Changes:
- Use mkstemp() for temp file in edit command. Put temp file in "/tmp",
instead of current directory.
- Created my_strlcpy(), since Linux doesn't have strlcpy().
- Made everything declared static that should be.
- Prime Number Tools cleaned up and compiled with "-Wall".
- Made them all successfully compile with g++.
- A bug fix for HTML mode.
- Completely removed sensitivity command and simplified parsing code.
- Fixed minor bug in fraction format display routine.
Overview of changes in gcalctool 5.5.41
* Version for GNOME 2.10.0 (final) call for tarballs.
----
Overview of changes in gcalctool 5.5.40
* Version for GNOME 2.10.0 (rc1) call for tarballs.
----
Overview of changes in gcalctool 5.5.39
* Partial fix for bug 168385. "F+F=" wasn't working in arithmetic operator
precedence mode.
----
Overview of changes in gcalctool 5.5.38
* Changes to make the trigonometric code work with hyp and inv properly.
----
Overview of changes in gcalctool 5.5.37
* Left and Right shift operations weren't working correctly. Menu
name comparison in mb_proc() in gtk.c was incorrect.
----
Overview of changes in gcalctool 5.5.36
* The fix for bug #162998 was incorrect when in Arithmetic Precedence mode.
* Exch functionality was incorrect in Arithmetic Precedence mode.
----
Overview of changes in gcalctool 5.5.35
* Fixed bug 167124. One of the entries in the words[] array in
syntax_translation.c was accidentally commented out.
----
Overview of changes in gcalctool 5.5.34
* Fixed bug 166634. Call to create the About box was missing a
NULL terminator.
----
Overview of changes in gcalctool 5.5.33
* Adjusted all copyright messages to be 2005.
* Version for GNOME 2.10.0 (beta2) call for tarballs.
----
Overview of changes in gcalctool 5.5.32
* Fixed bug #165522. Made the status bar non-resizable.
----
Overview of changes in gcalctool 5.5.31
* Various code cleanups. It moves some declarations of static functions
to the corresponding .c files. It also cuts down the number of
included files, moving some includes to where they are needed from
calctool.h.
----
Overview of changes in gcalctool 5.5.30
* Adjusted the About dialog to use the latest GtkAbout widget.
----
Overview of changes in gcalctool 5.5.29
* Version for GNOME 2.10.0 (beta1) call for tarballs.
----
Overview of changes in gcalctool 5.5.28
* Fixed bug #162998. The "menu" buttons in gcalctool now responds
to the "clicked" event rather than the "button-press-event" to
allow gok, the on-screen keyboard to function properly.
* Fixed bug #159957. Input number in FIX mode now make mimimal changes
to (i.e append a single character) the display rather than completely
delete and insert it. The is helpful to assistive technologies like
gnopernicus.
----
Overview of changes in gcalctool 5.5.27
* Fixed bug #164184. Adjusted location of comments in syntax_translation.c
so that the L10N message translation software isn't confused.
----
Overview of changes in gcalctool 5.5.26
* Fixed bug #163468. Pressing "(<Del>)" would cause gcalctool to crash.
----
Overview of changes in gcalctool 5.5.25
* Version for GNOME 2.9.4 call for tarballs.
----
Overview of changes in gcalctool 5.5.24
* Couldn't use the "#" key to select constants as this symbol was
a lower-case symbol on his keyboard.
* Powers that were a multiple of 30 (10 as well as 3) would not
display correctly, having a 4 digit significand. Further testing
showed that without the fix the following powers had a 4 digit
significand: 27,30, 57,60, 90, 96 (and every 3 from here on).
----
Overview of changes in gcalctool 5.5.23
* Fixed bug #161976. configure.in now checks to make sure that atk >= 1.5
----
Overview of changes in gcalctool 5.5.22
* Fixed bug #155101. Made the gcalctool window resizable.
----
Overview of changes in gcalctool 5.5.21
* Version for GNOME 2.9.3 call for tarballs.
----
Overview of changes in gcalctool 5.5.20
* Fixed bug #160929. Removed the "Mode" word from each of the different
mode menu items in the gcalctool View menu.
----
Overview of changes in gcalctool 5.5.19
* Fixed bug #157701. gcalctool can now use real multiplication and division
symbols instead of "*" and "/".
----
Overview of changes in gcalctool 5.5.18
* Fixed bug #160088. Adjusted the titlebar text so that it doesn't include
the word "Mode" which was getting chopped when the calculator was in
Basic mode.
----
Overview of changes in gcalctool 5.5.17
* Fixed bug #157962. Added in a set of three colors for gcalctool that
will work with the default theme. To use them, the "gcalctoolrc" file
supplied with the source distribution needs to be moved to ~/.gcalctoolrc
----
Overview of changes in gcalctool 5.5.16
* Generated gcalctool tarball for the GNOME 2.9.2 release.
----
Overview of changes in gcalctool 5.5.15
* Adjusted the scroll_right() routine to only scroll right if the
horizontal scroll bar is visible.
