Interview
mit / with
Steen Suurballe

http://www.lokasoft.nl
http://www.lokasoft.nl/gandalf6 (details to Gandalf 6.0)
http://www.lokasoft.nl/order.htm#321 (order information)

Last update: FQ February 09th, 2005 (13:34 MEZ)

Analyse power!
Middlegame strategist!

That's GANDALF

 

Steen Suurballe, Interview
Der mit dem W"u"lf tanzt
Dance with W"u"lfes
 

by Frank Quisinsky, helps by Alexander Schmidt
http://www.uciengines.de

Translations by Prof. Dr. Djordje Vidanovic and Lisa Thul

         

Hinweis:
Eine weitere Veröffentlichung des Interviews ist erwünscht, sofern die Quellenangabe erfolgt:
Quelle: Arena Chess GUI: http://www.playwitharena.com
Ferner unter der Voraussetzung des vollständigen Abdruckes (z. B. in Printmedien)
Frank Quisinsky, den 08. Februar 2005.

Note:
Further publication of this interview is allowed and wished as long as full a indication of the sourse is made:
Source:
Arena Chess GUI: http://www.playwitharena.com
Furthermore one condition is the complete casting (e.g. inprinted media)
Frank Quisinsky, February 08th, 2005

German: Für weitere Informationen: Frank Quisinsky
English: For further information: Frank Quisinsky

 

Fragen, questions: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23

 

Vorgeschichte / Previous history:

German: Eins der stärksten und beliebtesten Schachprogramme ist Mittelspielstratege und Endspielkünstler Gandalf von Steen Suurballe (Dänemark). Das Programm von Steen wurde in den letzten Jahren nicht nur unterschätzt, sondern es mangelte auch an entsprechender sachkundiger Presseinformationen. Nach einer Entwicklungszeit von zwei Jahren wurde es zunächst ruhig um Gandalf und eine Legende schien sich zunächst in den Hirnen der sehnsüchtigen wartenden Fans zu bilden. Zum Jahreswechsel meldete sich Dänemark samt Steen zurück und atemberaubende Ergebnisse schockierte die gesamte Konkurrenz. Gandalf ist wieder da ...

English, translation by Prof. Dr. Djordje Vidanovic:
One of the strongest and most popular chess programs is the middle game strategist and the end-game wizard, Gandalf, by Steen Suurballe (Denmark). The program by Steen was apparently underestimated in the recent years, partly because it lacked adequate professional press coverage. After a two-year development period a sort of a legend formed in the minds of its long-awaiting fans. Around the beginning of this new year some breath-taking results shocked the entire competition including even its author, Steen Suurballe. We now present Gandalf ...

 

 

Interview
Der mit dem W"u"lf tanzt
Dance with W"u"lfes

01. Frank Quisinsky:

German: Auf das Interview mit Dir freue ich mich ganz besonders! Aufgrund unserer damaligen Gandalf 4.32f-h CD, welche von Gambit-Soft vertrieben wurde, und der ganzen diesbezüglichen Kontakte mit Dir bin ich selbst zu einem großen Gandalf Fan geworden.  Ich weiß gar nicht wo ich mit meinen Fragen beginnen soll! Gandalf 3.0 (DOS) ist vielleicht ein guter Startpunkt für unser Interview. Gandalf 3.0 erreichte schon vor Jahren eine sehr hohe Spielstärke und konnte sich mit den seinerzeit besten verfügbaren Schachprogrammen messen. Lediglich im Endspiel hat Gandalf seine Schwächen zum Teil sehr deutlich offenbart. Heute zählt Gandalf 6.0 zu den zähesten „Endspielkünstlern“. Hast Du Bücher über Endspiele gelesen? Du hast sicherlich einen persönlichen Berater, der Dir Material und Ideen zu Verbesserungen des Endspiels von Gandalf gegeben hat? Die Implementierung von Endspielwissen stellte bestimmt den Löwenanteil Deiner Arbeit der letzten Jahre, oder?

