1 2008-01-dd, version 1.2.0
2 =========================
7 * Save big amounts of memory in computation of some functions and constants by:
8 - Avoiding pre-computation of series terms and instead computing them in a
10 - Avoiding computation with excess precision in binary splitting algorithm
11 by coercion of intermediate integer values to floating-point as early as
14 Implementation changes
15 ----------------------
17 * Added support for huge numbers:
18 - intC used for all counter loops.
19 - intE is now a 64-bit type on all 64-bit platforms and even on selected
22 * CLN now uses C++ exceptions for error handling. The cl_abort() hook is not
23 supported any more. Please refer to the documentation to learn about existing
26 * Fixed a bug on i386 where comparing Z/2Z ring zeros returnd random results.
28 * Removed cl_boolean. Use built-in C++ bool instead.
33 * Dropped the cln-config script. Please use pkg-config instead.
35 * Updated infrastructure to that of libtool-1.5.24.
37 * Changed encoding of all files to UTF-8.
39 * Fix compilation issues with GCC-4.3.
42 2006-08-08, version 1.1.13
43 ==========================
45 * Compilation fixes for 64-bit brokenness introduced in last release.
48 2006-08-06, version 1.1.12
49 ==========================
51 Implementation changes
52 ----------------------
54 * Fix rare assertion when printing quite large floats.
59 * Compilation fixes for several platforms: *BSD, Intel Mac, and MinGW.
62 2005-11-23, version 1.1.11
63 ==========================
68 * Considerably improved performance of number input.
71 2005-10-22, version 1.1.10
72 ==========================
74 Implementation changes
75 ----------------------
77 * Removed the vptr of modular integer and univariate polynomial
78 classes in order to fix some crashes in that sector of CLN. Code
79 using this needs to be recompiled.
81 * Many more operator signatures make the integration of CLN types with
87 * Several compilation fixes.
89 * Made it possible to cross-compile CLN.
92 2004-11-03, version 1.1.9
93 =========================
98 * Input of numbers in bases 2, 4, 8, 16 and 32 is now done in linear bit
99 complexity as opposed to O(N^2). Useful for all kinds of persistency.
101 Implementation changes
102 ----------------------
104 * Fixed several bugs in the integer input and output routines that could be
105 blamed for random crashes in the following cases: output in base 32 for
106 quite large numbers, input in base 2 for fixnums and input in base 3 for
107 fixnums on a 64 bit target.
109 * Fixed crash when radix specifiers were used in input streams.
111 * Speed up on x86_64 and ia64 by adding some inline assembly.
116 * Fixes for compilation on MacOS X and little endian Mips.
119 2004-07-01, version 1.1.8
120 =========================
122 Implementation changes
123 ----------------------
125 * Fix bug in log(a,b) when b is 1/n.
127 * Fix crash in shared library on AMD64.
132 * Fixes for some language conformance issues.
135 2004-05-02, version 1.1.7
136 =========================
138 Implementation changes
139 ----------------------
141 * Fixed two corner case bugs in cl_LF multiplication and sqrt.
143 * Workaround GCC 3.[123] compiler bug.
145 * Parallel build should work reliably now.
148 2004-01-01, version 1.1.6
149 =========================
155 `sintL ldegree(const cl_UP& x)'.
156 (Suggested by Munagala Ramanath.)
158 Implementation changes
159 ----------------------
161 * Fixed a bug in polynomial subtraction.
162 (Reported by Munagala Ramanath.)
167 * Provide a package control file cln.pc.
169 * Updated infrastructure to autoconf-2.57.
171 * Fixed compilation issues with GCC-3.4 prereleases and on Mac OSX.
174 2002-05-28, version 1.1.5
175 =========================
180 * Fix compilation issues with GCC-3.1.
183 2002-01-04, version 1.1.4
184 =========================
189 * Fix compilation issues with GMP-4.0 and on IBM s390.