----
Overview of changes in gcalctool 5.5.14
* Fixed bug #158357. Removed the -DGNOME_DISABLE_DEPRECATED and the
-DGTK_DISABLE_DEPRECATED flags from the INCLUDES definition in
.../gcalctool/Makefile.am for now, to allow it to successfully build
against the latest GNOME/Gtk+.
* Slight adjustment to correct check the Advanced mode menu item when
the user had previously terminated gcalctool in that mode.
----
Overview of changes in gcalctool 5.5.13
* Changed the names of the gconf mode type key so that there is backward
compatibility.
* Reset the initial default mode to BASIC.
----
Overview of changes in gcalctool 5.5.12
* Fixed bug #157757. The horizontal scrollbar for the display area will
now only be visible when needed. It will also display the rightmost part
of the display as you are entering new input.
----
Overview of changes in gcalctool 5.5.11
* Added in a very basic mode for gcalctool. Rather than calling this a
"Light" mode, the existing "Basic" mode is now renamed to "Advanced"
and the new basic mode is called "Basic".
----
Overview of changes in gcalctool 5.5.10
* Fix for bug #157021. X^Y now works for -X with integer Y. -X with
non-integer Y will now display a descriptibe error in the status bar.
----
Overview of changes in gcalctool 5.5.9
* Generated gcalctool tarball for first GNOME 2.9 release.
----
Overview of changes in gcalctool 5.5.8
* Slight adjustment to bug #152301. The tooltip message had been tweaked to
"Show the About Gcalctool dialog".
----
Overview of changes in gcalctool 5.5.7
* Fixed bug #152301. Adjusted the tooltip message for Help->About... from
"Show about help" to "Show the about dialog", to make it more understandable.
----
Overview of changes in gcalctool 5.5.6
* Fixed bug #153726. Gcalctool will now build correctly when using the
-fno-common option for gcc (3.3.4).
----
Overview of changes in gcalctool 5.5.5
* Numerical answer localization and thousand separator support for
arithmetic precedence mode.
* If the accuracy was changed via one of the ten preset menu items in the
Acc menu, the tooltip wasn't getting correct updated with the new accuracy.
----
Overview of changes in gcalctool 5.5.4
* Further changes to the "Set Precision" popup (rfe #147803):
- "Significant places" not "Significant Places".
- Change "Set Precision..." to be "Other (12) ..." (where "(12)" is the
current precision, and be in the same radio box group as the other menu
items. Check the appropriate menu entry.
- Adjust the Acc button tooltip to show the current accuracy too.
- When the user attempts to enter an invalid entry in the Set Precision
popup, an error message is shown in the main gcalctool status bar.
- the other Acc radio button items (0-9) need to have mnemonics (_0, _1 ...).
* Fixed bug #152790. Most of the changes needed were already in CVS; there
was one compiler warning that still needed to be removed.
* Storing the display value to a memory register was not updating the memory
register window, if it was visible.
----
Overview of changes in gcalctool 5.5.3
* Added in functionality to set the number of significant places > 9.
There is a new menu item under the Acc menu called "Set Precision..."
that brings up a small popup allowing you to set precision between 0
and MAXACC significant places (where MAXACC is currently 30).
----
Overview of changes in gcalctool 5.5.2
* Fixed up numerous warnings generated by running lint on the gcalctool code.
Removed unused variables and functions. Declared routines statically where
possible.
----
Overview of changes in gcalctool 5.5.1
* Fixed bug 148581. Changes to get latest gcalctool compiling on a BSD system.
----
Overview of changes in gcalctool 5.5.0
* Major new release of gcalctool. Incorporates an arithmetic operator
precedence mode (currently the default while it's been debugged). To get
the old style, uncheck View->Use Arithmetic Precedence from the menu bar.
* The "(" and ")" buttons have moved from the scientific mode into the
Basic mode so that they are now available to users who need them but not
everything that the Scientific mode provides.
* There is a footer message area that provides feedback.
* The display area now has an horizontal scrollbar.
* Typing in the keyboard shortcuts for the various gcalctool buttons
autocompletes their entry into the display area.
* Tooltips show button's shortcut.
and ${CFLAGS}. This fixes the build of net/djbdns, as well as any
other of these packages passing down PKG_SYSCONFDIR via CFLAGS, as
well as being more generally correct for arbitrary user-defined
CFLAGS. Suggested by jlam.
For consistency across djbware in pkgsrc:
* In math/djbfft's and sysutils/daemontools's do-configure targets,
remove leading @ from ${ECHO} lines; from the former, also remove
unneeded single quotes from one such line.
* Rename net/publicfile's pre-build and sysutils/service-config's
post-patch targets to do-configure.
* In sysutils/checkpassword's do-configure target, reorder creation
of conf-cc, conf-ld, and conf-home.
All of the affected packages have been verified to compile.
XXX These packages probably have enough build goo in common to
XXX warrant an mk/djbware.mk. I'll investigate this post-freeze.
Changes since 11.7:
- Increased precision and optimization of code command.
- Prepared code for future improvement to integrate command.
- Integrate (a*constant^x) for x has been added.
- Allow differentiation and integration of -constant^x.
- Disabled power factoring of k^x*k^y, where k is a constant.
