ChangeLog 1.05 MB
Newer Older
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
2018-09-20  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
	doc/implementconstant.tex, doc/remez.tex, doc/sollya.pdf,
	doc/time.tex, doc/timing.tex, doc/www/bashevaluate.php,
	doc/www/bashexecute.php, doc/www/implementconstant.php,
	doc/www/remez.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the documentation

2018-09-20  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf, doc/sollya.tex, doc/www/sollya.php, implement.c,
	mpfi-compat.c, polynomials.c, sollya.1,
	tests-lib/timplementpoly.reference,
	tests-lib/tv_implementpoly.reference, tests-tool/libraryexample.c,
	version.h: Updated copyrights

2018-09-19  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Adding non regression tests for bug #21716

2018-09-19  Christoph Lauter <christoph.lauter@lip6.fr>

	* mpfi-compat.c: Patching a small bug in sollya_mpfi_div_z.

2018-09-18  Christoph Lauter <christoph.lauter@lip6.fr>

	* mpfi-compat.c: Still trying to patch '#21716 evaluate returns
	incoherent results'  And some clean-up for the MPFI compat
	functions.

2018-09-17  Christoph Lauter <christoph.lauter@lip6.fr>

	* mpfi-compat.c, polynomials.c: Trying to patch '#21716 evaluate
	returns incoherent results'  Non-regression tests still need to be
	added.

2018-08-24  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog: Regenerated ChangeLog

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
2018-08-24  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
	doc/introExample51.tex, doc/printxml.tex, doc/readxml.tex,
	doc/remez.tex, doc/sollya.pdf, doc/time.tex, doc/timing.tex,
	doc/www/bashevaluate.php, doc/www/bashexecute.php,
	doc/www/introExample51.php, doc/www/printxml.php,
	doc/www/readxml.php, doc/www/remez.php, doc/www/time.php,
	doc/www/timing.php, help.h, sollya.1: Recompiled the documentation.


2018-08-24  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/readxmlexample.xml, general.c,
	tests-tool/check008.sollya.reference, xml.c: Changed
	sollya.gforge.inria.fr to sollya.org

2018-08-22  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Made help2man work in a VPATH build context.

2018-08-09  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, configure.ac: Yet improved support of locale with
	help2man  We now test whether the --locale=C.UTF-8 option works or
	not in configure.

2018-08-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Generalized the policy of displaying a meaningful
	answer in configure before exiting.  AC_MSG_ERROR and AC_MSG_WARN
	must be used to display additional information on stderr.

2018-08-01  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: configure must stop whenever no version of fplll is
	found.

2018-08-01  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Display no as a result of a test, before printing a
	more detailed warning.  Notices that AC_MSG_RESULT displays on
	stdout, whereas AC_MSG_WARN displays on stderr.

2018-07-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog: Regenerated ChangeLog

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
2018-07-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/bashevaluate.tex, doc/bashexecute.tex, doc/remez.tex,
	doc/sollya.pdf, doc/time.tex, doc/timing.tex,
	doc/www/bashevaluate.php, doc/www/bashexecute.php,
	doc/www/remez.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the doc.

2018-07-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Initialized a variable to make a compiler happy.

2018-07-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* execute.c: Removed a warning signaled by gcc.  No need to update
	copyrights: it is a minor fix.

2018-07-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, sollya.map: We were wrong: the libtool version of the
	lib is 7.0.0  because the function behind the exported symbol
	'sollya_tool_main' changes compltely from Sollya 6.0 to Sollya 7.0
	(hence no backward compatibility => age=0)

2018-07-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf, doc/sollya.tex, doc/www/sollya.php: Fixed a few
	typos and converted the doc of exit status to html.

2018-07-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf, doc/sollya.tex: Documented the Sollya exit status
	codes.

2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf, doc/sollya.tex, doc/www/sollya.php: Clarified the
	phrase about gnuplot being mandatory-but-not-really in the doc

2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf, doc/sollya.tex, doc/www/sollya.php: Fixed a few
	typos in the doc. HTML and PDF are synchronized.

2018-07-04  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf, general.c: Patched a small bug with the return
	code returned when the tool just executes for the --help or
	--version options.

2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf, doc/sollya.tex: Followed advice of Jérôme Benoit
	to make the package reproducible

2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog: Regenerated ChangeLog

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/timecheck.sollya: Updated reference timings in
	timecheck

2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/timecheck.sollya: Added manual affectation of variables
	before doing the actual time checks.  Indeed, in some cases, we were
	measuring the time required to deaffect a variable rather than the
	time required to compute its new value.

2018-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/timecheck.sollya: Fixed an issue introduced with commit
	5f10b52d49fe.  The 3rd timecheck was intenteded to work with the
	polynomial p defined in the 1st check. But the modification
	introduced with commit 5f10b52d49fe changed that p in the 2nd check.


2018-07-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check004.sollya, tests-tool/check005.sollya,
	tests-tool/check012.sollya, tests-tool/check013.sollya,
	tests-tool/check029.sollya: Clean up files generated by the tests
	after their use.

2018-07-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Made sollya.1 depends on version.h rather than the
	sollya binary.  This is to ensure that it won't be regenerated by
	default in a distributed release. I followed recommendations of

	https://www.gnu.org/software/automake/manual/html_node/Errors-with-distclean.html

2018-07-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Setting the locale for help2man to something that
	should work on all systems

2018-07-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog: Regenerated ChangeLog

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
2018-07-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.h: Updated copyrights

2018-07-03  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Disabling locale support in help2man...  ... as some
	versions of help2man are not configured to work with this support,
	and we are not really using it anyway. The only thing we care are
	the environment variables that set the language which is used by
	help2man for generating its header.

2018-07-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* Makefile.am, base-functions.c, base-functions.h, expression.h: 
	Solved that fu**ing multiple definition issue seen on Mac OS X.

2018-07-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* configure.ac: Again trying to do something for certain versions of
	lld and llvm on Mac OS X.

2018-07-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* base-functions.c, base-functions.h, chebyshevform.c,
	expression.c, expression.h, implementconst.c, taylorform.c: Trying
	to work around a problem seen on Mac OS X with the fixif
	integration.

2018-07-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog: New ChangeLog

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330
2018-06-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: Fixed a memory leak

2018-06-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, help.h, hooks.c: Copyrights are now up to date

2018-06-29  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
	doc/implementconstant.tex, doc/introExample51.tex, doc/remez.tex,
	doc/sollya.pdf, doc/time.tex, doc/timing.tex,
	doc/www/bashevaluate.php, doc/www/bashexecute.php,
	doc/www/implementconstant.php, doc/www/introExample51.php,
	doc/www/remez.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the documentation.

2018-06-29  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/check030.sollya, tests-tool/check030.sollya.reference: 
	Added a non-regression test for an issue where annotatefunction is
	used to annotate a constant expression.

2018-06-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/CommandsAndFunctions.tex, doc/Makefile.am, doc/{eucldiv.shlp
	=> diveucl.shlp}, doc/{eucldiv.tex => diveucl.tex}, doc/gcd.shlp,
	doc/gcd.tex, doc/keywords.def, doc/{euclmod.shlp => modeucl.shlp},
	doc/{euclmod.tex => modeucl.tex}, doc/sollya.pdf,
	doc/www/ListOfCommands.php, doc/www/{eucldiv.php => diveucl.php},
	doc/www/gcd.php, doc/www/{euclmod.php => modeucl.php}, help.h,
	parser.y: Changed keywords euclmod and eucldiv into modeucl and
	diveucl respectively.  This is to ensure a natural sorting position
	of the corresponding entry in the table of contents of the pdf
	documentation.

2018-06-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* NEWS: typo in the NEWS

2018-06-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check025.sollya, tests-tool/check025.sollya.reference: 
	Non-regression tests for the bugs fixed at revisions 6450330 and
	d4eb1bc

2018-06-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: Another instance of the bug addressed in commit
	6450330.

2018-06-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* expression.c, help.h: Some errors of copyrights I just detected.
	Copyrights are normally up-to-date.

2018-06-29  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/tget_element_in_list_timecomplexity.c: Patched a couple
	of issues in the complexity test for get_element_in_list.

2018-06-29  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, expression.c: Improved hook handling for cases when the
	function to be annotated is constant.

2018-06-29  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/tget_element_in_list_timecomplexity.c: Patched
	timecomplexity test again.

2018-06-29  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, hooks.c: Patched a small issue with annotatefunction
	used on constant expressions that are zero.

2018-06-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* base-functions.c: Patched a bug found with floor(c), when c =
	log2(13)/log2(17) - log(13)/log(17);

2018-06-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/tget_element_in_list_timecomplexity.c: Made test
	tget_element_in_list_timecomplexity work on a Raspberry Pi 3 running
	jessie.