191 * Updated infrastructure to autoconf-2.52.
194 2001-11-05, version 1.1.3
195 =========================
197 Implementation changes
198 ----------------------
200 * Fix a computation error of sin and sinh for small arguments.
201 (Reported by Christopher Kennedy.)
203 * Module ordering works with gcc-3.0 -fuse-cxa-atexit now.
205 * -DNO_ASM works properly on Sparc.
208 2001-07-25, version 1.1.2
210 Implementation changes
211 ----------------------
213 * Minor cleanups and portability issues.
214 (No need to upgrade if you have been happy so far.)
217 2001-05-31, version 1.1.1
218 =========================
220 Implementation changes
221 ----------------------
223 * Minor cleanups for GCC 3.0 and libstdc++-v3.
225 * Fixes several little bugs and portability issues.
228 2000-12-14, version 1.1.0
229 =========================
234 * ISO/IEC 14882 fashion adjustments:
235 Put everything into namespace cln. All fundamental data types still
236 carry their old names. Other non-macro identifiers are now written as
237 cln::foo instead of cl_foo, except where the cl_ comes from a data type
238 (as in cl_I_to_int()). Headers are installed into a separate
239 directory, #include <cln/foo.h> instead of <cl_foo.h>. Applications
240 must be manually ported to the new style. We apologize for the
243 Implementation changes
244 ----------------------
246 * Removed internal copy of mpn subdirectory tree from libgmp-2.0.2.
247 Configuring with --with-gmp now relies on an installed libgmp library
248 version 3 or above. We recommend libgmp-3.1 or newer.
250 * Adjusted several break-even points to make better use of a faster libgmp and
251 better match present-day CPUs.
253 * Fix several errors for architectures with 64-bit wide words.
254 (Others than Alpha, which worked already.)
256 * Fix compilation errors with current GCC-3.0 snapshots.
261 * Added package tools: script cln-config and autoconf macro AC_PATH_CLN
265 2000-01-13, version 1.0.3
266 =========================
272 `cl_I doublefactorial (uintL n)'.
274 Implementation changes
275 ----------------------
277 * Fix several little configuration errors.
279 * Fix some compilation errors with gcc-2.96 prereleases.
282 1999-09-07, version 1.0.2
283 =========================
289 `cl_I numerator (const cl_RA&)'
290 `cl_I denominator (const cl_RA&)'.
291 Suggested by Richard Kreckel and Sylvain Pion.
293 * New function `cl_equal_hashcode' for the classes cl_N, cl_R, cl_RA, cl_I,
294 cl_F, cl_SF, cl_FF, cl_DF, cl_LF.
295 Suggested by Richard Kreckel.
297 Implementation changes
298 ----------------------
300 * Fix an endless loop when either of the functions
301 `cl_boolean rootp (const cl_RA&, uintL, cl_RA*)',
302 `cl_boolean rootp (const cl_RA&, const cl_I&, cl_RA*)'
303 was called with an integer argument. Thanks to Richard Kreckel.
305 * Fix a bug in the addition and subtraction of rational numbers which could
306 cause unnormalized rational numbers like 3/1 to be created.
309 1999-06-09, version 1.0.1
310 =========================
315 * Speed up the functions `logand', `lognand', `logandc2'.
317 Implementation changes
318 ----------------------
320 * Fix compilation errors with gcc-2.7.2, egcs-1.1.2 and gcc-2.95.
322 * Fix compilation errors on HPPA, MIPS, some versions of DEC Alpha, OpenBSD,
323 and SPARC with LiDIA.
325 * Fix compilation errors on Win32. Thanks to Mumit Khan.
328 1999-01-12, version 1.0
329 =======================
334 * New include files, containing I/O functions:
338 <cl_sfloat_io.h> <cl_ffloat_io.h> <cl_dfloat_io.h> <cl_lfloat_io.h>
341 The file <cl_io.h> now contains only I/O functions for characters and
344 * To access the base ring of a univariate polynomial ring, now use
345 `R->basering()' instead of `R->basering'.