This fixes a small bug.
- Reduced stack usage by almost 2/3.
- Reduced stack usage another large amount to the smallest it can be.
- Laplace (a*constant^x) and (a/constant^x) for x has been added.
- Allow integration of (a/constant^x) for x.
- Moved all numeric GCD routines to "gcd.c".
- Improved limit command for "limit x inf".
- Halved the default equation size and doubled the number of equation
spaces to 40.
- Fixed bug that caused infinite loop for "2^(.5*n)" and incorrect answer
for "-1^(2*n)".
- Allow edit command to work with no equations defined.
- Allow typing "abort" or "quit" during pause command.
- They both abort the current operation.
- Removed usage displayed every time a command fails.
- Simplify command once again prefers 1/(constant^n) instead of (1/constant)^n.
- Added "set finance" option.
- Some improvement to simplification of constants.
- Improved "factor number" command. Made self-testing and allow
multiple integers.
- Strip trailing newlines from string returned by getstring().
- Re-enabled power factoring of k^x*k^y, where k is a constant.
- The bug it had caused just disappeared.
- This is because a lot of changes have been made.
- Some speed up in polynomial factoring made by ignoring anything that
is not a polynomial in any variable.
- Improved denominator rationalizer.
- Disabled power factoring of -1^x*-1^y. This time it is done right.
- Added new function debug_string(). It simplifies the debugging code.
- Simplified pf_sub() (the polynomial factoring subroutine).
- Fixed up Prime Number Tools man pages.
- Improved matho-twin.
- Added simplification of roots like 12^.5 to 2*(3^.5) in simplify command.
- Code cleanup.
- Improved the results of solving by not always factoring constants.
- Added semicolon (";") to end of each line when using "list export".
- Some improvement made to the simplify command.
- More tests added to test suite.
- Fixed bug in patch_root_div().
- Fixed infinity/nan bug in uf_repeat().
PyTables 0.9.1:
This release is mainly a maintenance version. In it, some bugs has been
fixed and a few improvements has been made. One important thing is that
chunk sizes in EArrays has been re-tuned to get much better performance
and compression rations. Besides, it has been tested against the latest
Python 2.4 and all test units seems to pass fine.
PyTables 0.9:
In this release you will find a series of quite exciting new features,
the most important being new indexing capabilities, in-kernel selections,
support for complex datatypes and the possibility to modify values in both
tables and arrays (yeah, finally :).
Support Python 2.4
1.1.1 changes:
numarray-1.1.1 is a bugfix release to numarray-1.1. Notable bugs
fixed include memory leaks in matrixmultiply and comparison ufuncs.
1.1 changes:
I. ENHANCEMENTS
CharArray eval() sped up
Document memmap.py (memory mapping)
Unsigned int type support limited
Add kroenecker product
II. BUGS FIXED / CLOSED
max.reduce of byteswapped array
numeric compatibility byteoffset
matrixmultiply (a,b) leaves b transposed
random_array.randint exceeds boundaries
buffer not aligned on 8 byte boundary (Windows-98 broken)
Object Array repr for >1000 elements
Invalid sequences errors
Segfault in array element deletion
Incorrect handling of overlapping assignments in Numarray
Weirdness with 'new' method
searchsorted bug and fix
randint bug fix patch
a.is_c_array() mixed int/bool results
argsort of string arrays
III. CAUTIONS
1. This release is binary incompatible with numarray-1.0. Writers of
C-extensions which directly reference the byteoffset field of the
PyArrayObject should be aware that the data pointer is now the sum of
byteoffset and the buffer base pointer. All C extensions which use
the numarray C-API must be recompiled. This incompatibility was an
unfortunate consequence of the fix for "numeric compatibility
byteoffset".
* Show base (if not decimal, hexadecimal or non-standard) as subscript
* Use singular instead of abbreviated name for completion
* Fix "Convert result" button
* Use the last operator for leftover stack values
* Fix RPN crash and don't add lonely stack value to itself
* Workaround freeze with message dialog without buttons
Changes:
Another minor adjustment to simplify command to make everything work
just right.
Trap "derivative x" of (negative constant or zero)^x.
Solve "(constant)^y=constant" for y.
Moved all html files and the man page to subdirectory "doc".
Code cleanup.
Improved modulus solving.
Trap "x%0". Simplify "x%inf".
Fix fraction list display of -0.
Change for better simplification.
Fix for division by roots of fractions.
Collection.
Set::Infinite is a Set Theory module for infinite sets.
It works with reals, integers, and objects.
When it is used dates, this module provides schedule
checks (intersections), unions, and infinite recurrences.
python*-pth packages into meta-packages which will install the non-pth
packages. Bump PKGREVISIONs on the non-pth versions to propagate the
thread change, but leave the *-pth versions untouched to not affect
existing installations.
Sync all PYTHON_VERSIONS_AFFECTED lines in package Makefiles.
- bug fix in handling of -pi (previous version changed constant pi)
- bug fix in division of integer vectors with negative coefficients
- for loop can now handle >= 2**32 iterations
- fixed bug which occurred in V 1.50 while parsing certain
parenthized expressions