2018-06-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/eucldiv.shlp, doc/eucldiv.tex, doc/euclmod.shlp,
	doc/euclmod.tex, doc/gcd.shlp, doc/gcd.tex, doc/sollya.pdf,
	doc/www/eucldiv.php, doc/www/euclmod.php, doc/www/gcd.php, help.h: 
	Changed documentation of div, mod, gcd again a little bit.

2018-06-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* NEWS: Modified NEWS.

2018-06-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/plot.shlp, doc/plot.tex, doc/sollya.pdf, doc/www/plot.php,
	help.h: Added documentation of SOLLYA_GNUPLOT in the doc of the plot
	command

2018-06-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/eucldiv.shlp, doc/eucldiv.tex, doc/euclmod.shlp,
	doc/euclmod.tex, doc/gcd.shlp, doc/gcd.tex, doc/sollya.pdf,
	doc/www/eucldiv.php, doc/www/euclmod.php, doc/www/gcd.php, help.h: 
	Modified a little bit the documentation of gcd, div and mod

2018-06-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/timecheck.sollya: Improved stability of the timecheck.


2018-06-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, sollya.map: Updated library version info

2018-06-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* NEWS: Proposition of a NEWS file for the release 7.0

2018-06-27  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* sollya.1: Updated man page

2018-06-27  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, configure.ac: Trying to use -no-undefined
	unconditionally. Let's see what happens.

2018-06-26  Christoph Lauter <christoph.lauter@lip6.fr>

	* configure.ac: Made a warning message slightly less confusing for
	Cygwin users.

2018-06-26  Christoph Lauter <christoph.lauter@lip6.fr>

	* Makefile.am, configure.ac: Made shared libraries build on Cygwin.


2018-06-26  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Fixed a small mistake in the detection of pdflatex
	The message indicating whether pdflatex comes along with the
	required packages was displayed in any case, even when pdflatex
	itself had not been detected.

2018-06-25  Christoph Lauter <christoph.lauter@lip6.fr>

	* configure.ac: Just warn if libmpfi is not found at runtime.

2018-06-25  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, expression.h, infnorm.c, printf.c: Worked again on the
	fallthrus in switch statements.

2018-06-25  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* execute.c: Some fall-through signaled by gcc

2018-06-22  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Fixed a bad syntax in configure.ac  I do not fully
	understand what happens here, but I think one pair of brackets is
	eaten by the autoconf logic, so we need three pairs of brackets to
	make two in effect.

2018-06-21  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf: Strange recompilation.

2018-06-21  Christoph Lauter <christoph.lauter@lip6.fr>

	* configure.ac, execute.c, expression.c, expression.h, infnorm.c,
	match.c, polynomials.c, printf.c: Tried to implement logic to use
	gcc's attribute for fall-thrus in switches.

2018-06-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* autodiff.c, autodiff.h, base-functions.c, base-functions.h,
	chain.c, chain.h, chebyshevformaux.c, chebyshevformaux.h,
	doc/conversionIntro.sh, doc/sollya.tex, doc/www/sollya.php,
	double.c, double.h, execute.c, expression.c, expression.h,
	fplll_wrapper.cpp, fplll_wrapper.h, fpminimax.c, general.c,
	general.h, help.h, hooks.c, implement.c, implementconst.c,
	implementconst.h, infnorm.c, infnorm.h, integral.c, lexer.l,
	match.c, mpfi-compat.c, mpfi-compat.h, plot.c, remez.c, remez.h,
	sollya-library-wrappers.c, sollya-messages.h, sollya-messaging.h,
	sturm.c, supnorm.c, taylorform.c,
	tests-lib/timplementpoly.reference,
	tests-lib/tv_implementpoly.reference, version.h, xml.c: 
	Systematically changed Sophia-Antipolis into Sophia Antipolis

2018-06-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* autodiff.c, execute.c, execute.h, expression.c, expression.h,
	external.c, general.c, general.h, hash.c, help.h, hooks.c,
	infnorm.c, infnorm.h, internlexer.l, lexer.l, library.c,
	minilexer.l, mpfi-compat.h, parser.y, plot.c, polynomials.c,
	polynomials.h, remez.c, signalhandling.c,
	sollya-library-wrappers.c, sollya-messages.h, sollya-messaging.h,
	sollya.h, sturm.c, supnorm.c: Updated copyrights changed between
	sollya-4.0 and sollya-6.0  The rule had been badly applied at this
	time (we should keep the list of all former affiliations, even when
	they denote the same entity)

2018-06-21  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/timplementpoly.reference,
	tests-lib/tv_implementpoly.reference: Small patch to make the tests
	work again.

2018-06-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf: And recompiled the doc

2018-06-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* autodiff.c, autodiff.h, base-functions.c, base-functions.h,
	chain.c, doc/sollya.tex, doc/www/sollya.php, double.c, double.h,
	execute.c, execute.h, expression.c, expression.h, external.c,
	fpminimax.c, general.c, general.h, help.h, hooks.c, implement.c,
	implement.h, infnorm.c, internlexer.l, internparser.y, lexer.l,
	library.c, match.c, minilexer.l, miniparser.y, mpfi-compat.c,
	mpfi-compat.h, parser.y, plot.c, polynomials.c, polynomials.h,
	remez.c, sollya-library-wrappers.c, sollya.h, sturm.c, sturm.h,
	taylor.c, taylorform.c, taylorform.h, version.h, xml.c: Updated all
	copyrights since the release of sollya 6.0

2018-06-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* .gitignore: Updates .gitignore with tget_constant_as_uint64_array


2018-06-20  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf: Recompiled the PDF.

2018-06-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* implement.c: Proposition of a more robust patch to bug [#21376].

2018-06-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Non-regression tests for bug [#21680]

2018-06-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/sollya.pdf, doc/sollya.tex, doc/www/sollya.php: Slight
	modifications in the recent documentation of
	sollya_lib_get_constant_as_uint64_array + update of the HTML doc
	accordingly.

2018-06-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-lib/Makefile.am, tests-lib/tget_constant_as_mpz.c,
	tests-lib/tget_constant_as_uint64_array.c,
	tests-lib/tget_constant_as_uint64_array.reference: Added tests for
	the new sollya_lib_get_constant_as_uint64_array function.  Also
	added an extra check on the tests of sollya_lib_get_constant_as_mpz.


2018-06-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Non-regression test for bug fixed with commit 4d1761b.

2018-06-12  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c: Patched a bug with faithful evaluation containing a
	sub-expression for which both childs can be evaluated exactly but
	that would yield a (signed) infinity that makes a limit to be
	reached in the whole expression.  Example:  f = erf(1/x);
	evaluate(f,[0]);

2018-06-12  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
	doc/introExample51.tex, doc/remez.tex, doc/sollya.pdf,
	doc/subpoly.tex, doc/taylor.tex, doc/time.tex,
	doc/www/bashevaluate.php, doc/www/bashexecute.php,
	doc/www/introExample51.php, doc/www/remez.php, doc/www/subpoly.php,
	doc/www/taylor.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the documentation.

2018-06-11  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* fpminimax.c: Fixed a segfault issue.  When pstar is not provided
	to fpminimax, while providing a list of points in a fixed-point
	context, a segfault happens due to an unprotected call to isError.
	This fixes the issue.

2018-05-25  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf, doc/sollya.tex: Documented
	sollya_lib_get_constant_as_uint64_array in the PDF version of the
	documentation.  The HTML documentation still needs to be adapted.

2018-05-23  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* hooks.c: Fixed some memory leaks in descartesRuleBasic.  I did not
	review the code in full details and just applied the obvious fixes.
	Some other leaks might remain. Especially, I have some doubts with
	variable 'coefficients'.

2018-05-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, polynomials.c: Patching memory leaks.

2018-05-15  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-lib/tbuild_function_libraryconstant.c,
	tests-lib/tbuild_function_libraryfunction.c,
	tests-lib/tget_element_in_list_timecomplexity.c,
	tests-lib/tlibraryconstant.c, tests-lib/tlibraryfunction.c: Patched
	minor memory leaks.

2018-05-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, expression.h, general.c: Made variable caching again
	a little bit more robust.

2018-05-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, expression.c: Made memory management for variable nodes
	slightly more robust.

2018-05-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Optimized precision handling in an old function.

2018-05-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c, tests-tool/check025.sollya.reference: Improved
	polynomial degree handling for the case c * p(x)

2018-05-14  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Patched a memory leak.

2018-05-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* expression.c: Patched a memory leak introduced with the branch
	unifiedBaseFunction

2018-05-11  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.h: Improved documentation of getRecurseCutoff and
	getRecursePrec in base-function.h now that their role has been
	clarified by Christoph in commit c155bf2.

