Sun 5 Oct 2025 : removed excluded_source_files from words.gpr.

Sat 8 Feb 2025 : added ts_modbal.adb to test the balancing of a sequence
  of doubles by the addition of a modifier constant; updated words.gpr.

Tue 14 Jan 2025 : new ts_wallvhda.adb to measure wall clock time,
  updated test_vectored_hexa_doubles and words.gpr.
  New ts_mtvhda.adb for a high level parallel version of ts_wallvhda,
  calls new procedure in test_vectored_hexa_doubles, updated words.gpr.
  Updates words.gpr to compute the number of cores also on Windows.

Sun 12 Jan 2025 : added ts_leibniz.adb to experiment with the reformulation
  of Leibniz rule to approximate pi/4.

Sat 4 Jan 2025 : added to balanced_quarter_doubles an algorithm to split
  into quarters, tested in test_balanced_quarter_doubles.adb.

Thu 2 Jan 2025 : to test_vectored_{double, quad, octo, hexa}_doubles.adb,
  added frequency and timers.  New vectored_doubles, test_vectored_doubles,
  and ts_vda.adb to provide the baseline to compare with double double times.
  Updated words.gpr, edit in test_vectored_hexa_doubles.ads.
  Removed redundant with/use clauses in vectored_doubles.adb.

Wed 1 Jan 2025 : to balanced_quarter_doubles added the definition of
  balanced quarter doubles, tested by test_balanced_quarter_doubles.
  In balanced_quarter_doubles, in is_balanced, allow f's to be negative.
  Added timers to test_vectored_hexa_doubles.adb and changed the
  compiler switches in words.gpr to full optimization.

Tue 31 Dec 2024 : improved the wrappers in balanced_quarter_doubles,
  used in test_vectored_quad_doubles.adb, in a balanced product procedure
  in vectored_quad_doubles.  Fixed the test on the balanced product in
  test_vectored_octo_doubles.adb, after updating vectored_octo_doubles.
  Updated vectored_double_doubles and test_vectored_double_doubles.adb.
  Fixed balanced_quarter_doubles.adb, edit in vectored_octo_doubles.ads.
  New vectored_hexa_doubles, test_vectored_hexa_doubles, ts_vhda.adb,
  updated words.gpr.

Mon 30 Dec 2024 : tested the wrappers in test_balanced_quarter_doubles
  and recalibrated the exponents in balanced_quarter_doubles.adb.
  To test_vectored_{double, quad, octo}_doubles, added tests on product 
  of balanced products.  Modified vectored_{quad, octo}_doubles.adb.

Sun 29 Dec 2024 : new balanced_quarter_doubles and ts_qtrbal.adb,
  calling the main in test_balanced_quarter_doubles.  Updated words.gpr.
  To balanced_quarter_doubles, added procedures to make balanced quarter
  double doubles, quad doubles, octo doubles, and hexa doubles.
  Added wrappers to balanced_quarter_doubles.

Sat 28 Dec 2024 : introduced more accumulators in vectored_octo_doubles.
  Added test on distribution of the numbers to test_bits_of_doubles.

Fri 27 Dec 2024 : factored out the sign balance procedures out of the
  bits_of_doubles into a new package sign_balancers, moved the tests on
  the sign balancers from test_bits_of_doubles into test_sign_balancers,
  making a new ts_sgnbal.adb main test procedure.  Updated words.gpr
  and the packages vectored_{double, quad, octo}_doubles.adb.
  To bits_of_doubles, added a function to count the number of last
  zero bits in the fraction of a double, tested in test_bits_of_doubles.
  Applied last_zero_count in vectored_{double, quad, octo}_doubles.adb
  to compute the filling of the fractions in the subsums.
  Improved the accuracy of the product in vectored_quad_doubles
  via the addition of extra accumulator variables.

Thu 26 Dec 2024 : in vectored_quad_doubles.adb, improved the accuracy of
  to_quad_double with 15 additions in quad double precision.
  Shortened definition of sign_quarter in vectored_quad_doubles.adb.
  New vectored_octo_doubles, test_vectored_octo_doubles, ts_voda.adb,
  updated words.gpr.

Wed 25 Dec 2024 : new vectored_quad_doubles, test_vectored_quad_doubles, and
  ts_vqda.adb, with updated words.gpr, for vectored quad double arithmetic.
  Applied sign balancing in test_vectored_quad_doubles.adb.
  Corrected documentation in test_vectored_quad_doubles.ads.

Tue 24 Dec 2024 : improved verbose runs in vectored_double_doubles
  with the writing of the subsums.  Added sign balance procedures
  for quad doubles, octo doubles, and hexa doubles to bits_of_doubles;
  and extended test_bits_of_doubles.

