Bredemeier:

Please tell us briefly when you started developing GNU Backgammon and how long it took you to finish it. When was it first possible to download a complete version of the program?

Wong:

I was working on backgammon software long before anything was named GNU Backgammon. I had learned the game as a child, but became especially interested about 5 years ago, after seeing a co-worker playing on FIBS. Not long after that, I began work on a simple FIBS bot (Abbott, for those who remember it), which was based on Gerry Tesauro's "pubeval". Abbott was finished in early 1998, and by October that year, I had a prototype of a custom neural network called Costello. (Costello was influenced by earlier work by Hans Berliner, Tesauro, and Eric Groleau.) In 1999, I contacted the GNU project, offering to contribute my program, and Richard Stallman dubbed it GNU software that February. By the end of 1999, several other users had downloaded it, and with their contributions, I was no longer the only developer. I wouldn't say that a complete version of GNU Backgammon has been released yet, but the first program which could be played against was released in early 1999.

Bredemeier:

How did you become involved in the project? Please describe the role you played in the development and tell us how many people you worked with. Did you work in a team?

Wong:

I worked on the initial versions alone, and the first other contributors were Jørn Thyssen and Joseph Heled, in December 1999. Since then we have been working as a loose-knit team. Jørn added cubeful evaluations (and has since worked on all kinds of additional features), and Joseph has made extensive progress with neural network training and other evaluation improvements. By the middle of 2000, Øystein Johansen had ported GNU Backgammon to MS Windows, and has been leading the Windows version ever since. Dozens of others have made other contributions, but I am afraid they are too numerous to list here.

Bredemeier:

Where do you live and how old are you? Which software projects have you worked on before, and why did you find it interesting to work on this project?

Wong:

I've lived in several places since starting work on GNU Backgammon -- I started in Auckland, New Zealand, and since then have lived in various cities in the U.S.A. (Tucson, Randolph, and Boston, to be precise). I am now 27, and will turn 28 in less than a week.

I have worked on a wide range of other software projects, ranging from retail and reservation systems for use by the hospitality industry to accounting web proxies and research projects in highly configurable systems and transport protocols.

I have found it interesting to work on GNU Backgammon largely because I love backgammon, and I love free software. I started work on it because there was no expert-level backgammon programs available as free software, and previous work (such as Tesauro's) had demonstrated that such a project ought to be feasible.

Bredemeier:

Do you like playing backgammon in your free time?

Wong:

Definitely! I often play against GNU Backgammon, and frequently play against my fiancée as well (whom I will marry in two days). I occasionally play in real-life tournaments, with mixed results.

Bredemeier:

GNU Backgammon, as a free software, has many features we only know from commercial programs such as Snowie or Jelly Fish, as, for instance, the extensive analysis tools. Where do you see GNU Backgammon in comparison with Jelly Fish or Snowie? Can a professional player ameliorate his playing skills in the same or, even, in a better way with GNU Backgammon than with Jelly Fish or Snowie?

Wong:

I have never seen any satisfactory evidence which would lead me to conclude that any of GNU Backgammon, Jellyfish or Snowie were significantly stronger or weaker than another -- I have seen several examples of one program tending to play a particular type of position better or worse than other software, but overall, the three appear very close.

As far as features go, I believe that the most recent pre-releases of GNU Backgammon provides most or all of the important features offered by JF and SW, as well as a few new ones. I expect that several more will be available by the time gnubg is eventually released, so yes, I do think that gnubg ought to be as useful a program as JF or SW to even the most serious player.

Bredemeier:

Could you assess the significance of GNU Backgammon? Tell us about user reactions. What have been the major points of critique? What has been the most positive feedback you received so far?

Wong:

I believe that the real significance of GNU Backgammon is that it brings free software to the backgammon world. Whatever contributions in makes by virtue of its playing strength or feature set pale in comparison to the fact that gnubg is free software, while all comparable predecessors were proprietary.

The majority of user reactions have been positive. Most of the criticism has been directed at the lack of good documentation; in addition, some people dislike the user interface.

While several users have written in with praise for gnubg, the kind of feedback I have personally found the most positive has been the modifications and improvements that others have contributed. It is very difficult to describe how rewarding it is to go to sleep one night, wake up the next morning and check the CVS repository, and find that a developer on the other side of the world has implemented a useful new feature overnight. This has happened repeatedly over the last few years.