2018-05-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, base-functions.h, expression.c, infnorm.c: 
	Clarified the semantic of point_eval.  This is an evaluation "à la"
	MPFR, with an implicit rounding mode MPFR_RNDN. The rounding
	functions of double.c is patched in master in order to conform this
	semanantic, whatever the precision of the ouput variable (see commit
	87d9b3b2).

2018-05-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.h: Changed a little bit the comment added with
	56cf7f2.  Indeed, before being passed to evalsign the sub-tree is
	simplified with obvious simplifications, but no deep simplification.
	For instance, rounding functions know that their result fit on some
	explicit precision format and the simplification does compute the
	exact result in general. A similar argument holds for abs: evalsign
	is called by the simplification function in order to try to simplify
	it. Therefore, in those cases, having still the symbol (ABS, DOUBLE,
	etc.) as head symbol likely means that the sign will not be easy to
	guess and it is legitimate to return without even trying.  This
	argument does not hold for integer rounding functions: here the
	result is exactly representatble on *some* precision format, but we
	do not know which format. Hence, the simplification procedure will
	try to compute the exact value, but keeping a reasonably small
	precision format. It can fail if this trial precision is too small,
	in which case we keep the head symbol (FLOOR, etc.). But it does not
	mean that the sign cannot be easily computed.

2018-04-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: abs_evalsign is now trivialized.  Now that
	abs(expr) is simplified whenever expr is a constant expression whose
	sign can be detected, there is no reason to keep a non-trivial
	evalsign function (same reason as for rounding functions).

2018-04-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: abs(expr) is now simplified when expr is a
	constant expression whose sign can be guessed.  I took this
	opportunity to clean up a bit the code.

2018-03-31  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: Improved readibility of
	simplify_rounding_function.  Also, avoid allocating and initializing
	an mpfr when non-necessary in simplify_integer_rouding_functions

2018-03-30  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: All rounding functions now share the same
	simplification procedure.

2018-03-30  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, tests-lib/tdiff.reference,
	tests-tool/check012.sollya.reference: Rationalized the expressions
	for the derivative of base functions.  * Some derivatives (tan,
	asin, acos and tanh) were composed of x^2, whereas others (atan,
	asinh, acosj, atanh) were composed of x*x. This is now unified with
	x^2 evertywhere. * The derivative of erf and erfc were composed with
	sqrt(atan(1)) to mean sqrt(pi)/2... * Except for sqrt, all base
	functions whose derivatives were of the form 1/f constructed the
	derivative of a composition as (1/f(g))*g'. Now this is
	systematically g'/f(g). * The derivative of acos is now structurally
	the negation of the derivative of asin. The same holds for erfc wrt
	erf. * The derivative of acosh is now 1/sqrt(x^2-1) (with a
	subtraction) instead of 1/sqrt(-1 + x^2)

2018-05-11  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.h, infnorm.c, tests-tool/check029.sollya,
	tests-tool/check029.sollya.reference: Made an important function
	DAG-aware and added some tests for that.

2018-05-09  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c: Internally documented two functions.

2018-05-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, expression.c, match.c: Some modifications that I
	forgot to propagate to the branch unifiedBaseFunction following the
	previous merge of master into it.

2018-05-03  Christoph Lauter <christoph.lauter@lip6.fr>

	* assignment.c, double.c, execute.c, expression.c, expression.h,
	fpminimax.c, general.h, implement.c, infnorm.c, match.c, remez.c,
	sollya-library-wrappers.c, taylor.c: Refactored code for node memory
	management again.

2018-05-03  Christoph Lauter <christoph.lauter@lip6.fr>

	* assignment.c, execute.c, expression.c, expression.h, hooks.c,
	infnorm.c, polynomials.c: Refactored code working with caches held
	in memref'ed nodes.

2018-05-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* double.c: Reworked the double-rounding warnings for the rounding
	functions.

2018-05-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* double.c, expression.c, hooks.c, implement.c: Reworked the
	rounding functions for D, DD, TD, SG, QD, halfprecision.

2018-05-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* double.c, double.h: Cleaned up declaration of private functions in
	double.h

2018-04-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.h: Added a comment about the semantic of evalsign.
	This explains why for functions such as double or ceil, evalsign
	does not even try to do something.

2018-04-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, base-functions.h: Corrected a type error.  It
	was a historical mistake that I did when I ported code of infnorm.c
	from master to the branch unifiedBaseFunction

2018-04-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: Fixed a bug with acosh_evalsign when the
	argument equals 1.  The returned value used to be 1 instead of 0

2018-04-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: abs is not differentiable everywhere on its
	domain.  After discussion with Christoph, we agreed that the
	semantic given as a comment of the definition of the field
	isDifferentiableEverywhere is correct. Consequently, the
	corresponding field in abs becomes false. The previous behavior was
	copied-pasted from the corresponding part in the master, and has
	been introduced with commit e322baf7 in order to prevent a bad use
	of Interval Taylor Evaluation for non-differentiable functions.
	Notice that we were not able to spot a tracking number, nor an email
	or ICQ thread about it, and we could not reproduce a situation
	showing a difference of behavior between right before and right
	after commit e322baf7.  It might appear later that abs was correctly
	handled there, in which case, the present commit should be reverted
	and the semantic (and probably the name) of
	isDifferentiableEverywhere should be patched accordingly.

2018-04-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* chebyshevform.c, double.c, double.h, execute.c, expression.c,
	expression.h, implement.c, infnorm.c, mpfi-compat.c, mpfi-compat.h,
	remez.c, sollya-library-wrappers.c, sturm.c, taylorform.c,
	taylorform.h, worstcase.c: Function prefixed with mpfr_ should
	belong to the MPFR API.  Consequently, all functions specific to
	sollya, even though they are kind of upgrading the API of MPFR for
	Sollya are now named sollya_mpfr_*

2018-04-06  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c, tests-tool/check020.sollya.reference: Increased verbosity
	level of most remez messages.

2018-04-06  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* execute.c, expression.c, lexer.l, polynomials.c: Removed unused
	variables, as signaled by gcc

2018-04-12  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, infnorm.c: Small changes discussed with Sylvain:
	invisible hidden bugs that are bugs nevertheless

2018-03-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: Corrected a mistake: log1p vanishes only at
	zero, so set onlyZeroIsZero to true

2018-03-30  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Slight change to correct an error found by chance
	while discussing with Sylvain.

2018-03-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, base-functions.h, expression.c: Updated
	copyrights and removed unrelevant TODOs

2018-03-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c, base-functions.h, expression.c: Re-designed the
	signature of try_exact_rational_eval in order to unify with the
	style of other functions.  Instead of asking the end-user to pass a
	pointer to the baseFunction object itself, this is now done manually
	in base-function.c, inside individual functions that encapsulate the
	call to the generic try_exact_rational_eval_* functions.

2018-03-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, expression.c, expression.h, taylor.c,
	tests-tool/check011.sollya.reference: Made a lot more operations
	DAG-aware.  Now things like:  > p = taylor(asin(x),9,0.5); >
	degree(p); 9 >  have become feasible.

2018-03-27  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.h: Some progress on substitution cache.

2018-03-27  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Started working on a substitution cache.

2018-03-26  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, expression.c, expression.h, general.c,
	tests-tool/check024.sollya.reference: Unified the creation of
	VARIABLE Sollya objects wherever possible.  The change induces an
	about 5% performance gain with metalibm. The reasons for that gain
	are not really understood but there is a gain.

2018-03-23  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, expression.c, implement.c, infnorm.c, remez.c,
	taylor.c: Started work on unification of all VARIABLE nodes.

2018-03-22  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: With commit da252f1a41af5baed6374a0bba1842970f7976d8, a
	left-over of #20094 has reappeared. This commit patches the
	resulting regression.

2018-03-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/equal.shlp, doc/equal.tex, doc/isevaluable.shlp,
	doc/isevaluable.tex, doc/neq.shlp, doc/neq.tex, doc/sollya.pdf,
	doc/www/equal.php, doc/www/isevaluable.php, doc/www/neq.php,
	help.h, tests-tool/check029.sollya,
	tests-tool/check029.sollya.reference: Modified the documentation
	after the changes introduced with commit da252f1a  Added new tests
	to ensure non-regression of this new behavior. This allowed to
	detect a bug (that look like a regression). The checks currently DO
	NOT PASS and won't until this bug is fixed.

2018-03-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check002.sollya, tests-tool/check002.sollya.reference: 
	Improved a test that commit ce762ff revealed to be badly designed.

2018-03-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check012.sollya, tests-tool/check012.sollya.reference: 
	Non regression test for bug patched with commit 51a3907

2018-03-15  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Non-regression test for bug #21659

2018-03-15  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check022.sollya, tests-tool/check022.sollya.reference: 
	Added non regression test for bug #21658

2018-03-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Trying to patch the left-overs of the bug [#21658]
	Elliptic lists do not support arbitrary precision integers

2018-03-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Patching bug #21659 Misleading error message
	Completely rewrote dirtyIsConstant.