347 * Implement `plusp', `max', `min' for the classes cl_F, cl_SF, cl_FF, cl_DF,
350 * Implement `abs' for the class cl_RA.
352 * Replaced `read_number' with specialized functions `read_complex',
353 `read_real', `read_float', `read_rational', `read_integer'.
355 * Replaced the functions
356 `void fprint (cl_ostream stream, int x)'
357 `void fprint (cl_ostream stream, unsigned int x)'
358 `void fprint (cl_ostream stream, long x)'
359 `void fprint (cl_ostream stream, unsigned long x)'
361 `void fprintdecimal (cl_ostream stream, int x)'
362 `void fprintdecimal (cl_ostream stream, unsigned int x)'
363 `void fprintdecimal (cl_ostream stream, long x)'
364 `void fprintdecimal (cl_ostream stream, unsigned long x)'
369 * The function `cl_I xgcd (const cl_I& a, const cl_I& b, cl_I* u, cl_I* v)'
370 now normalizes the resulting u and v to be of minimum absolute value.
371 (Suggested by the LiDIA group.)
373 * The conversion from string to cl_F, cl_R, cl_N now chooses the format of
374 the resulting float depending on the number of significant digits present
377 * Speed up the functions
378 `cl_R operator* (const cl_RA&, const cl_LF&)'
379 `cl_R operator* (const cl_LF&, const cl_RA&)'
380 `cl_R operator/ (const cl_RA&, const cl_LF&)'
381 `cl_LF operator/ (const cl_LF&, const cl_RA&)'
383 Implementation changes
384 ----------------------
386 * Fix for `cl_I ash (const cl_I& x, const cl_I& n)' when |n| is between
388 (Reported by Thomas Papanikolaou.)
390 * Fixed a memory leak in the long-float truncate function.
391 (Reported by Keith Briggs.)
393 * Fixed a memory leak in the modular integer exponentiation.
395 * Reduced the stack space consumption. The default stack size (typically
396 8 MB on Unix) should now be sufficient in order to run all CLN programs.
398 * Portability fixes for m68k platforms.
399 (Reported and fixed by Roman Hodek.)
404 * Restructured the sources. The subpackages
407 - base + integer + rational
408 - base + integer + rational + float
409 - base + integer + rational + float + real
410 - base + integer + rational + float + real + complex
411 - base + integer + modinteger
412 are now self-contained.
414 * The configuration script can be called with option `--enable-shared',
415 to build CLN as a shared library.
417 * Updated the LiDIA interface.
420 1998-05-07, version 0.98
421 ========================
423 Implementation changes
424 ----------------------
426 * Portability fixes for 64-bit platforms like DEC Alpha.
427 (Reported by John Cremona and Thomas Papanikolaou.)
436 * Portability fixes for new ISO/ANSI C++ compilers.
438 * Updated the LiDIA interface for LiDIA version 1.3 and fixed a couple of
445 Implementation changes
446 ----------------------
448 * Portability fix for i686 platforms.
449 (Reported by Michael Pfeifer.)
459 <cl_ring.h>, declaring general rings,
460 <cl_GV.h>, <cl_SV.h>, defining vectors,
461 <cl_univpoly.h>, defining univariate polynomial rings.
466 * Speed up the internal computation of ln(2) and ln(10) by about 20%.
468 * Speed up the computation of exp (for 1000 digits or more) by about 10%.
470 Implementation changes
471 ----------------------
473 * Portability fix for 64-bit platforms like DEC Alpha:
474 Fixed a bug: endless loop in `sqrt' for large precision.
475 (Reported by Blake Jones.)
477 * Portability fix for RS/6000 platforms.