 

English, translation by Prof. Dr. Djordje Vidanovic:
I've been looking forward to interviewing you!  I became a big Gandalf fan during the publication of the Gambit-Soft edition of Gandalf 4.32f-h and the contacts I had with you at the time. I am not sure now where to begin with my questions…  Anyway, let's start with Gandalf 3.0 (DOS) as it seems to be a good starting point for our interview.  Already years ago Gandalf 3.0 achieved a high playing strength and could hold its own against the best chess programs at that time. It was mainly the endgame that revealed Gandalf's weaknesses. Today Gandalf 6.0 ranks among the few real "endgame artists". Did you read books pertaining to endgames? You surely have a personal advisor, who offers you ideas and advice on how to improve Gandalf's endgame? Did the implementation of the endgame knowledge constitute the bulk of your work in the past years, or was it something else?

 

 

Steen Suurballe:

The Major thing that happened between 5.1 and 6.0 is that I met Christian Kongsted, a danish journalist and very strong chess player (2506 Correspondance player). He asked me 2 read at least 2 chess books. Secrects of modern chess strategy + Fundamental Chess Endings. So the answer is yes, I have gained alot of new chessknowledge from Kongsted + books. But gandalf has not only just improved in the endgame. i believe all parts of the game has improved (positional, tactics and endgame).

 

German, translation by Lisa Thul:

Das Treffen mit Christian Kongsted, dänischer Journalist und sehr starker Schachspieler (2506 Fernschach ELO) war das Wichtigste was zwischen den Versionen 5.1 und 6.0 passierte. Er sagte mir, dass ich lediglich nur zwei Schachbücher lesen solle. 'Secrets of modern chess strategy' und 'Fundamental chess Endings'. Also meine Antwort ist: Ja. Ich erwarb einiges an neuem Schachwissen von Christian Kongsted und diesen Büchern. Aber Gandalf ist nicht nur im Endspiel verbessert worden. Ich glaube, dass alle Spielbereiche verbessert wurden (Position, Taktik und Endspiel).

 

start of page

 

02. Frank Quisinsky:

German: Ganz offenbar hat Dich die Entwicklung der eigenen GUI nach Gandalf 3.0 nicht mehr interessiert. Eigentlich kaum vorstellbar, denn bei Gandalf 3.0 konnten wir Deine diesbezüglichen Fähigkeiten bewundern. Vielleicht nutzt Du ja auch privat eine eigene Benutzeroberfläche, die Du heimlich weiterentwickelst. Verschweigst Du uns diese Information? Klären wir diesen Punkt endgültig ab !?"?"

 

English, translation by Prof. Dr. Djordje Vidanovic:
It's quite obvious that you were not interested in developing your own GUI after Gandalf 3.0. This is actually difficult to understand because you showed an admirable programming ability in Gandalf 3.0. Perhaps you use your own private interface, having further developed it for your purposes. Is this information a secret and could you clarify this point, please?

 

Steen Suurballe:

Im not at all interested in GUI's. The ones that exists are all excellent. I want to use my time on chess-programming only.

 

German, translation by Lisa Thul:

Ich bin nicht wirklich an GUI's interessiert. Alle Existierenden sind wirklich großartig. Ich bevorzuge es meine Zeit ausschließlich mit Schachprogrammierung zu verbringen.

 

start of page

 