2018-03-14  Christoph Lauter <christoph.lauter@lip6.fr>

	* external.c, general.c, general.h, plot.c: Updating gnuplot
	handling behavior in order to implement Sylvain's dreams as he set
	them out in his email on 2018-03-10 22:36 CET.  Typically:  *
	gnuplot is now called without the persist option  * Any interactive
	gnuplot script contains a gnuplot command to keep   the plot window
	open  * it is ensured that the gnuplot file does not go away until
	the   gnuplot process dies  * it is ensured that the gnuplot process
	dies on its own once the plot   window is closed  This allows
	zooming in the gnuplot window and other manipulations.  For cases
	when the user does not want to call gnuplot but another compatible
	plotting tool, an environment variable SOLLYA_GNUPLOT is read at the
	moment of the first plot.

2018-03-14  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Patches bug #21658 Elliptic lists do not support
	arbitrary precision integers  No non-regression tests have been
	added yet.

2018-03-08  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
	doc/introExample43.tex, doc/introExample51.tex, doc/remez.tex,
	doc/sollya.pdf, doc/time.tex, doc/timing.tex,
	doc/unsuppressmessage.tex, doc/www/bashevaluate.php,
	doc/www/bashexecute.php, doc/www/introExample43.php,
	doc/www/introExample51.php, doc/www/remez.php, doc/www/time.php,
	doc/www/timing.php, doc/www/unsuppressmessage.php, help.h: 
	Recompiled the documentation after patching a bug.

2018-03-08  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c: Patched a bug with evaluation over intervals.

2018-03-08  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/bashevaluate.tex, doc/bashexecute.tex, doc/equal.shlp,
	doc/equal.tex, doc/evaluate.tex, doc/introExample33.tex,
	doc/introExample43.tex, doc/introExample51.tex,
	doc/isevaluable.tex, doc/neq.shlp, doc/neq.tex, doc/remez.tex,
	doc/sollya.pdf, doc/time.tex, doc/timing.tex,
	doc/unsuppressmessage.tex, doc/www/bashevaluate.php,
	doc/www/bashexecute.php, doc/www/equal.php, doc/www/evaluate.php,
	doc/www/introExample33.php, doc/www/introExample43.php,
	doc/www/introExample51.php, doc/www/isevaluable.php,
	doc/www/neq.php, doc/www/remez.php, doc/www/time.php,
	doc/www/timing.php, doc/www/unsuppressmessage.php, help.h: 
	Recompiled the documentation after a small change for == and !=.

2018-03-08  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, tests-lib/tcmp_not_equal.reference,
	tests-tool/check013.sollya.reference,
	tests-tool/check029.sollya.reference: Changed error and NaN behavior
	again, according to our new understanding of IEEE754-2008 and the C
	standard.  NaN == NaN -> false NaN != NaN -> true [NaN] == [NaN] ->
	false [NaN] != [NaN] -> true error == error -> false error != error
	-> false

2018-03-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Patching a strange bug where [NaN;NaN] != [NaN;NaN]
	evaluates to true.

2018-03-06  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check.proto, tests-tool/check029.sollya,
	tests-tool/check029.sollya.reference: Non-regression test for bug
	21654

2018-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* plot.c: Patching a bug with the file option of plot.

2018-03-01  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Improved polynomial division when divisor has just
	one monomial.

2018-02-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Implemented polynomial multiplication for
	polynomials with integer coefficients using Kronecker substitution.


2018-02-27  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Further improved polynomial multiplication.

2018-02-27  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Added special code for rational polynomials.

2018-02-23  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, polynomials.c: Implemented Karatsuba polynomial
	multiplication.

2018-02-12  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Made simplifications in rationalmode = on; work
	better on expressions like floor(log10(10^-30 * (10^15 + 1)))

2018-01-24  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check007.sollya, tests-tool/check007.sollya.reference: 
	Non-regression test for bug #21592

2018-01-08  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Patched a bug with differentiation of polynomials.
	The bug was found with evaluation of an expression where the use of
	Hopital's rule is necessary.

2017-09-27  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c, general.h, internparser.y, lexer.l, miniparser.y,
	parser.y: Removed deprecated 'read' macro.

2017-08-24  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-lib/tbuild_function_libraryconstant.c,
	tests-lib/tbuild_function_libraryconstant.reference,
	tests-lib/tbuild_function_libraryfunction.c,
	tests-lib/tbuild_function_libraryfunction.reference,
	tests-lib/tlibraryconstant.c, tests-lib/tlibraryconstant.reference,
	tests-lib/tlibraryfunction.c, tests-lib/tlibraryfunction.reference: 
	Made tests for library constants and library functions more robust
	J. Benoit discovered false positives while testing sollya on armhf
	architecture for Debian. Indeed, I do not completely understand the
	logic behind the tests (it seems to me that we were testing twice
	the same thing with stupid6 and stupid7). Anyway, the code was
	intended to give a pointer to a random place in the memory by
	considering stupid6 + 17 and stupid7 - 17, and it appears that we
	might (by chance) end up on the address of a true function, whence
	the false positive. I made the code more robust by using a freshly
	allocated pointer, which is the technique already used for another
	test.

2017-06-16  Christoph Lauter <christoph.lauter@lip6.fr>

	* implement.c, implement.h: Trying to patch "#21376 Abandon (core
	dumped) dans implementpoly"

2017-05-19  Christoph Lauter <christoph.lauter@lip6.fr>

	* chain.c: Replaced stupid bubble sort by qsort from stdlib.

2017-05-18  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, tests-tool/check009.sollya: Made simplify less
	precision-hungry.

2017-05-11  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: Implemented the bisection algorithm on the exponents for
	Newton's algorithm when the zero is very close to 0

2017-05-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* remez.c: No, 8 is better.

2017-05-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* remez.c: Back to 512.

2017-05-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* remez.c: The value 512 * prec was too much. There should be a
	bisection algorithm.  In cases when the expression is something like
	((cos(x))^ - 1) / (x^2)  evaluation is easy at 0 (by Horner's rule)
	but not at 2^(-512 * prec) due to cancellation.

2017-05-09  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c, remez.c: Patched a bug in faithful evaluation. Patched
	an issue in dirtyfindzeros.

2017-05-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* infnorm.c, remez.c: Precision was way too large to simply
	determine the sign of a function.  This affected dirtyfindzeros and
	dirtyinfnorm

2017-05-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Patching a performance issue with tree
	simplification.

2017-05-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* remez.c: Avoiding simplifications that are not really needed.

2017-05-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Improving the memory behavior of bind().

2017-05-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c, tests-tool/check001.sollya.reference,
	tests-tool/check002.sollya.reference: Still improving performance of
	evaluate().

2017-05-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c, tests-tool/check001.sollya.reference,
	tests-tool/check003.sollya.reference,
	tests-tool/check005.sollya.reference,
	tests-tool/check024.sollya.reference: Improving the performance of
	evaluate.

2017-05-04  Christoph Lauter <christoph.lauter@lip6.fr>

	* remez.c: dirtyfindzeros just needs to know the sign of the
	function to detect sign changes between sampling points.

2017-03-23  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c: Better precision handling in hook evaluation +
	Hopital's method combination.

2017-03-23  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c, tests-tool/timecheck.sollya,
	tests-tool/timecheck.sollya.reference: Again working on complexity
	issues with annotate.

2017-03-22  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* .gitignore: Updated .gitignore

2017-03-22  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference,
	tests-tool/libraryexample.c: Non-regression test for bug #21283

2017-03-22  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c: Added an extra safety net for external procedures that
	like to fork.

2017-03-21  Christoph Lauter <christoph.lauter@lip6.fr>

	* external.c, general.c, plot.c: Patching an issue with plot
	stemming from a correct usage of the Sollya library in a forking
	programm resp. from an unorthodox combination of external
	procedures, fork and plot.

2017-03-21  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c: Patching an issue in the parser stemming from an
	unorthodox combination of external procedures, fork and plots.

2017-03-20  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/check017.sollya.reference: Changed reference for
	bashevaluate again.

2017-03-17  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c, tests-tool/check017.sollya.reference: Trying to patch
	bug #21247 "Strange 'contact lost' phenomenon when --warnonstderr is
	used"

2017-03-08  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* execute.c, infnorm.c: Some changes forgotten in the previous
	merges

2017-03-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c, mpfi-compat.c, mpfi-compat.h: Made calls to
	libraryfunctions try not to waste precision.

2017-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c: Patching some strange bug in faithful evaluation based
	on hooks that I remember to have fixed already but for which I find
	no clear commit message stating that fact.

