CHANGES 8.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227
  1. GNU CHESS HISTORY
  2. (current version 1.2.87)
  3. January 2, 1987 -- Stuart Cracraft at UCLA
  4. He added a few more Tal games to the collection.
  5. January 2, 1987 -- Jim Aspnes at MIT
  6. He contributed MCO variations for the Catalan,
  7. Queen's Indian, and Reti openings.
  8. December 29, 1986 -- Jim Aspnes at MIT
  9. He contributed all MCO variations of the Najdorf
  10. to the opening book.He also contributed a LISP
  11. macro (written in GNU Emacs Lisp) to convert
  12. xchess game formats to GNU Chess opening book
  13. format.
  14. December 14, 1986 -- Ken Thompson at Bell Labs
  15. He contributed almost 200 games by Tal to
  16. our collection of Tal-games, bringing the
  17. total number of Tal positions in the book
  18. to 10,692. Total book positions now 13,207.
  19. These reside in bookin, bookin.bdg, bookin.tal.
  20. Note that presently, only bookin and bookin.tal
  21. can be used. The new Tal positions came in a
  22. slightly different format, which we have chosen
  23. to adopt as our standard format. All book
  24. games in bookin and bookin.bdg will gradually
  25. change into the new standard format.
  26. December 11, 1986 -- Stuart Cracraft at UCLA
  27. Added "averages" for node-count per move,
  28. cpu per move, rate per move to list_history
  29. and write_history.
  30. New version of Xchess installed.
  31. Started typing in Tal games into "bookin.tal".
  32. Added "total book positions" printout to "book"
  33. and "enter" statistics printout.
  34. December 10, 1986 -- Stuart Cracraft at UCLA
  35. Implemented aspiration search in normal
  36. alpha-beta search. Speedups of 3% to 40%
  37. have been noticed in most positions.
  38. Occasionally a slower search will result,
  39. but it is thought these are worth the
  40. usual speedups.
  41. December 9, 1986 -- Stuart Cracraft at UCLA
  42. Fixed minor bug in write_history()
  43. Added another Tal game, 2nd game of 1st world
  44. championship match with Botvinnik, a Benoni.
  45. December 9, 1986 -- Stuart Cracraft at UCLA
  46. Enhanced parallelism. All parallel processors
  47. now communicate via a shared data file and
  48. are kept running (in idle loops watching the
  49. shared data file). This saves us a few seconds
  50. on each move since the 'rsh' need not be invoked
  51. more than once (at the beginning). Since the
  52. shared data file is now implemented, we will
  53. next work towards a "parallel anarchy" in which
  54. any processor can use any other processor in
  55. order to reduce its search. The current scheme
  56. with the program being only as fast as its slowest
  57. processor, is quite inefficient.
  58. December 1, 1986 -- ASP at MIT Athena
  59. Added a couple of Master games from
  60. Modern Chess Openings 12 (a Fischer game,
  61. and a Matanovic game).
  62. November 30, 1986 -- Stuart Cracraft at UCLA
  63. Added parallelism. Can now handle multiple
  64. processors (sharing same disk). Later we will
  65. add the capability to use processors not sharing
  66. the same disk. Modified README and MAN-PAGE.
  67. November 26, 1986 -- Stuart Cracraft at UCLA
  68. Fixed a few bugs in book-mailing mechanism.
  69. Fixed a bug regarding situations where only
  70. one move is available.
  71. Fixed a bug in read_history() that caused
  72. Black queenside castles to be mishandled.
  73. November 25, 1986 -- Stuart Cracraft at UCLA
  74. Added two pawn heuristics. Reward pawns moving into
  75. a phalanx of pawns. A phalanx is two or more
  76. horizontally-connected pawns. Likewise, penalize
  77. pawns leaving a phalanx of pawns. The penalty for
  78. leaving is a little more than the reward for
  79. entering.
  80. November 24, 1986 -- Stuart Cracraft at UCLA
  81. A user reported an unbelievable bug. Investigation
  82. of this bug led to the discovery that GNU Chess was
  83. not picking the move judged best by the tree search
  84. in all cases. This resulted in the bug showing
  85. itself which further showed that the program was
  86. selecting an inferior move. This may result in an
  87. improvement to the program's play.
  88. November 24, 1986 -- Stuart Cracraft at UCLA
  89. Added two heuristics. Penalize king moves if
  90. the king hasn't castled yet. Also, penalize pawn
  91. moves which produce doubled pawns. Should
  92. probably have something for isolated pawns
  93. too.
