing
in input, the display will now show the thousands separator
(if checked).
o On a Solaris keyboard, Shift-"^" is
GDK_asciicircum/GDK_SHIFT_MASK. This keyval/state pair needed to
be added to the Xor recognized mappings.
o gcalctool can now optionally show the thousands seperator
character for fixed point numbers displayed in the decimal
base. Same for numbers stored in the memory registers. There is
a new "Show Thousands Separator" menu entry in the View menu to
set this.
o Online help updated.
o Updated the gcalctool man pages.
o Install "gnome-calculator" symlink as reqested by Glynn Foster.
Pkgsrc fixes:
1) Install trans directory.
Official bug fixes since bug fix 4:
1) A wrong return format for `IsomorphicSubgroups' applied to cyclic groups.
2) A wrong `true' result of `IsSubset' for certain algebras.
3) Segmentation faults and other strange behaviour when assigning
finite field elements of different characteristics into compressed vectors
4) A bug in the subgroup conjugation test for permutation groups that are
not subgroups.
5) A missing method for `BaseOrthogonalSpaceMat'.
6) A missing `Set' call in the construction of the global variable
`AUTOLOAD_PACKAGES'.
7) A wrong display string of the numerator in rational functions returned
by `MolienSeries' (in the case that the constant term of this numerator
is zero).
8) A strange behaviour of `Intersection' for the case that a strictly
sorted list is the unique entry of the list that is given as the
argument; in this situation, this entry itself was returned instead of
a shallow copy.
9) An error in the basis of a product space of algebras.
10) Possibly wrong result of `Centre' for pc groups.
11) Possibly wrong result of `DirectSumDecomposition' for matrix Lie algebras.
12) An error in `LieNormalizer', `LieCentralizer' for zero subspaces.
13) An error in the computation of matrices of adjoint modules.
14) A strange error message when constructing the simple Lie algebra of type
B1.
15) An error in `ModuleByRestriction'.
16) An error in `IrrBaumClausen' for the trivial group.
17) An error with vector space bases of row spaces over fields which
neither are prime fields nor contain all entries of the vectors.
18) An error with `IsMonomial', when it uses the function
`TestMonomialFromLattice' (i.e., in hard cases, likely for characters of
nonsolvable groups).
Errors 1, 3, 5, 6, 9, 12, 13, 14 and 16 may cause GAP to stop with an error
message, errors 2, 4, 7, 10, 17, and 18 may cause wrong results.
- teTeX 1.0.x packages have been moved to teTeX1* directories.
- teTeX 2.0.x packages have been moved to teTeX* directories and their base
name has been changed to teTeX (instead of teTeX2).
This python module implements constants and functions for working with IEEE754
double-precision special values. It provides constants for Not-a-Number (NaN),
Positive Infinity (Inf), and Negative Infinity (-Inf), as well as functions to
test for these values.
The code is implemented in pure python by taking advantage of the 'struct'
standard module. Care has been taken to generate proper results on both
big-endian and little-endian machines. Some efficiency could be gained by
translating the core routines into C.
--------------------
Fixed bugs
o There was several memory leaks due to bad management of Python
refcounts.
o There was wrong synchronization of the thread running the R event
loop and the calls to the R interpreter.
o There was a bug with signal handling which caused segfault when
'Ctrl-C' were pressed.
o The R event loop shouldn't be run in non-interactive sessions.
o The i/o Python callbacks were wrong synchronized with the R event
loop thread.
New features
o Added a rpy_version symbol.
* Emacs fixes
* The ++ and -- operators now work for indexed matrices
* The built-in variable argv is now a list of strings instead of a string vec.
* The new built-in variable `crash_dumps_octave_core' controls
whether Octave writes user variables to the file `octave-core'
when it crashes or is killed by a signal.
* Loops of the form `for i = STRING ... endfor' are now allowed.
* It is now possible to set the iteration limit for lsode using
lsode_options ("step limit", N).
* New functions
* The save command now accepts the option -append to save the
variables at the end of the file, leaving the existing contents.
* New command-line option --no-history (also available using the
single character option -H) inhibits saving command history.
* The built-in variable `PWD' has been removed.
* New operators.
* New built-in data types.
* Commas in global statements are no longer special.
* It is now possible to declare static variables that retain their
values across function calls.
More --> See "NEWS"
Changes since 1.6.2
Many new features.
Language changes.
Additional standard package 'mle'.
Some old features deprecated or defunct.
Many many bug fixes.
Update to 1.8.0 from Ray Brownrigg in private communication, cleaned up
and updated to 1.8.1 by me.