2017-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c: Again hooks.

2017-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c: Hooks again.

2017-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* sturm.c, tests-lib/tnumberroots.reference: Improved behavior of
	numberroots.

2017-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c: Again a slight improvements to annotatefunction.

2017-03-06  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf, hooks.c: Slight change to hooks for better debug
	support.

2017-03-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-lib/tremez.c, tests-lib/tremez.reference,
	tests-lib/tv_remez.c, tests-lib/tv_remez.reference,
	tests-tool/check020.sollya, tests-tool/check020.sollya.reference: 
	Fixed tests so as to make them pass again after the recent changes
	in remez

2017-02-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/library.shlp, doc/library.tex, doc/www/library.php, help.h: 
	Made it clear in the doc of library command that one can use
	sollya.h but without calling sollya_lib_[init|close]

2017-03-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c: Correcting the last commit.

2017-03-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf, hooks.c: Added a possiblity to debug hooks.

2017-03-02  Christoph Lauter <christoph.lauter@lip6.fr>

	* autodiff.c, execute.c, expression.c, general.c, general.h,
	infnorm.c, library.c, sollya-library-wrappers.c: Added an additional
	safety net for external procedures, library functions and library
	constants  External code called from within interactive Sollya may
	now spuriously invoke sollya_lib_init and sollya_lib_close without
	negative consequences.

2017-03-01  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check020.sollya: Slight modification of check020 in
	order to ease future maintenance.

2017-03-01  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c: Again some improvements to evaluation hook adding.

2017-02-28  Christoph Lauter <christoph.lauter@lip6.fr>

	* hooks.c: Made annotations with polynomials faster to be installed
	in certain cases.

2017-02-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: Continuing the series of possible evaluation outside the
	domain.  dirtyinfnorm was evaluating at one point juste outside the
	domain as the last step of its loop. The result of this evaluation
	was not considered, anyway, but this behavior might be a problem for
	functions very hard to evaluate outside their domain.

2017-02-28  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: A few other places with the same problem as in the
	previous commit de56e8f

2017-02-27  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: Fixed bug in remez leading to evaluating the function on
	points possibly outside its domain.

2017-02-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* expression.c: Fix stupid bug with
	containsNonDifferentiableSubfunctions.  By the way, to handle this
	function correctly, I had to introduce with commit  f8207dd a new
	field in the type baseFunction. I named it
	isDifferentiableEverywhere, but I must say that I am not 100% sure
	to understand the semantic of containsNonDifferentiableSubfunctions
	and so this field is documented in a way that do not correspond to
	the true behavior. This is a thing that should be clarified with
	Christoph.

2017-02-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* execute.c: Fix a small bug introduced with commit f8207dd  Hash
	computations stopped distinguishing between different base
	functions. This should now be fixed.

2017-02-21  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* base-functions.c: Fix a bug introduced with commit 252474e  I did
	not fix completely properly a conflict. Hopefully, the tests
	discovered it.

2017-02-08  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/Makefile.am, tests-lib/teuclidian_div.c,
	tests-lib/teuclidian_div.reference, tests-lib/teuclidian_mod.c,
	tests-lib/teuclidian_mod.reference, tests-lib/tgcd.c,
	tests-lib/tgcd.reference: Added tests for div and mod. Modified
	tests for gcd.

2017-02-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/Makefile.am, tests-lib/tgcd.c, tests-lib/tgcd.reference: 
	Added tests for gcd.

2017-02-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/eucldiv.shlp, doc/eucldiv.tex, doc/euclmod.shlp,
	doc/euclmod.tex, doc/gcd.shlp, doc/gcd.tex, doc/sollya.pdf,
	doc/www/eucldiv.php, doc/www/euclmod.php, doc/www/gcd.php, help.h: 
	Finished documenting gcd, div and mod.

2017-02-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/CommandsAndFunctions.tex, doc/Makefile.am, doc/eucldiv.shlp,
	doc/eucldiv.tex, doc/euclmod.shlp, doc/euclmod.tex, doc/gcd.shlp,
	doc/gcd.tex, doc/keywords.def, doc/numberroots.shlp,
	doc/numberroots.tex, doc/sollya.pdf, doc/www/ListOfCommands.php,
	doc/www/eucldiv.php, doc/www/euclmod.php, doc/www/gcd.php,
	doc/www/numberroots.php, help.h: Started documenting gcd, div and
	mod.

2017-02-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, execute.h, expression.c, expression.h, internlexer.l,
	internparser.y, lexer.l, library.c, minilexer.l, miniparser.y,
	parser.y, sollya-library-wrappers.c, sollya.h: Continuing on the
	binding for div and mod.

2017-02-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Started working on the binding for div and mod.

2017-02-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c: Patched a memory leak in gcd.

2017-01-31  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c, polynomials.h: Started working on div and mod for
	numbers and polynomials.

2017-01-20  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Patched a performance bug with multiplication of
	polynomials.

2016-12-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/head.shlp, doc/head.tex, doc/revert.shlp, doc/revert.tex,
	doc/sollya.pdf, doc/sort.shlp, doc/sort.tex, doc/tail.shlp,
	doc/tail.tex, doc/www/head.php, doc/www/revert.php,
	doc/www/tail.php, help.h: Added some seealso in the doc

2016-12-14  Christoph Lauter <christoph.lauter@lip6.fr>

	* sturm.c: Slightly improved numberroots for particular situations.


2016-12-01  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Slight improvement on how externalprocedure receive
	their result arguments. This is because certain compilers seem to
	have trouble taking a pointer on a void * variable.

2016-12-01  Christoph Lauter <christoph.lauter@lip6.fr>

	* sollya-library-wrappers.c, sollya.h, sollya.map: Added some
	additional functionality into the Sollya library that allows the
	Python wrapper to support Python long integers.

2016-11-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog, Makefile.am: Updated ChangeLog and -version-info in
	master branch.

1331 1332 1333 1334 1335 1336 1337
2016-10-17  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: The man page was not distributed when help2man was
	not present...

2016-10-17  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1338 1339
	* doc/sollya.pdf, doc/sollya.tex, doc/www/sollya.php: Added missing
	documentation for the sollya_lib_range function
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364

2016-10-17  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Patched a small bug in an evaluation function.

2016-10-17  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: A certain cache in the polynomial code needs to be
	invalidated in certain situations.

2016-10-11  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Performance optimizations in polynomial code.

2016-10-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, expression.h, polynomials.c: Performance
	optimizations with a code that uses rational arithmetic on
	polynomials a lot.

2016-10-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Slight performance improvement to functions
	handling polynomials with rational constants.

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1365 1366 1367 1368 1369
2016-10-17  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: A certain cache in the polynomial code needs to be
	invalidated in certain situations.

1370 1371 1372
2016-10-17  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1373 1374 1375 1376 1377
	doc/introExample51.tex, doc/remez.tex, doc/sollya.pdf,
	doc/time.tex, doc/timing.tex, doc/www/bashevaluate.php,
	doc/www/bashexecute.php, doc/www/introExample51.php,
	doc/www/remez.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the documentation.
1378 1379 1380 1381 1382 1383 1384 1385

2016-10-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* NEWS: Updated NEWS file with the modification we did since last
	time.

2016-10-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1386 1387
	* doc/sollya.pdf: Reintroduced sollya.pdf that has mysteriously be
	removed in a previous commit
1388 1389 1390 1391

2016-10-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Configure checks that pdflatex comes with the
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1392 1393 1394
	required packages.  Otherwise, when pdflatex was detected, make
	tried to compile the doc when doing make all, and this could fail,
	leading `make all` to fail without even trying to compile sollya.
1395 1396 1397 1398 1399 1400 1401 1402 1403

2016-10-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check024.sollya, tests-tool/check024.sollya.reference: 
	Introducing a non-regression test for the bug patched at commit
	e34e380

2016-10-16  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1404 1405
	* Makefile.am, doc/Makefile.am, doc/sollya.pdf: In silent-mode,
	pdflatex and help2man now display nice messages.
1406 1407 1408 1409

2016-10-15  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac, tests-lib/Makefile.am: Patched a bug preventing make
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1410 1411
	check-lib to build.  Of course, I reverted the temporary patch
	committed by Christoph with commit d632607.
1412 1413 1414 1415 1416 1417

2016-10-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c: Slightly changed the logic with srand/rand in
	__unique_id generation.

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1418 1419 1420 1421
2016-10-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/sollya.pdf: Automatic recompilation of the documentation.

1422 1423 1424 1425 1426 1427 1428 1429
2016-10-15  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-lib/Makefile.am: F**ing temporary solution to make the tests
	compile again.

