[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bug-gnubg] BUG: Crash using one-sided database on disk
From: |
Joern Thyssen |
Subject: |
Re: [Bug-gnubg] BUG: Crash using one-sided database on disk |
Date: |
Wed, 28 Apr 2004 18:00:33 +0000 |
User-agent: |
Mutt/1.4.2.1i |
On Wed, Apr 28, 2004 at 03:50:58PM +0100, Ian Shaw wrote
>
> Joern Thyssen [mailto:address@hidden wrote
>
> > Hmm, I can only find the 7pt to 12pt databases. There is no 13pt
> > database?! You must have generated it yourself or gotten it somewhere
> > else?!
>
>
> D'oh! I misunderstood my own directory structure. It looks like I generated
> the 8 - 13 one-sided files myself in Dec 2002.
>
> Does anyone have the correct checksum? Didn't Achim create some
> database files a while ago?
I don't know.
Can anyone list the MD5 timestamp for the 13pt database?
Morten, have you generated it?
> The first line of the file is gnubg-OS-13-15-1-1-0xxxxxxxxxxxxxxxxxxx
>
> followed by a load of gibberish, so it looks like a binary file.
That's correct. The first 40 bytes is some internal information used by
gnubg: OS = one sided, 13 = 13pt, 15 = 15 chequers, 1 = include gammon
distributions, 1 = the file is "compressed" meaning that only non-zero
values are stored, 0 = actual distribution stored rather than just
average and std.dev.
> It could very well be that the problem is in my file rather than in
> gnubg. Ought I to try regenerating it?
If you have the patience :-) If the MD5 checksum is identical to the one
you generated back in 2002, the bug must be in gnubg or makebearoff.
You're welcome to burn the database onto a DVD or a few CDs and
snail-mail it to me. This will help me debugging it :-)
I may also be able to write a small debugging program that can dump
positions from the database.
Joern