| 1 |
Revision history for Data-Float: |
| 2 |
|
| 3 |
v0.015 2025-04-02 15:06+01:00 Europe/London |
| 4 |
[Documentation] |
| 5 |
- Update security policy and fix inconsistent contact instructions. |
| 6 |
|
| 7 |
This clarifies the license of the security policy. GH#3. |
| 8 |
|
| 9 |
- Update Changes to conform to the CPAN::Changes spec. GH#2. |
| 10 |
|
| 11 |
[Other] |
| 12 |
- Issues have been moved to GitHub. GH#1. |
| 13 |
|
| 14 |
v0.014 2025-03-28 17:08+00:00 Europe/London |
| 15 |
|
| 16 |
[Documentation] |
| 17 |
- Added a security policy. |
| 18 |
|
| 19 |
- Updated maintainer information. |
| 20 |
|
| 21 |
[Toolchain] |
| 22 |
- Removed use of Module::Build, and re-add Makefile.PL to the distribution. |
| 23 |
|
| 24 |
v0.013 2017-07-29 |
| 25 |
|
| 26 |
- no longer include a Makefile.PL in the distribution |
| 27 |
|
| 28 |
- in documentation, use four-column indentation for all verbatim |
| 29 |
material |
| 30 |
|
| 31 |
- in META.{yml,json}, point to public bug tracker |
| 32 |
|
| 33 |
v0.012 2012-02-04 |
| 34 |
|
| 35 |
- include META.json in distribution |
| 36 |
|
| 37 |
- convert .cvsignore to .gitignore |
| 38 |
|
| 39 |
- add MYMETA.json to .gitignore |
| 40 |
|
| 41 |
v0.011 2010-10-21 |
| 42 |
|
| 43 |
- port to Perl 5.13.6, where changed behaviour of signed zeroes in |
| 44 |
Perl arithmetic broke nextdown() and nextafter() and caused false |
| 45 |
test failures for several other functions |
| 46 |
|
| 47 |
v0.010 2010-07-26 |
| 48 |
|
| 49 |
- use simpler "parent" pragma in place of "base" |
| 50 |
|
| 51 |
- in documentation, use the term "truth value" instead of the less |
| 52 |
precise "boolean" |
| 53 |
|
| 54 |
- check for required Perl version at runtime |
| 55 |
|
| 56 |
- use full stricture in test suite |
| 57 |
|
| 58 |
- use full stricture in Build.PL |
| 59 |
|
| 60 |
- in Build.PL, explicitly declare configure-time requirements |
| 61 |
|
| 62 |
- remove bogus "exit 0" from Build.PL |
| 63 |
|
| 64 |
- add MYMETA.yml to .cvsignore |
| 65 |
|
| 66 |
v0.009 2008-04-06 |
| 67 |
|
| 68 |
- refer to optional constants more carefully in the code, to avoid |
| 69 |
syntax problems on systems that lack infinities and NaNs |
| 70 |
|
| 71 |
- fix a skip count that caused false test failures on systems lacking |
| 72 |
signed zeroes |
| 73 |
|
| 74 |
v0.008 2008-04-02 |
| 75 |
|
| 76 |
- bugfix: in initialisation, correctly override any ambient |
| 77 |
$SIG{__DIE__} |
| 78 |
|
| 79 |
v0.007 2007-10-02 |
| 80 |
|
| 81 |
- close DATA filehandle when finished reading from it |
| 82 |
|
| 83 |
- fix some tests that were producing false failures on perl 5.6 |
| 84 |
|
| 85 |
v0.006 2007-10-01 |
| 86 |
|
| 87 |
- bugfix: change behaviour of nextafter() when both arguments are |
| 88 |
numerically zero to that specified by IEEE 754: the second argument |
| 89 |
is returned rather than the first |
| 90 |
|
| 91 |
- bugfix: in hex_float(), cleanly reject non-ASCII digits in exponent |
| 92 |
|
| 93 |
- bugfix: in float_hex(), cleanly reject non-ASCII digits in an |
| 94 |
"EXPONENT="-type "zero_strategy" setting |
| 95 |
|
| 96 |
- add IEEE 754r functions nextup() and nextdown() |
| 97 |
|
| 98 |
- add IEEE 754r function totalorder() |
| 99 |
|
| 100 |
- add constant max_number |
| 101 |
|
| 102 |
- hex_float(): accept IEEE 754r special input strings "infinity" and |
| 103 |
"snan", in addition to the existing "inf" and "nan" |
| 104 |
|
| 105 |
- hex_float(): make the "0x" prefix in hexadecimal input optional |
| 106 |
|
| 107 |
- float_hex(): new option "hex_prefix_string" to control the "0x" prefix |
| 108 |
|
| 109 |
- test classification functions, examination functions, string |
| 110 |
conversion functions, and manipulation functions (all the functions |
| 111 |
that were not being tested) |
| 112 |
|
| 113 |
- test all functions for lack of side effects on zero arguments and |
| 114 |
purity of zero results |
| 115 |
|
| 116 |
- in documentation, note new standard-conforming behaviour of copysign() |
| 117 |
with a NaN second argument |
| 118 |
|
| 119 |
- in documentation, note that hex_float() accepts the special "0.0" |
| 120 |
form of input without a sign, as well as with one |
| 121 |
|
| 122 |
- in documentation, where the IEEE standard nature of functions is |
