TalkChess.com

Returned to the code after a time and
found some big bugs.

A code-walkthrough by either oneself
later or by others is pretty helpful!

Definitely... when you're looking at the same code every day you stop questioning things, and the mistakes just get ignored.

Another good way to find bugs is to try to explain your code to others. Quite often you'll go "so this happens this way and then... oh... wait...", finding a bug without the other person even saying anything.

rbarreira wrote:Another good way to find bugs is to try to explain your code to others. Quite often you'll go "so this happens this way and then... oh... wait...", finding a bug without the other person even saying anything.

This is one of the things that make code reviews useful in a programming job. Usually the reviewer is unfamiliar with the code, and you have to explain to them exactly what the code *actually* does. So you systematically work through all the possibilities, and it will usually be obvious if there is some edge case where the code doesn't do what you intended.

smcracraft wrote:Returned to the code after a time and
found some big bugs.

A code-walkthrough by either oneself
later or by others is pretty helpful!

By how much do you estimate play was improved from the fixes?

Are you talking about GNU Chess 5, or is this a new program?

hgm wrote:Are you talking about GNU Chess 5, or is this a new program?

IIRC, Stuart never worked on version 5, which was a completely new program in bit boards. Just the early versions.

mhull wrote:

hgm wrote:Are you talking about GNU Chess 5, or is this a new program?

IIRC, Stuart never worked on version 5, which was a completely new program in bit boards. Just the early versions.

Actually he did work on some parts of Gnu 5 like the winboard support and some of the evaluation code and maybe some other stuff which I no longer can remember .

I've been wondering for some time now if it make sense to *upgrade* Gnu 5. It seems easier to start a new engine from scratch (which is what I am doing now). There are so many new ideas that trying to fit them into Gnu is just too difficult.

Kong Sian

mhull wrote:

smcracraft wrote:Returned to the code after a time and
found some big bugs.

A code-walkthrough by either oneself
later or by others is pretty helpful!

By how much do you estimate play was improved from the fixes?

My tests are pretty basic and actually dropped very slightly in results. My former runs were on an earlier version of the development machine's OS...

I've heard that sometime results on formal tests can drop despite principal variations and move quality going up.

Knock on wood that's what happened here.

The bug was very glaring and in the quiescence search.

hgm wrote:Are you talking about GNU Chess 5, or is this a new program?

Just my own experimental program. It's bit-board-based (eval and move generator.)

Beats me 100% of the time but that doesn't say much since I'm just a patzer.

kongsian wrote:

mhull wrote:

hgm wrote:Are you talking about GNU Chess 5, or is this a new program?

IIRC, Stuart never worked on version 5, which was a completely new program in bit boards. Just the early versions.

Actually he did work on some parts of Gnu 5 like the winboard support and some of the evaluation code and maybe some other stuff which I no longer can remember .

I've been wondering for some time now if it make sense to *upgrade* Gnu 5. It seems easier to start a new engine from scratch (which is what I am doing now). There are so many new ideas that trying to fit them into Gnu is just too difficult.

Kong Sian

Antonio C. is working on a version of Fruit for GNU 6.



--Stuart