ICGA Journal 150 CHESS ENDGAME NEWS

For nearly a year, just 16 endgames at the top of the 3-3p mountain had not had their Nalimov DTM ‘EGT’ endgame tables promulgated, despite the fact (Haworth, 2005, 2006) that Nalimov himself completed them by 9th August 2005. Bourzutschky (2006a) repaired this lacuna by 18th July 2006, converting his FEG EGTs for these endgames to the Nalimov format: he verified them by 27th July with Nalimov’s code and distributed the MD5SUM file-signatures, Nalimov-style statistics and files. On 8th August 2006, Hernandez (2006) received the ‘last 16 ’ EGTs from Nalimov, also verified and with statistics: he in turn ftp’d them to Diepeveen in Holland and Price in the USA2. Clearly, the two sources had to be seen to confirm each other. Bourzutschky (2006b) detected that, as suspected from the v20.14 CRAFTY code and shown by smaller EGT file sets, Nalimov had reverted to 8-bit cells for the KQPK(B/R)P EGTs so he converted these EGTs again3, issuing new MD5SUM data. Price (2006) manifestly verified all the Nalimov EGT files against Bourzutschky’s revised MD5SUM data and triggered the confident promulgation of both sources ’ EGT files. With two offerings rather than one compensating for the delay, a scenario familiar to London bus queues, it has been possible to validate Nalimov’s own EGTs with the independent technology of J de Koning’s FEG. Fortu-nately, the simple MD5SUM file-comparison sufficed because neither author had optimised compression by changing the EGT values for broken, i.e., obviously illegal, positions. A major milestone has assuredly bee