  94. November 23, 1986 -- Wayne Christopher at Berkeley
  95. New version of X chess display front-end.
  96. Fixed bugs include multiple pieces, runs
  97. on SUNS & Bobcats, loads saved games.
  98. November 23, 1986 -- Stuart Cracraft at UCLA
  99. Cleaned up some minor bugs regarding history.
  100. Added "Illegal command" error message at Wayne's
  101. request.
  102. November 22, 1986 -- David Goldberg at SUN Microsystems
  103. He complained that GNU Chess was memory-hungry.
  104. A few minor modifications to hash.c reduced
  105. uninitialized data space 87% and text space
  106. 12%. This should make it easier for GNU Chess
  107. to run on small computers.
  108. November 22, 1986 -- Stuart Cracraft at UCLA
  109. "read" command was working, but needed
  110. additional tweaking so that history
  111. array would be printed by list_history().
  112. November 19, 1986 -- Stuart Cracraft at UCLA
  113. Added "read" command which reads a history
  114. file (game listing) and restores the board
  115. to as if the person was still playing that.
  116. particular game. Generally cleaned up
  117. history mechanism, made it more orthogonal.
  118. Revised README. Added doc to MAN-PAGE.
  119. November 16, 1986 -- Stuart Cracraft at UCLA
  120. More opening book bugs found and fixed.
  121. Added capability to accept abbreviated-algebraic notation
  122. for entering "book" games from files.
  123. Added approximately 2500 new positions to
  124. opening book from games involving the
  125. opening called Blackmar-Diemer Gambit,
  126. a hoary line developed by Diemer in
  127. Germany years ago.
  128. November 15, 1986 -- Wayne Christopher at Berkeley
  129. He modified the move generator, resulting in
  130. a 28% speedup.
  131. November 14, 1986 -- Stuart Cracraft at UCLA
  132. He documented a lot of the GNU Chess modules
  133. with brief comments for each function. More
  134. extensive internal documentation may go in
  135. later.
  136. November 14, 1986 -- Wayne Christopher at Berkeley
  137. He created the Xchess interface for
  138. GNU Chess to have windowing with X windows.
  139. November 14, 1986 -- Stuart Cracraft at UCLA
  140. He added a "randomization" feature to
  141. the opening book. This will cause the
  142. program to select randomly from alternate
  143. variations whenever -DBEST is removed
  144. from Makefile's CFLAGS. If this is not
  145. removed, the opening play selects the
  146. first move found in the book as it appears
  147. "in order" in the human-readable book.
  148. November 14, 1986 -- David Goldberg at SUN Microsystems
  149. He responded to a query about dbm(3) which
  150. eventually resulted in the fixing of a subtle
  151. bug in the book code which was causing the
  152. program to sometimes hash to the incorrect
  153. address and thereby produce a book move which
  154. didn't even exist in the book. Thanks David!
  155. November 14, 1986 -- Stuart Cracraft at UCLA
  156. He added the "oboard" routine in util.c. This
  157. is the reverse of the already extant "iboard"
  158. (same module). These two routines translate
  159. between GNU Chess internal format and
  160. Forsythe notation.
  161. November 10, 1986 -- Stuart Cracraft at UCLA
  162. He added the "enter" command. This causes
  163. the current game to be entered in the book.
  164. Then, GNU Chess tries to mail this new entry
  165. to the book maintainers (for inclusion in
  166. the master copy of the book).
  167. November 9, 1986 -- Stuart Cracraft at UCLA
  168. He added code for an opening book. MAN-PAGE
  169. and README were modified accordingly.
  170. November 8, 1986 -- Stuart Cracraft at UCLA
  171. Checks and mates are now noticed at ply-1.
  172. This is a more complete fix to the Oct 31 fix.
  173. October 31, 1986 -- Stuart Cracraft at UCLA
  174. First attempt at fix to bug which causes
  175. program to check human's king when program
  176. itself is in check.
  177. October 31, 1986 -- Mly at MIT
  178. Reported a bug which caused program to crash
  179. when an illegal human move was played. Fixed.
  180. Also, program was unable to play as White. Fixed.
  181. October 22, 1986 -- Stuart Cracraft at UCLA
  182. Pps now rewards moves which liberate bishops.
  183. October 19, 1986 -- Stuart Cracraft at UCLA
  184. Added bitmapper routines to distribution.
  185. Added version notice.
  186. October 19, 1986 -- David Goldberg at SUN Microsystems
  187. Interfaced GNU Chess with SUN's chesstool.
  188. October 18, 1986 -- Initial release date.