[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Comment in SearchRoot

From: Simon Waters
Subject: Re: Comment in SearchRoot
Date: Thu, 20 Feb 2003 14:17:52 +0000
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2.1) Gecko/20021130

Hash: SHA1

Tim Bedding wrote:
> In search.c there is the following comment
> /* XXX It seems that ply == 1 always at this point */
> My reading of the SearchRoot function is that ply
> is set to 1 and never changed.
> Should the comment be altered?
> Perhaps
> /* We need to change the code so that ply is not always
>    1 here */
> would be better. It is hard to say because I am not
> sure of the thinking behind it.

The XXX comment are usually (always?) Lukas documenting his
thought processes, and probably should be tidied up for release
code, although I'm not promising they will be.

In this case the code block following the assertion was derived
from the code for the Search routine, where ply can be odd or
even, so the test makes sense in Search, but not in the
derivative SearchRoot routine (where as you'd expect ply is 1).

I don't feel a burning urge to tidy up SearchRoot in ways that
make it less like Search. As SearchRoot is pretty easy to
understand already but Search.....

Although I'm always open to better ways of doing things.


PS: How are people getting on with building 5.06, would it help
if I stuck a tar ball on my website?

The Cygwin build works but mops about an integer type being
defined twice - I'm happy to live with that.

All my GNU/Linux builds work.

Benchmarks suggest no significant change in strength on test
cases. I can't demonstrate any improvement in play with thinking
in opponents time, but the analysis mode works well and
encourages me to use the program more, so I'd be keen to get it
out and get onto 5.07....

Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org


reply via email to

[Prev in Thread] Current Thread [Next in Thread]