Changes:
* Version 3.0, 10-15-03
- add "options" and "stat" argument for the driver routines
DGSSV/DGSSVX. This interface is more user-friendly and flexible.
- add more examples in EXAMPLE/
- add a "symmetric mode" with better performance when the matrix is
symmetric, or diagonal dominant, or positive definite, or nearly so.
Also, make 'test' target do something.
want to patch all this), so bail out if we detect a gcc-3 compiler.
Suggest octave-current instead. Closes my own PR pkg/21874.
While here, allow dependecy on either ghostscript or ghostscript-nox11.
From release announcemenet:
After a lengthy QA process, PHP 4.3.4 is finally out!
This is a medium size maintenance release, with a fair number of bug fixes.
All users are encouraged to upgrade to 4.3.4.
Bugfix release
PHP 4.3.4 contains, among others, following important fixes, additions
and improvements:
* Fixed disk_total_space() and disk_free_space() under FreeBSD.
* Fixed FastCGI support on Win32.
* Fixed FastCGI being unable to bind to a specific IP.
* Fixed several bugs in mail() implementation on win32.
* Fixed crashes in a number of functions.
* Fixed compile failure on MacOSX 10.3 Panther.
* Over 60 various bug fixes!
For full list of changes in PHP 4.3.4, see ChangeLog:
http://www.php.net/ChangeLog-4.php#4.3.4
curses.buildlink2.mk. This was wrong because we _really_ do want to
express that we want _n_curses when we include the buildlink2.mk file.
We should have a better way to say that the NetBSD curses doesn't
quite work well enough. In fact, it's far better to depend on ncurses
by default, and exceptionally note when it's okay to use NetBSD curses
for specific packages. We will look into this again in the future.
changes:
Version 23.1 August 2003
[unreported] fixed assignment to matrix.
[unreported] added check so that non-character array cannot be interpreted
as a character buffer.
[unreported] array([m,m,m]) when m is a rank-0 array now works as expected.
Patches:
[ 782001 ] Add true and floor divide to MA
[ 781215 ] true_divide bug
[ 776991 ] multiplying small UINT values gives random results
[ 776467 ] remove string exceptions
Inspired by FreeBSD "ports".
Fix the PLISTs accordingly.
Also, while at it, remove now obsolete compileall.py calls in post-install
targets and insure that extension.mk is in included before builinlinks of
other Python modules.
Discussed with/ok'ed by drochner@.
USE_GCC2 or USE_GCC3 where appropriate.
the functionality of the old gcc.buildlink2.mk has been rolled into
compiler.mk now, which is automatically used.
more changes to come later...
Some highlights of changes since 4.2.3:
* PCRE updated to 4.3, GD to 2.0.15
* improved Apache2 support
* much improved stream & URL wrapper support, output compression support
* added CLI (Command Line Interface) SAPI
* debug_backtrace() backported from ZendEngine2
* faster build system
* huge number of other bug fixes and improvements
Packaging changes:
* 'pcre', 'xml', and 'session' modules folded back into main package -
'pcre' and 'xml' is required by PEAR, and 'session' is just too essential
to be separate
* 'gd' module now uses bundled PHP GD library, which is better integrated
* PHP modules use shared distinfo when possible to ease future PHP updates
* ${PREFIX}/bin/php is now CLI version, ${PREFIX}/libexec/cgi-big/php
remains CGI version
Clarify that the internal hash as well as the hash builtin
function used by calc, while based on the Fowler/Noll/Vo
hash is NOT an FNV hash.
Made slight performance improvements to calc by an optimization of how
calc's internal hash is computed. The "make chk" regression test
runs about 1.5% faster (when compiled with -O3 on an AMD Athlon)
NO_HASH_CPU_OPTIMIZATION is not defined. Calc's internal hash values
have not changed. By default, NO_HASH_CPU_OPTIMIZATION is NOT defined
and the slightly faster expression is used.
A slight modification of what was known as the "calc new stardard"
configuration (calc -n or config("all", "newstd")) is now the default
calc configuration. The flag:
calc -O
was added to get the old classic calc configuration. The flag command
line flag, -n, now does nothing. Use of -n is deprecated and may go
away / be used for something else in the future.