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* general.c: Replaced calls to rand_r by calls to rand (rand_r is
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1430 1431 1432 1433 1434
	marked as obsolete from POSIX.1-2008).  I am not sure that this is
	related to the fact that it is obsolete, but we had a warning
	'implicitely declared function' under Debian about rand_r which was
	mysterious. It did not appear clearly to me that rand_r was really
	necessary in our context, so I switched to the couple srand/rand.
1435 1436 1437 1438 1439 1440 1441 1442 1443

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Replaced a trick that we used to do before we knew
	about abs_srcdir variable

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* general.c, parser.y, sollya.1: Integrated last patch proposed by
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1444 1445
	J. Benoit.  Do not include the built information in the version and
	help texts used by help2man to generate the man page.
1446 1447 1448

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1449 1450
	* sollya.1, version.h: Integrated 7th patch proposed by J. Benoit
	Enhanced formatting of the version text, so as to generate a nicer
1451 1452 1453 1454 1455
	looking man page.

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/Makefile.am: Integrated 6th patch proposed by J.
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1456 1457 1458
	Benoit.  Simple reformatting of a long line in
	tests-tool/Makefile.am in order to break it into several shorter
	ones.
1459 1460 1461 1462

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac, doc/Makefile.am: Integrated 5th patch proposed by J.
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1463 1464 1465
	Benoit  Nicer way of compiling the documentation through make
	sollya.pdf I understood that `make install` should now install
	sollya.pdf but did not test this feature yet.
1466 1467 1468 1469 1470 1471 1472 1473

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, configure.ac: Integrated 4th patch suggested by J.
	Benoit.  Enhanced configure.ac with modern practices

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1474 1475 1476
	* doc/sollya.pdf, doc/sollya.tex: Integrated 3rd patch suggested by
	J. Benoit.  Nicer formatting of the pdf title page, licence page,
	and table of contents.
1477 1478 1479

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1480 1481
	* doc/sollya.pdf, doc/sollya.tex: Integrated 2nd patch proposed by
	J. Benoit.  Enhanced table of contents in sollya.pdf
1482 1483 1484 1485

2016-10-14  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/dirtyintegral.shlp, doc/dirtyintegral.tex,
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1486
	doc/honorcoeffprec.tex, doc/implementpoly.tex, doc/sollya.pdf,
1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497
	doc/www/dirtyintegral.php, doc/www/honorcoeffprec.php,
	doc/www/implementpoly.php, general.c, help.h, implement.c,
	infnorm.c, sollya-library-wrappers.c, sollya-messaging.h, sollya.1,
	sollya.h2m, tests-lib/timplementpoly.c,
	tests-lib/timplementpoly.reference, tests-lib/tv_implementpoly.c,
	tests-lib/tv_implementpoly.reference,
	tests-tool/check004.sollya.reference,
	tests-tool/check005.sollya.reference,
	tests-tool/check029.sollya.reference: Integrated 1st patch suggested
	by J. Benoit.  Spelling mistakes.

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530
2016-10-14  Christoph Lauter <christoph.lauter@lip6.fr>

	* internlexer.l: Patching a strange bug that made
	"showmessagenumbers = on/off" constructions to be ignored in Sollya
	files when these files were read not with read nor parse but with
	execute.

2016-10-13  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c: Trying to patch a regression in the Sollya library
	introduced with commit f175f2b938036f38a027adf2f0f381db87bd1e21

2016-10-11  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Performance optimizations in polynomial code.

2016-10-11  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Added heuristic algorithm to compute the gcd of two
	polynomials. Corrected a bug in a function used kinda everywhere.

2016-10-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, expression.h, polynomials.c: Performance
	optimizations with a code that uses rational arithmetic on
	polynomials a lot.

2016-10-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Slight performance improvement to functions
	handling polynomials with rational constants.

2016-10-10  Christoph Lauter <christoph.lauter@lip6.fr>
1531

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1532
	* polynomials.c: Changed normalization rule for gcd of polynomials.
1533 1534


CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1535 1536 1537 1538 1539 1540 1541
2016-10-10  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, execute.h, expression.h, internlexer.l,
	internparser.y, lexer.l, library.c, minilexer.l, miniparser.y,
	parser.y, polynomials.c, sollya-library-wrappers.c, sollya.h: Again
	some progress with gcd. Binding done, untested, no tests, no
	documentation.
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563

2016-10-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Added the man page to the distributed tarballs

2016-10-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* sollya.1: and the corresponding generated sollya.1...

2016-10-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Yet another small fix for help2man

2016-10-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* sollya.1: Added the manual page to the git repository.

2016-10-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, version.h: Fixed minor issues with the generation of
	the man page

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1564 1565 1566 1567
2016-10-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* expression.c, polynomials.c: Some progress with gcd.

1568 1569 1570
2016-10-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/Makefile.am, doc/bashevaluate.tex, doc/bashexecute.tex,
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1571 1572 1573 1574 1575
	doc/introExample51.tex, doc/remez.tex, doc/sollya.pdf,
	doc/time.tex, doc/timing.tex, doc/www/bashevaluate.php,
	doc/www/bashexecute.php, doc/www/introExample51.php,
	doc/www/remez.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the documentation.
1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588

2016-10-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Patching a small issue with parse(), used in
	combination with DOS-like CR/LF sequences at the end of the parsed
	string.

2016-10-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/Makefile.am, tests-tool/check031.sollya,
	tests-tool/check031.sollya.reference: Attempt for non-regression
	tests regarding correct handling of "\r\n"

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1589 1590 1591 1592 1593
2016-10-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c, polynomials.h: Starting working on polynomial gcd.


1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608
2016-10-06  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check031.sollya: Proposition of tests for the '\r\n'
	compatibility.

2016-10-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* NEWS: Slightly changed the NEWS file.

2016-10-06  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* NEWS: Concrete proposition of a NEWS file.

2016-10-05  Christoph Lauter <christoph.lauter@lip6.fr>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1609 1610 1611 1612 1613 1614 1615 1616 1617
	* internlexer.l, lexer.l, minilexer.l: Trying to make the different
	lexers silently ignore the carriage return symbol found in DOS
	formatted files before newline. I hope that this change does not
	interfere with Cygwin's (and MS Win*s POSIX subsystems) inherent
	ways of translating back-and-forth between both formatting rules, as
	the C standard mandates it to do so.  Feature untested; testing
	should be performed for all imaginable combinations of read, execute
	and parse invokations, with Sollya's original input either coming
	from a file, given in argument, or from standard input.
1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634

2016-10-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* NEWS: Updated the proposition of NEWS file with the bug fixed a
	few days ago

2016-10-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check024.sollya, tests-tool/check024.sollya.reference: 
	Added non-regression test for bug [#20818]

2016-10-04  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c, tests-tool/check013.sollya.reference,
	tests-tool/timecheck.sollya.reference: Trying to patch "[#20818]
	Expression tree changes when it is written in a file and read back".

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1635

1636 1637 1638 1639 1640 1641
2016-09-30  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Corrected a typo.

2016-09-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1642 1643
	* configure.ac: Completely reworked the configure tests for FPLLL.
	The recent change of API in fplll introduce new combinatoric
1644 1645 1646 1647
	possibilities: one might or now want to add the -std=c++11 option,
	and one might or not use the new API. This is to add to the old
	possibilities regarding the position of fplll.h in the include
	directory tree and the question of whether -lfplll flag should be
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662
	passed or not the linker.  At first, I thought that things were not
	much more difficult: the new -std requirement comes with the new API
	and that's it. It turns out that fplll API is currently not stable
	and the new API for instance was the one to use with fpllll 5.0 but
	not with the current git repository of fpllll. Nevertheless, the
	requirement of -std=c++11 is stays present with the current
	repository, although the old API is used.  To make things more
	complicated, we have examples of versions of fplll (observed when
	trying to compile under Christoph's Cygwin system) where passing
	-std=c+11 breaks the compilation while not passing it works fine.
	Also, recently Victor Magron signaled that versions of g++ around
	4.8 and 4.9 works fine with modern versions of fplll *only* is the
	option -std=c++11 is explicitely passed (although with more recent
	versions of g++, it compiles fine with or without the option).  So,
	I decided to more or less systematically test the possible
1663 1664 1665 1666 1667 1668 1669 1670 1671
	combinations of these four parameters. Some combinations really seem
	to be unlikely because they refer to behaviors of fplll at very
	different time of its history. So we indeed try only 7 possibilities
	over the 16 possibilities. Hopefully, this should allow us to
	compile sollya with more or less any version of fplll.

2016-09-26  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/timecheck.sollya: Non-regression tests for the couple
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1672 1673 1674
	of bugs hidden behind the same example.  These two bugs have been
	addressed (without a bug report) during the past three weeks.
	Contrary to what we once thought, there is absolutely no cache issue
1675 1676 1677 1678 1679 1680 1681 1682 1683 1684
	involved, so I reverted commit 1b7a727 on my way.