| 123 |
noted, add discussion of non-conforming behaviour on NaNs |
| 124 |
|
| 125 |
- in documentation, change some variable names in the synopsis for |
| 126 |
clarity |
| 127 |
|
| 128 |
- test POD syntax and coverage, and rename some internal functions to |
| 129 |
satisfy the coverage test |
| 130 |
|
| 131 |
- tweak tests on constants to avoid infinite loops if importing constant |
| 132 |
functions fails |
| 133 |
|
| 134 |
- build with Module::Build instead of ExtUtils::MakeMaker |
| 135 |
|
| 136 |
- complete dependency list |
| 137 |
|
| 138 |
- include signature in distribution |
| 139 |
|
| 140 |
- in documentation, separate "license" section from "copyright" section |
| 141 |
|
| 142 |
v0.005 2007-01-25 |
| 143 |
|
| 144 |
- bugfix: change behaviour of copysign() with a NaN as the second |
| 145 |
argument to that specified by IEEE 754: it is not an error but |
| 146 |
(in the context of Perl's opaque NaNs) results in copying an |
| 147 |
unpredictable sign |
| 148 |
|
| 149 |
- add hex_float() function to input floating point values in hexadecimal |
| 150 |
|
| 151 |
- add IEEE 754 function signbit() |
| 152 |
|
| 153 |
- float_id_cmp(): tighten specification of return values to match |
| 154 |
Perl's <=> operator (actual behaviour always matched the tighter spec, |
| 155 |
it just wasn't documented and tested) |
| 156 |
|
| 157 |
- in documentation, note that the string values of float arguments |
| 158 |
are ignored |
| 159 |
|
| 160 |
- in documentation, clarify note about implicit conversion of integer |
| 161 |
to float |
| 162 |
|
| 163 |
- in documentation, note that both arguments to nextafter() must be |
| 164 |
floating point values |
| 165 |
|
| 166 |
- in documentation, note standard nature of signbit(), copysign(), |
| 167 |
and nextafter() |
| 168 |
|
| 169 |
- in documentation, note float_id_cmp()'s relation to the IEEE 754r |
| 170 |
function totalorder() |
| 171 |
|
| 172 |
- in documentation, note that Perl does not distinguish between |
| 173 |
different NaNs |
| 174 |
|
| 175 |
- in documentation, give a second definition of max_integer (equivalent |
| 176 |
to the first) |
| 177 |
|
| 178 |
- add test for consistency of constants |
| 179 |
|
| 180 |
- in documentation, reference Scalar::Number |
| 181 |
|
| 182 |
- remove now-useless test t/use.t |
| 183 |
|
| 184 |
v0.004 2007-01-12 |
| 185 |
|
| 186 |
- bugfix: correct value for max_integer to 2^(significand_bits+1), |
| 187 |
and correct its description in the documentation |
| 188 |
|
| 189 |
- make pos_zero and neg_zero constants into non-constant functions |
| 190 |
that return fresh floating point zeroes, if running on a Perl where |
| 191 |
floating point zeroes can transmogrify into integer zeroes |
| 192 |
|
| 193 |
- add float_id_cmp() function for total ordering of floats |
| 194 |
|
| 195 |
- in documentation, more details of behaviour of zeroes |
| 196 |
|
| 197 |
- in documentation, discussion of interaction with native integer values |
| 198 |
|
| 199 |
- in documentation, note truncating behaviour of the % operator |
| 200 |
|
| 201 |
- in documentation, note the slightly misleading nature of the names |
| 202 |
"significand_bits" and "max_integer" |
| 203 |
|
| 204 |
- reference Data::Integer and perlnumber(1) in documentation |
| 205 |
|
| 206 |
v0.003 2006-08-08 |
| 207 |
|
| 208 |
- float_hex(): add OPTIONS parameter to control details of output |
| 209 |
formatting |
| 210 |
|
| 211 |
- when looking for NaNs, see whether the string "nan" qualifies |
| 212 |
|
| 213 |
- slight clarification to documentation of significand_step |
| 214 |
|
| 215 |
v0.002 2006-08-03 |
| 216 |
|
| 217 |
- bugfix: in mult_pow2(), copysign(), and nextafter(), take care |
| 218 |
to return a pristine signed zero when returning zero: they were |
| 219 |
returning zeroes that got broken (due to the Perl bug noted in the |
| 220 |
documentation) by internal arithmetic |
| 221 |
|
| 222 |
v0.001 2006-08-01 |
| 223 |
|
| 224 |
- add IEEE 754 functions copysign() and nextafter() |
| 225 |
|
| 226 |
- test sign of zero using sprintf() to avoid being confused by dualvars |
| 227 |
|
| 228 |
- slight expansion of initial paragraph of documentation |
| 229 |
|
| 230 |
v0.000 2006-07-30 |
| 231 |
|
| 232 |
- initial released version |