Bredemeier:

In Germany, not many people know GNU Backgammon even if it corresponds to the likings and wishes of the ambitioned backgammon player in many ways. Could you imagine to turn GNU Backgammon into a commercial project in order to be able to invest money in a further development of the program, or do you rely entirely on the philosophy of the GNU Project which is to have independent developers generate a free program?

Wong:

The philosophy of the GNU Project is not necessarily incompatible with commercial interests -- for instance, hardware vendors have provided some financial support for GNU compilers in the past, and commercial operations such as Red Hat provide distributions of the GNU/Linux operating system for a fee. There is no reason why anybody who is interested in providing some sort of commercial GNU Backgammon project could not distribute copies of it and charge money for marketing and user support services. I have no personal interest in starting such a scheme, however.

We do rely entirely on the philosophy of the GNU Project in that GNU Backgammon is and will always remain free software -- where the word "free" refers to freedom, not price. Its users will always have the freedom to use, to study, to modify and to redistribute GNU Backgammon, regardless of where they got it from or whether they paid for it.

Bredemeier:

What do you think of Snowie? Would you agree to call it the world's best and strongest backgammon program? Would you say the program is cost-effective, even when users have to pay $380 for the Professional Edition?

Wong:

It is certainly conceivable that Snowie might be the best and strongest backgammon program in the world, but I can't say I would agree to give it that title without adding several qualifiers.

It seems clear that many users think SW is cost-effective, given that many have paid that $380. I am not one of them, but who am I to argue with their chequebooks?

Bredemeier:

The development of Jelly Fish is stopped. Would you see this as a proof that Snowie is holding a monopoly?

Wong:

In the sense that SW currently enjoys a large market share, I suppose that could be regarded as a monopoly. But since there are no unusual barriers to the entry of new software (such as GNU Backgammon), I don't believe that SW's position is harmful or abusive in any way.

Bredemeier:

Thinking back to the time when you developed GNU Backgammon, did you have aims and wishes you weren't able to realize? And if you had, what were they?

Wong:

No, I think that even the current pre-releases of GNU Backgammon have achieved all the important aims and wishes I had four years ago when first imagining the project. There are still many sub-goals which will require work for a long time to come, and most of these are written up in our list of things to do, available online at:

http://savannah.gnu.org/cgi-bin/viewcvs/gnubg/gnubg/TODO?rev=HEAD
Bredemeier:

Tell us about the project you are currently working on.

Wong:

The only projects I am working on are GNU Backgammon, and my wedding!

Bredemeier:

Suppose someone provided you with all the staff and money you liked, which project would you engage in?

Wong:

I would certainly continue to dedicate effort toward GNU Backgammon, which I see as an ongoing project for the foreseeable future. Beyond backgammon, an interest of mine I would like to investigate further is the GNU Hurd kernel. I do not particularly want to manage any staff, but if money were no concern, I expect I could happily work on GNU Backgammon and the Hurd for years to come.

Bredemeier:

One point of critique is that the GNU Backgammon manual is only available in English. Are you planning to change this situation? A multi-lingual manual could surely help to make GNU Backgammon more widely known, as it would help foreign players to understand the numerous features to analyse the games and matches. As a result, you would also be one step ahead of other backgammon programs which rely on the language skills of their users.

Wong:

GNU Backgammon itself is gradually becoming more multilingual as additional translations are contributed -- so far Achim Mueller and Jørn Thyssen have provided German and Danish translations, respectively.

The manual has not yet been completed in any language, but it is in much the same position as the software itself: we are happy to offer it in as many languages as we can find translators for.

Bredemeier:

Do you know the German backgammon program 'Backgammon Pro'? And if you do, how do you like it?

Wong:

I presume you are referring to the game available for Palm OS handhelds? I've heard of it, but I'm afraid I don't have access to that platform, and have never used the program.

Bredemeier:

Please tell us your favourite links.

Wong: Here are ten of them: Bredemeier:

And now, as a last point, I would like to give you the possibility to add or comment on anything which hasn't been said so far during the interview:

Wong:

I would like to thank everybody who has contributed to GNU Backgammon over the years -- primarily the developers, but also anybody who has sent us a bug report, a patch, or a helpful suggestion.

And I should probably mention that all of my responses in this interview are my opinions alone, and do not necessarily reflect the views of the other GNU Backgammon developers or the GNU Project as a whole.

Bredemeier:

Thank you very much for the interview. I should think that it will be much appreciated.

Wong:

Thank you!