2016-09-26  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* sturm.c: Updated copyrights

2016-09-23  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* ChangeLog, NEWS: Selection of items to put in the NEWS of 6.0

1685 1686 1687 1688
2016-09-19  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/Makefile.am, tests-tool/check031.sollya,
	tests-tool/check031.sollya.reference: Removed non-regression test
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1689 1690 1691
	for the bug fixed at commit 9b578b8.  It seems really not easy to do
	a robust test that would also work on platforms without an X server.

1692

1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
2016-09-23  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* fpminimax.c: Better formatting of fpminimax messages at verbosity
	level >= 4

2016-09-23  Christoph Lauter <christoph.lauter@lip6.fr>

	* sturm.c: Slight performance improvement to numberroots.

2016-09-22  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c, tests-tool/timecheck.sollya: Patching the most
	difficult bug ever to be patched inside Sollya.

2016-09-21  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Revert "Patched a small bug with the reuse of a
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1710 1711
	global variable."  This reverts commit
	b5cbc68511a8a5178a8044cdf43c9e04bdbe0608.
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726

2016-09-20  Christoph Lauter <christoph.lauter@lip6.fr>

	* polynomials.c: Patched a small bug with the reuse of a global
	variable.

2016-09-19  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/timecheck.sollya: Get rid of certain caching effects.

2016-09-19  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Updated tests for min and max.

1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750
2016-09-13  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Added additional checks for min and max.

2016-09-13  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Patching a bug with infinities in min and max.

2016-09-13  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/check031.sollya: Some improvements to check031, making
	it run when no X is available and improving locking behavior.

2016-09-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/Makefile.am, tests-tool/check031.sollya,
	tests-tool/check031.sollya.reference: Kind of non regression test
	for the bug patched with commit f175f2b.

2016-09-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/timecheck.sollya,
	tests-tool/timecheck.sollya.reference: Non-regression test for a
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1751 1752 1753
	performance bug addressed at commit e0b3184.  The test currently
	does not pass as the performance issue seems to still exist,
	although it has been already much improved by commit e0b3184.
1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766

2016-09-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: Patched a memory leak in remez

2016-09-09  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check030.sollya, tests-tool/check030.sollya.reference: 
	Added non-regression tests for bugs #12892 and (parts of) #17200.

2016-09-09  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c, sollya-messages.h, sollya-messaging.h: Kind of patched
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783
	bug #12892.  The patch is not perfect, but works and will work in
	most situations, although this is not the smartest thing we could
	do. The problem was that multi-point exchange algorithm sometimes
	behaves badly when Haar condition is not fulfilled, leading to a
	situation where epsilon could suddenly decrease (instead of
	constantly increase), therefore leading to possible pseudo-cycles in
	the algorithm.  When we detect that epsilon has decreased, we now
	simply switch to the single point exchange algorithm and keep this
	choice active until the end of the algorithm. This ensures that no
	cycle can happen.  We could do better by starting over from the
	previous set of points (instead of simply continuing with the set of
	points that led to epsilon decreasing) but the patch to do it would
	be more involved, and in practice it does not seem to be very
	harmful to simply continue the algorithm with the points obtained
	after epsilon has decreased. Indeed, this patch also fixes bug
	#17200 except for one case where, precisely, it is not a good idea
	to continue the algorithm after epsilon has decreased.
1784 1785 1786 1787

2016-09-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Removed definitions of sollya_C*FLAGS in Makefile.am
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1788 1789 1790 1791 1792 1793 1794 1795 1796 1797
	(they seem useless).  Christoph introduced them when we started
	compiling the interactive tool from the library. He could not come
	with any precise argument why he put these definitions, except that
	he remembers having copied it from some more or less official
	example found on the web.  As far as I understand the logic of
	automake, this is useless because AM_C*FLAGS are automatically added
	to any target. I tried removing it and compiling sollya on a
	computer where the usual libs are installed in some prefix,
	providing them through the --with- mechanism and it worked
	perfectly.
1798 1799 1800

2016-09-07  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816
	* infnorm.c: Removed code that seems useless.  While investigating
	the bug addressed by Christoph at commit e0b3184, I found a piece of
	code I could not understand the logic of in infnorm.c. This code
	adds a complementary test to an existing test deciding whether we
	should run Newton's algorithm within dirtyinfnorm algorithm. The
	addition has been made by Christoph at commit 9b1d5e6. This commit
	performs many non trivial modifications on many files, including
	infnorm.c, but this precise modification is completely isolated (in
	particular, it is the only modification to the dirtyinfnorm
	algorithm). There is no comment that would provide us with an idea
	of the intent behing this modification, and the commit log is not
	more helpful.  I cannot see any logic in this modification, although
	it is rather harmless in the sense that it would simply conclude
	that it is necessary to run Newton's algorithm in situations where
	it does not seem quite useful.  The present commit reverts this
	modification. All the tests keep passing.
1817 1818 1819 1820 1821 1822 1823 1824 1825

2016-09-07  Christoph Lauter <christoph.lauter@lip6.fr>

	* infnorm.c: Trying to patch a strange performance bug in interval
	evaluation with low precision.

2016-09-05  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya: Changed variable names to uppercase in
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1826 1827 1828 1829 1830
	the systematic tests for objectname.  This is to clearly separate
	variable names in the generated script from variable names in the
	generating script. This is made necessary by Christoph's change of
	commit 10f890b that removed a 'restart' that would be otherwise
	required.
1831 1832 1833 1834 1835 1836 1837 1838

2016-09-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Still improving debuggability of a test.

2016-09-05  Christoph Lauter <christoph.lauter@lip6.fr>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1839 1840 1841
	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Improved test for objectname to allow for debugging of issues with
	that test in valgrind.
1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855

2016-08-31  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* external.c, infnorm.h: Updated copyrights.

2016-08-30  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/shlp2tex.sh: shlp2tex.sh now automatically adds new .tex files
	to Makefile.am

2016-08-30  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-lib/tget_element_in_list_timecomplexity.c: Made
	tget_element_in_list_timecomplexity test be skipped instead of
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1856 1857 1858 1859
	failure  Whenever this complexity check does not pass, it returns a
	code meaning SKIP instead of FAIL, so as not to disturb users or
	Debian robots. This is similar to what is performed in tests-tool
	with the MAYFAIL_TESTS
1860 1861 1862 1863

2016-08-30  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Attempt to patch configure issue with fplll on
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1864 1865 1866 1867
	Cygwin.  The changes introduced at commit ea4ede2 break configure
	with some versions of fplll on Cygwin. Indeed, the option -std=g++11
	was now systematically added when supported by the compiler, but it
	turns out that the compilation of some older versions of fplll under
1868
	Cygwin raise an error when compiled with this option, although they
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1869 1870 1871
	compile fine without the option...  This commit attempts to fix this
	issue by adding the -std=g++11 option only when the new API is
	detected, indicating that fplll is recent enough.
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882

2016-08-30  Christoph Lauter <christoph.lauter@lip6.fr>

	* fpminimax.c, general.c, plot.c: Worked around some compiler
	warnings.

2016-08-30  Christoph Lauter <christoph.lauter@lip6.fr>

	* external.c, general.c, general.h, lexer.l, plot.c: Still trying to
	fix a strange issue with fork() and the input streams used for read.

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1883

1884 1885 1886 1887 1888 1889
2016-08-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* fplll_wrapper.cpp: Support both old and new API for fplll.

2016-08-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1890 1891
	* Makefile.am, fplll_wrapper.cpp, fplll_wrapper.h, fpminimax.cpp =>
	fpminimax.c: Separated C++ from C in fpminimax.
1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904

2016-08-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Prepared configure, so as to detect rather FPLLL
	should be used with the old API

2016-08-29  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* configure.ac: Option -fPIC was given twice.

2016-08-26  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am, general.c, sollya.h2m, version.h: Minor modifications
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1905 1906 1907 1908 1909
	to restore proper behavior of help2man  It turns out that an
	innocent and small modification to Jerome Benoit's proposition when
	merging it back to the trunk at commit 972f2d6 was completely
	messing up the work of help2man. This commit restores a nice
	behavior.
1910 1911 1912 1913 1914 1915 1916 1917 1918

2016-08-25  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/check029.sollya, tests-tool/check029.sollya.reference: 
	Added non-regression test for bug #8867

2016-08-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* execute.c, implement.c, infnorm.c, infnorm.h, remez.c: (Kind of)
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1919 1920 1921 1922 1923 1924
	fixing bug [#8867] Possible infinite loop in Remez  This led me to
	introduce a new argument to uncertifiedInfnorm that allows one to
	store an argmax.  The patch is a patch over a patched algorithm, and
	all this does not really make sense any more. At least, this patch
	makes both examples reported in the bug report terminate. I have no
	proof of any kind that it completely patches the problem.
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940

2016-08-02  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: Reworked verbosity levels for some messages of remez
	algorithm.