Mon 23 Dec 2024 : added sign balance procedure to bits_of_doubles,
  tested in test_bits_of_doubles.  Simplified the signed quarter procedure
  in vectored_double_doubles, updated test_vectored_double_doubles.adb.
  Added wrapper for complex product to vectored_double_doubles,
  extending test_vectored_double_doubles.adb.

Sun 22 Dec 2024 : initialized development of sign aware wrappers for the
  real product in vectored_double_doubles, with updates in the tests in
  test_vectored_double_doubles.adb.

Sat 21 Dec 2024 : added wrapper to vectored_double_doubles to compute the
  complex squared norm, updated test_vectored_double_doubles.adb.

Fri 20 Dec 2024 : extended vectored_double_doubles with a wrapper for
  compute squared norms, taking signs of the quarters into account,
  extended test_vectored_double_doubles.

Wed 18 Dec 2024 : added sign aware wrapper to vectored_double_doubles
  and extended test_vectored_double_doubles.adb.
  Bug fixed in to_double_double of vectored_double_doubles,
  updated test_vectored_double_doubles.adb again.
  Added wrapper for complex sum to vectored_double_doubles,
  tested in test_vectored_double_doubles.adb.

Tue 17 Dec 2024 : extended vectored_double_doubles with the inner product
  of two vectors, tested by test_vectored_double_doubles,
  and its main test procedure is called in ts_vdda.adb.
  Added test on norm to test_vectored_double_doubles.
  In vectored_double_doubles.adb, used the error in the conversion of
  the sums to a double double, in an attempt to be more accurate.

Mon 16 Dec 2024 : moved the code in ts_splitdbl.adb into test_bits_of_doubles,
  and removed ts_splitdbl from words.gpr.  Similarly, moved ts_maskbits.adb
  into test_bits_of_doubles, removing ts_maskbits from words.gpr.
  The files ts_splitdbl.adb and ts_maskbits.adb are then also removed.
  New package vectored_double_doubles with code from ts_ddbits.adb
  and ts_vdda.adb.  Removed ts_ddbits from words.gpr, updated ts_vdda.adb.

Thu 12 Dec 2024 : fixed the Split procedure in bits_of_doubles so there
  is no roundoff when adding the four parts of the split.
  Fixed ts_splitdbl.adb to work out a particular case.

Wed 11 Dec 2024 : factored out fraction_exponent() function out of the
  code in the package bits_of_doubles.  Added split_bits() to the package
  bits_of_integers, tested by test_bits_of_integers.

Tue 10 Dec 2024 : new Add with base 2^52 in double_integer_arithmetic,
  tested by test_double_integers.

Sat 7 Dec 2024 : extended bits_of_integers with a bit_size function,
  used in test_double_integers.adb to see that, with 60 bits or less,
  there is no carry over in the multiplication function.
  Add multiplication of double integers to double_integer_arithmetic,
  with added tests in test_double_integers.

Fri 6 Dec 2024 : bug fixed in the Value function of test_double_integers.adb,
  updated printing in double_integer_arithmetic.adb.
  Define a multiplication in double_integer_arithmetic,
  tested by test_double_integers.adb.

Thu 5 Dec 2024 : extended bits_of_integers and test_bits_of_integers
  with splitters of unsigned 64-bit integers into 30-bit words.
  Added double_integer_arithmetic and test_double_integers,
  with main test called in ts_dblint.adb, updated words.gpr.

Tue 3 Dec 2024 : new packages bits_of_integers, test_bits_of_integers,
  with procedure ts_intbits.adb, updated words.gpr.

Sat 2 Nov 2024 : new ts_splitdbl, extended bits_of_doubles, moved 
  procedures of ts_ddbits into bits_of_doubles.

Thu 31 Oct 2024 : new ts_vdda.adb to test vector sum of double doubles.

Mon 28 Oct 2024 : move split procedures from ts_ddbits.adb into the
  package bits_of_doubles, adding a check on a zero last part.
  Fixed ts_ddbits.adb by application of Double_Double_Basics.split.

Sat 26 Oct 2024 : new mask_bits_of_doubles to get the last and first bits
  of a 52-bit integer, as a fraction of doubles, tested by ts_maskbits.
  Added two procedures to test_bits_of_doubles, to avoid the arrays of
  naturals when chopping or inserting bits to fractions of doubles.
  New ts_ddbits to split high and low parts of double doubles.

Fri 25 Oct 2024 : extended test_bits_of_doubles.adb and bits_of_doubles
  with wrappers to insert bits to the fraction of a double.
