Author: Robert Hyatt
Date: 11:48:01 03/19/02
Go up one level in this thread
On March 19, 2002 at 14:15:37, bob o wrote: >On March 19, 2002 at 13:45:32, Gian-Carlo Pascutto wrote: > >>When the compression software was being developed, Robert did tests >>with several different blocksizes to determine which was the most efficient >>one. It turned out to be 8192 bytes. If you use a different size, your >>tablebases will be unusable by most programs. >> >>The problem of using larger blocksizes is that the whole block has to >>be read in and decompressed in order to find a single position. For large >>sizes, this will slow down the search too much. > >Prof. Hyatt: > >would it be possible to compress the files as much as possible and then provide >a small utility (probably a batch file) to decompress and recompress with >e=8192? I would be willing to write the batch file for this, although it looks >to be a 2- or 3-line file. I figure if this is done, then you could save space >on your ftp site, several gigs by my estimate. That is always doable. However, for me, the ftp space is not as important as the email traffic. And mixing blocksizes will definitely increase my email volume. :) > >Thanks! > >Bob
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.