486 * Speed up the computation of the elementary transcendental functions:
487 exp, sinh, cosh, asinh, atanh (for 100 digits) by about 30%,
488 log (for 100 digits) by about 25%,
489 sin, cos (for 1000 or 10000 digits) by about 20%,
490 asin, acos, atan (for 1000 digits) by about 25%.
492 Implementation changes
493 ----------------------
495 * Portability fix for SunOS 4 and other BSD platforms.
504 * Fixed a bug: endless recursion in `cl_boolean sqrtp (const cl_RA&, cl_RA*)'.
506 * Fixed a bug: Only the first 85% of the digits returned by `cl_eulerconst'
507 for moderate precision (< 11370 digits) were correct.
508 (Reported by Thomas Papanikolaou.)
510 Implementation changes
511 ----------------------
513 * Fix static initialization order bug. Requires g++ 2.7.0 or newer
514 (g++ 2.8.0 or newer on HPPA).
516 * New method `debug_print()', for pretty printing of CLN objects, intended
517 to be called from the debugger.
526 * New functions `float cl_float_approx (const cl_R&)',
527 `double cl_double_approx (const cl_R&)'
528 for converting a real number to a C `float' or `double'.
533 * Speed up `cl_zeta' using Cohen-Villegas-Zagier convergence acceleration.
535 Implementation changes
536 ----------------------
538 * New function `cl_print', for pretty printing of CLN objects, intended to be
539 called from the debugger.
541 * Portability fixes for 64-bit platforms like DEC Alpha.
551 <cl_string.h>, defining strings,
552 <cl_symbol.h>, defining symbols,
553 <cl_proplist.h>, defining property lists,
554 <cl_condition.h>, defining conditions/exceptions.
556 * New functions `cl_F cl_catalanconst ()',
557 `cl_F cl_catalanconst (const cl_F&)',
558 `cl_F cl_catalanconst (cl_float_format_t)'
559 which return Catalan's constant.
560 (Suggested by Thomas Papanikolaou.)
562 * New functions `cl_F cl_zeta (int s)',
563 `cl_F cl_zeta (int s, const cl_F&)',
564 `cl_F cl_zeta (int s, cl_float_format_t)'
565 which return the Riemann zeta function at an integral point s>1.
567 * New functions `cl_F cl_exp1 ()',
568 `cl_F cl_exp1 (const cl_F&)',
569 `cl_F cl_exp1 (cl_float_format_t)'
570 which return e = exp(1).
572 * New function `cl_I binomial (uintL n, uintL k)' which returns the
573 binomial coefficient (n choose k).
575 * New functions `int cl_I_to_int (const cl_I&)',
576 `unsigned int cl_I_to_uint (const cl_I&)',
577 `long cl_I_to_long (const cl_I&)',
578 `unsigned long cl_I_to_ulong (const cl_I&)'
579 for converting an integer to a C `int' or `long'.
581 * New functions `float cl_float_approx (const cl_I&)',
582 `float cl_float_approx (const cl_RA&)',
583 `double cl_double_approx (const cl_I&)',
584 `double cl_double_approx (const cl_RA&)'
585 for converting a rational number to a C `float' or `double'.
587 Implementation changes
588 ----------------------
590 * Moved the sources to subdirectories.
591 (Suggested by Jörg Arndt.)
596 * New benchmark for better comparing LiDIA, Pari and CLN.
598 * Added an interface to LiDIA, allows using CLN as a replacement of libI.
599 (Suggested by Thomas Papanikolaou.)
601 * Added an ILOG Talk interface, for interactive use of the library.
610 * New include file <cl_modinteger.h>, defining modular integer rings.
612 * New functions `cl_F cl_eulerconst ()', `cl_F cl_eulerconst (const cl_F&)',
613 `cl_F cl_eulerconst (cl_float_format_t)' which return Euler's constant.
618 * Speed up square root of large numbers by use of Newton's algorithm.
620 * Speed up multiplication and division of large long-floats by small integers.