2016-08-14  Christoph Lauter <christoph.lauter@lip6.fr>

	* execute.c: Patched a minor issue with displaying of code elements
	in match - with constructs.

2016-08-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* external.c, general.c, general.h, lexer.l, plot.c: Making the
	interworking between read and forking commands like plot extra-safe.

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1941

1942 1943 1944 1945 1946 1947 1948 1949
2016-07-27  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c: Message "function oscillates too much" is displayed at
	most once per remez step.

2016-07-27  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* remez.c, tests-tool/check029.sollya,
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1950 1951
	tests-tool/check029.sollya.reference: Fixed a subtle bug in remez
	When the function was not defined close after the right end-point of
1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
	the domain, the algorithm could fall into a more or less infinite
	loop, trying to evaluate it again and again at points not in the
	domain.  A non regression test was added.

2016-07-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* general.c, main.c, sollya-library-wrappers.c: Updated copyrights
	and checked that PDF and PHP are synchonized.

2016-07-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/Makefile.am: Makes it possible to recompile the doc within a
	VPATH build

2016-07-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/Makefile.am: Make automake happy (it gives a special meaning
	to underscores)

2016-07-13  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* doc/Makefile.am: Added source tex files into the distributed
	packages

2016-07-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* library.c: Removed a warning.

2016-07-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1982 1983
	* .gitignore: Yet another update of .gitignore  For files in
	tests-lib this time...
1984 1985 1986 1987 1988 1989 1990 1991

2016-07-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* general.c, sollya.map, tests-lib/Makefile.am,
	tests-lib/tget_element_in_list.c,
	tests-lib/tget_element_in_list.reference,
	tests-lib/tget_element_in_list_timecomplexity.c,
	tests-lib/tget_element_in_list_timecomplexity.reference: 
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
1992 1993 1994 1995 1996 1997 1998
	Time-complexity checks in tests-lib independent from Sollya library
	In order to avoid declaring symbols in sollya.map that are useless
	except for the checks of the lib, I propose the following patch:   -
	separate tests of tget_element_in_list into two separate files: one
	file is specifically for checking the time complexity.   - this time
	check is performed only on systems where gettimeofday is available
	Right now, if gettimeofday is available the test is performed and
1999
	may fail (as opposed to tests-tool/timecheck.sollya that is marked
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2000
	as 'SKIP' instead of 'FAIL' when it does not pass). Normally, the
2001 2002 2003
	check for tget_element_in_list_timecomplexity is designed to be
	really robust and measure a theoretical complexity, so I am in favor
	of really marking it as 'FAIL' in case it does not pass. But notice
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2004 2005 2006 2007
	that this behavior is easy to change if we wish.  Normally, I should
	have removed everything that had been added to the source code
	specifically to make sollya_gettime available and usable in the
	tests of the lib.
2008 2009 2010

2016-07-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2011 2012 2013 2014 2015
	* Makefile.am, libsollyaexample.c: Removed libsollyaexample.c  This
	file had been created initially to offer a panorama of how to used
	the library but it has never been really filled with something
	non-trivial.  The idea remains good in itself but would require some
	work. Currently, the official documentation does the job and the
2016 2017 2018 2019 2020 2021
	test-lib/*.c files offer a lot of practical examples, more valuable
	than the current libsollyaexample.c

2016-07-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* tests-tool/Makefile.am, tests-tool/check.proto: Less fragile
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2022 2023 2024 2025 2026 2027 2028 2029 2030
	support for timecheck  timecheck is designed to be quite robust,
	but, by its nature may fail on some unusual architectures. This
	could be misleading to people that might think that they have a
	problem with their compilation of Sollya.  An idea proposed by
	Jerome Benoit was to list it as XFAIL, but timecheck *is not*
	expected to fail: it is expected to pass, although, sometimes, it
	might not pass (notice that these situations are worth being
	reported). As a compromise, I propose in this patch simply to mark
	it as skipped when it does not pass.
2031 2032 2033 2034 2035 2036 2037 2038 2039

2016-07-12  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* .gitignore: Updated .gitignore with files produced my modern
	versions of automake

2016-07-11  Christoph Lauter <christoph.lauter@lip6.fr>

	* doc/bashevaluate.tex, doc/bashexecute.tex,
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2040 2041 2042 2043 2044
	doc/introExample51.tex, doc/remez.tex, doc/sollya.pdf,
	doc/time.tex, doc/timing.tex, doc/www/bashevaluate.php,
	doc/www/bashexecute.php, doc/www/introExample51.php,
	doc/www/remez.php, doc/www/time.php, doc/www/timing.php, help.h: 
	Recompiled the documentation.
2045 2046 2047 2048 2049 2050 2051 2052 2053 2054

2016-07-10  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Forgotten file in the EXTRA_DIST variable.

2016-07-10  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* sollya.map: Modified sollya.map in order to allow the tool to
	compile again.

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2055
2016-07-09  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>
2056

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2057 2058 2059 2060 2061 2062 2063
	* Makefile.am: Fixed a bug introduced with
	8d2d75bbafda71eb5446fcc9b0f96383a47ab370  In that commit, Jerome
	Benoit removed AM_LDFLAGS from the definition of
	libsollya_la_LDFLAGS. I do not know if this was intentional or
	accidental, but it turns out that this was necessary, e.g. to link
	against a library installed in a non-conventional place specified
	through the --with-xxx argument of configure.
2064 2065 2066 2067 2068 2069 2070

2016-06-08  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* TODO_sollya5.0: Updated TODO file

2016-07-08  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2071 2072
	* .gitignore, m4/ld-version-script.m4: Added missing definition of
	macro gl_LD_VERSION_SCRIPT
2073 2074 2075

2016-07-06  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086
	* Makefile.am, configure.ac, sollya.h2m: Fixed some problems with
	the patch introduced by Jerome Benoit  - there were a few typos -
	the patch was forcing anyone who wants to compile Sollya from the
	sources to have help2man. If it was not the case, compilation failed
	with an error. I chose to make the presence of help2man optional and
	deactivated the creation of the man page when help2man is not
	present. Alternatively, we could decide that help2man is mandatory
	(as flex and bison are) but in this case, configure must fail when
	help2man is unavailable. - I replaced the Makefile rule "%.1" into
	"sollya.1" since the '%' construction is a GNU make extension and
	there is no real reason to use it here.
2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105

2016-07-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* Makefile.am: Made the Sollya tool be built out of the Sollya
	library.

2016-07-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* plot.c: Patching a funny bug found on non-standard VT100 terminals
	(not enough memory).

2016-07-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* Makefile.am, general.c, general.h, main.c, main.h: Started working
	on making the Sollya tool be just a wrapper of the Sollya library.

2016-07-05  Christoph Lauter <christoph.lauter@lip6.fr>

	* general.c: Reintegrating Jerome's patches for Debian into Sollya
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2106 2107 2108
	Master.  There were serveral regressions induced by Jerome's patch,
	so we should run this commit by Jerome to see if he can integrate
	this commit into Debian, too.
2109 2110 2111 2112 2113 2114 2115 2116

2016-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* xml.c: Minor modif to xml.c so as not to perturb our copyright
	helper script.

2016-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2117 2118
	* libsollyaexample.c: Updated libsollyaexample.c so as to compile.
	We should definitely think of the status of this file before next
2119 2120 2121 2122 2123
	release.

2016-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* Makefile.am: Removed lines that have been commented by Jerome
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2124 2125 2126 2127 2128 2129 2130 2131 2132 2133
	Benoit.  If they are useless, there is no need to keep them as a
	comment. If they turn out to be useful for some reason, we can
	always revert the change.  Also, I did not remove the carriage
	return introduce by Jerome Benoit in Makefile.am before the
	-version-info flag. I assume he introduced it for some reason (maybe
	to prepare the use of version script?). I wonder why he removed
	$(AM_LDFLAGS) however. I would think we introduced it for some
	reason in the first place. The proof of the pudding is in the
	eating: let us see if we manage to compile everywhere without this
	parameter.
2134 2135 2136 2137

2016-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>

	* general.c, parser.y, version.h: Corrected an error introduced in
CHEVILLARD Sylvain's avatar
CHEVILLARD Sylvain committed
2138 2139 2140 2141 2142
	previous commit.  Contrary to what one could think at first sight,
	both version messages were not perfectly identical. This led to some
	globbish in the displayed message.  Now, both version messages are
	the same. Also, some indentation problem was introduced in previous
	commit and is now fixed.
2143 2144 2145 2146 2147 2148 2149 2150

2016-07-04  Sylvain Chevillard <sylvain.chevillard@ens-lyon.org>