Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fritz's Tablebase Initialisation

Author: Mike Hood

Date: 16:51:33 05/05/04

Go up one level in this thread


On May 05, 2004 at 18:35:38, Robert Hyatt wrote:

>On May 05, 2004 at 18:30:11, Mike Hood wrote:
>
>>I accept your point, Bob. Reading existing files takes much more time than
>>trying to open non-existent files, so the savings from optimizing the
>>initialisation code might not be noticeable.
>>
>>Maybe my own programming background has corrupted me. I've worked 15 years
>>programming car electronics, and I've had to optimize every last assembler
>>command to avoid any delay in the fuel injection. If you're working with
>>Windows, who cares about making software run faster? Whatever savings you make
>>will be killed by the next operating system version.
>>
>>Or am I being too sarcastic?
>>
>>:)
>
>I wouldn't disagree.  I did lots of that kind of process-control stuff in the
>past, and usually the vendor wanted to use a 2 dollar part when a 3 dollar part
>would have made the programming much less time-critical...

Too true. The most extreme case I experienced was a control unit for BMW, back
in the 1980s. We were using an 8-bit processor (an 8051 derivative), but some
smart thinker realised that money could be saved by only using 4 address lines
to access the data area. So every time I read a byte I had to read the upper 4
bits, shift them, then read the lower 4 bits and do an OR. The same game in
reverse for writing a data byte. And then my bosses complained that my software
was running too slow.....



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.