622 * Speed up the computation of pi, e, exp, log, sin, cos, tan, sinh, cosh, tanh,
623 asin, acos, atan, asinh, acosh, atanh.
624 All of them are now computed with asymptotically fast algorithms, of
625 bit complexity O(log(N)^2*M(N)) = O(N^(1+epsilon)) for N digits.
627 * Fixed several bugs in the transcendental functions routines. In particular,
628 the `log' function went into an endless loop.
630 * Fixed a bug: The cl_I -> cl_DF and cl_I -> cl_LF conversions didn't work
631 correctly for negative integers.
640 * New include file <cl_numtheory.h>, containing number theoretic functions,
641 for now only the Jacobi symbol.
646 * Speed up squaring of large numbers by 30%.
648 * Speed up division of large numbers by use of Newton's algorithm. The
649 improvement is noticeable for numbers with at least about 1000 words =
650 32000 bits = 10000 decimal digits.
652 * Speed up the binary-to-decimal conversion of large numbers by use of
653 a divide-and-conquer method. The improvement is noticeable for numbers
654 of at least 1250 words = 40000 bits = 12000 decimal digits.
656 * Fixed a bug: The functions `cl_boolean minusp (const type&)' and
657 `cl_boolean zerop (const type&)' were missing for type=cl_I and type=cl_RA.
659 * Fixed a bug: The function `cl_boolean logtest (const cl_I&, const cl_I&)'
660 didn't work correctly if both arguments were bignums.
662 * Fixed a bug: The function `cl_I mod (const cl_I&, const cl_I&)', when called
663 as `mod(-n*m,m)' (n>0), returned m instead of 0.
665 * Fixed a bug: The function `uintL power2p (const cl_I&)' returned wrong
666 values for arguments x = 2^n, n >= 29.
668 Implementation changes
669 ----------------------
671 * Speed up CLN by using the low-level addition/subtraction, shift and
672 multiplication loops from GNU gmp-2.0.2.
674 * Fixed a severe reference counting bug in the constructors
675 `type::type (const long)' and `type::type (const unsigned long)'
676 for type=cl_number, type=cl_N, type=cl_RA, type=cl_I that most often
679 * CLN's word sequences can be chosen to be stored big-endian in memory
680 (like it was up to now) or little-endian (needed for interfacing to GMP).
689 * New include file <cl_timing.h>, contains timing facilities.
694 * Speed up the multiplication of large numbers by use of an FFT based
695 algorithm (Schönhage-Strassen). The improvement is noticeable when both
696 factors have at least 1250 words = 40000 bits = 12000 decimal digits.
698 * Speed up the functions `cl_I gcd (const cl_I&, const cl_I&)' and
699 `cl_I xgcd (const cl_I&, const cl_I&, cl_I*, cl_I*)' by a factor
700 of 1.5. Simple trick.
702 * Speed up the function `cl_boolean sqrtp (const cl_I&, cl_I*)' using a
703 trick from Henri Cohen.
705 * Fixed an endless loop in the function `uint32 gcd (uint32, uint32)'
706 which occurred when one of the arguments is zero.
708 * Fixed an endless loop in the function `cl_I factorial (uintL)'.
710 Implementation changes
711 ----------------------
713 * <cl_io.h> now uses iostreams by default instead of stdio FILE pointers.
714 (Reported by Tito Flagella.)
716 * Fixed linking error when compiling without -O.
717 (Reported by Tito Flagella.)
719 * Fixed syntax error in <cl_input.h>.
720 (Reported by Tito Flagella.)
722 * Fixed syntax errors in src/cl_DS.h and src/cl_FF_plus.h.
723 (Reported by Marcus Daniels.)
725 * More portability fixes.
727 * Configuration now supports the --srcdir option.
728 (Reported by Marcus Daniels.)
734 * Compiles well on Sparc, using gcc.
736 * Mail to beta testers.
742 * Compiles well on Linux, using gcc.