Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Open source doesn't work

Author: Dann Corbit

Date: 11:20:50 11/08/01

Go up one level in this thread


On November 08, 2001 at 11:39:59, Georg v. Zimmermann wrote:

>Hey GCP,
>
>congrats for drawing the Tiger !
>>
>>The reason why you got where you are now so fast is because
>>Ben-Dean Kawamura released DeepBug with sources under the GPL
>>license, allowing you to take it and improve it.
>>
>
>That is true. And I very much appreciate that, plus his help in getting me
>started.
>
>>That sortof takes down your whole argument doesn't it?
>>
>>It may not help you _personally_ in a _direct_ way, but that
>>doesn't mean it doesn't work.
>
>Well it helped me (I am following your argument here to simplify, one could also
>say that I am helping him, since his project that way didn't die, and his effort
>is remembered better). But I am not helping anyone it seems :(. So the glas is
>half empty I guess.

I think it is a function of several variables.  I have helped at least one dozen
chess programmers improve their engines.  I am very sure that I am not alone.
In fact, when I look at the credits for chess programs, I often find at least a
dozen names credited for helping in a single engine.  So it does work, and
better engines get produced.

Here are some problems:
1.  You are working in a fringe area.  I (for instance) could care less about
chess variants.  They are totally uninteresting to me.  I suspect a lot of
people feel similarly, but (of course) there will be those with an interest.
Working in a fringe area really cuts down on the number of people who might help
you.  Also (for the other project -- if I understand it correctly), a Winboard
clone is uninteresting to me unless it has a promise to be better.  I have yet
to see one.
2.  Your projects are not well known.  High profile projects like TSCP, Crafty,
Sjeng, etc. will get a lot of scrutiny.
3.  There are a bazillion open source projects right now.  Go to source forge
and you will see boatloads of choices.  You better stand out in some interesting
way, or you won't get the cream of the crop interested in becoming involved.
4.  You have to be willing to accept the possibility that your engine will be
only for teaching, because mostly new programmers might be your audience.

There are other problems too.  Personally, I prefer a Crafty style copyright or
(even better Berkeley or ACE style) to GPL, but that is neither here nor there.



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.