The following table gives the summary of these changes:
pre v2.11.8 v2.11.8
default pre v2.11.8 -O & oldstd v2.11.8
and oldstd -n & newstd classic cfg default
--------------------------------------------------------
epsilon 1e-20 1e-10 1e-20 1e-20
quo 2 2 2 2
outround 2 24 2 24
leadzero 0 1 0 1
fullzero 0 1 0 0
prompt > ; > ;
more >> ;; >> ;;
With the exception of epsilon being 1e-20, and fullzero being unset,
the new default calc config is like it was (pre-2.11.8) with calc -n /
config("all", "newstd").
The new default config is the old classic config with outround being
24, leadzero being set, and the prompts being ;'s.
* What is new in gsl-1.4:
** Added cumulative distribution functions and their inverses for the
continuous random distributions including: gaussian, lognormal, gamma,
beta, cauchy, laplace, chisq, exponential, gumbel, weibull,
F-distribution, t-distribution, logistic, pareto and rayleigh.
** Added faster binomial random variates using the TPE rejection
algorithm, in the function gsl_randist_binomial_tpe.
** Added new functions gsl_rng_fwrite and gsl_rnd_fread for storing
the state of random number generators in a file.
** Added a new function gsl_combination_memcpy()
** Corrected values of electrical constants in CGS units. To take
account of different electrical systems of units the values are now
prefixed by GSL_CONST_MKSA (for the SI Metre, Kilogram, Second, Ampere
system) or GSL_CONST_CGSM (for the Centimetre, Gram, Second, Magnetic
system with the Gauss as the fundamental unit of magnetic field
strength). The previous GSL_CONST_MKS and GSL_CONST_CGS prefixes have
been removed, as have the permeability and permittivity constants in
the CGS system since this uses different defining equations.
** Fixed bugs in the random number generators gsl_rng_fishman18,
gsl_rng_fishman2x, and gsl_rng_knuthran2 which caused them to return
incorrect results. Minor corrections were made to the parameters in
the other Knuth generators borosh13, coveyou, fishman20, lecuyer21,
and waterman14.
** Fixed a missing transpose bug in the gsl_linalg_QR_QRsolve
and gsl_linalg_QRPT_QRsolve routines which were computing the
solution to Q^T R x = b instead of Q R x = b.
** Fixed gsl_sf_gammainv to return zero instead of a domain
error for arguments corresponding to singularities in gamma.
** Fixed a bug in the simplex minimization algorithm which
caused it to fail to find the second highest point correctly
when searching the set of simplex points.
** Fixed a bug in the conjugate gradient minimizers conjugate_pr,
conjugate_fr and vector_bgfs which caused the search
directions to be updated incorrectly.
** Fixed a bug in gsl_sf_psi_1_int(1) which caused it to
return the incorrect sign for psi(1,1).
** Fixed the simulated annealing routine gsl_siman_solve to use the
parameter iters_fixed_T for the number of iterations at fixed
temperature instead of n_tries.
** Fixed a bug in gsl_combination_valid which caused it to return the
incorrect status.
** Fixed a bug in gsl_permutation_canonical_to_linear which caused the
output to always be zero, and the input permutation to be incorrectly
replaced by the output.
** Fixed a bug is gsl_ran_discrete which could cause uninitialised
data to be returned for some distributions.
** Fixed the dependencies for gsl_chebyshev.h to include gsl_math.h.
** Fixed a bug in gsl_complex_arccsc_real which caused it to return
the incorrect sign for the imaginary part when -1<x<0.
** Fixed a bug in the QAWC Cauchy integration routine which could
allow the singularity to fall on an interval boundary, leading to
division by zero.
** Improved gsl_sf_gamma_inc_P(a,x) to avoid a domain error for x<<a
when a>10.
** Improved the accuracy of gsl_sf_coupling_3j for large arguments.
** Improved the performance of gsl_sf_choose(m,n) by separating the
calculations for small and large arguments.
** On platforms without IEEE comparisons gsl_{isnan,isinf,finite} will
fall back to the system versions of isnan, isinf and finite if
available.
** gsl_linalg_householder_hv now uses BLAS routines internally
** The script configure.in is now compatible with autoconf-2.50 and
later.
** Reduced the memory usage of the multifit algorithms from MxM to MxN
for large M by performing the QR decomposition of the Jacobian
in-place.
** IEEE modes now use the C99 fenv.h functions when platform spectific
functions are not available.
V1.30 (March 2002) to V1.40 (June 2003)
- simultaneous assignments
(x1,x2,...,xn) := (a1,a2,...,an)
- new functions
divide, bit_count,
set_workdir, get_workdir, gmtime,
stack2string, stack_arraypush, string_scan,
realloc, binsearch
- some bug fixes (regarding qs_factorize and
handling of special variable and value arguments
in user defined functions)