03. Frank Quisinsky:

German: Gandalf 3.0 und leider auch unser Projekt Gandalf 4.32x konnten sich nie so richtig auf dem Markt durchsetzen. Es gab jeweils zu viele bekanntere Konkurrenzprodukte und auch die Ergebnisse der SSDF spiegelten oftmals nicht die wahre Spielstärke von Gandalf wieder. Gandalf benötigte von je her mehr Zeit um wirklich auf Touren zu kommen, mochte den eingesetzten K6 der SSDF als Pentium optimiertes Programm nicht. Leider erkannten das viel zu wenige Personen und Dein Programm wurde von bekannten Schachredakteuren in der Presse förmlich zerrissen. Konntest Du mit diesem Umstand gut umgehen?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Both Gandalf 3.0 and, unfortunately, our project Gandalf 4.32x were never generally favoured on the market. This was so even when there weren't many well-known products by the competition. However, the results of the SSDF testing eventually reflected the true playing strength of Gandalf. Gandalf needed more testing to show its real performance as the Pentium optimized program did not much like the K6 processor used by the SSDF. Unfortunately, only few people recognized the merits of your program and the well-known chess editors were literally in a dilemma how to report on Gandalf. Could you deal with these circumstance well?

 

 

Steen Suurballe:

I had no problem with the press. The most important thing is that as many people as possble like gandalf.

 

German, translation by Lisa Thul:

Ich hatte keine Probleme mit der Presse. Das Wichtigste ist, dass so viele Anwender wie möglich Gandalf mögen.

 

start of page

 

04. Frank Quisinsky:

German: Dennoch interessierten sich berechtigter Weise immer mehr Personen für Gandalf. Gandalf 4.32f-h produzierte seinerzeit die schönsten Kurzpartien gegen Konkurrenzprodukte. Überhaupt sind die Aushängeschilder von Gandalf das strategische ausgeprägte Mittelspiel und enormes positionelles Wissen. Meines Erachtens war Gandalf 4.32f seinerzeit für Langzeitanalysen fast konkurrenzlos (auch heute fährt Gandalf 6.0 die besten Listenplätze bei Test-Suits ein). Es scheint offensichtlich zu sein, dass dich diese Spielphasen am königlichen Spiel faszinieren, oder bilde ich mir das als einer Deiner größten Fans nur ein?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Regardless of the mentioned issues more and more computer chess fans got interested in Gandalf, justifiably so. At that time, Gandalf 4.32f-h produced beautiful short games against the competition products. This indicated that Gandalf possessed good strategic mid-game knowledge and vast positional understanding. My feeling was that at that time Gandalf 4.32f was a cut above the competition regarding long-term analysis (similarly, nowadays Gandalf 6.0 leads the performance lists based on test suites).  It appears that the middle game in chess has a particular fascination for you, or am I only imagining things, being a huge fan your program as I am?

 

Steen Suurballe:

I have always wanted gandalf the play agressive, and yes it is me who just "love's" a good king-attack :-)

 

German, translation by Lisa Thul:

Mein Ziel war es immer, dass Gandalf aggressiv spielt. Und ja, ich bin es, der einen guten Königsangriff liebt. :-)

 

start of page

 

05. Frank Quisinsky:

German: Ich denke es gab bei Gandalf 5.x einen Einbruch. Gandalf 5.x verlor trotz vergleichbarer Spielstärke zu Gandalf 4.32h eine Menge an Feuer. Ich kann mir nicht vorstellen, dass dies wirklich in Deinem Interesse lag! Was hattest Du speziell zwischen Gandalf 4.32h und Gandalf 5.0 geändert?

 

English, translation by Prof. Dr. Djordje Vidanovic:

I think that Gandalf 5.x marked some sort of decline. Gandalf 5.x lost its internal fire despite a comparable playing strength to Gandalf 4.32h. I cannot imagine that this was really your intention! What were the specific changes between Gandalf 4.32x and Gandalf 5.0?

 

Steen Suurballe:

In fact very little changed. Almost all of my time was spend on "porting" gandalf to the ChessPartner interface.

 

German, translation by Lisa Thul:

Tatsächlich ist wirklich wenig verändert worden. Nahezu meine ganze Zeit habe ich dazu genutzt um Gandalf an die ChessPartner Oberfläche anzupassen.

 

start of page

 

06. Frank Quisinsky:

German: Was ist wichtiger für Dich: Die reine Spielstärke oder eher die Spielweise?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Which is more important for you: pure playing strength or rather the playing style?

 

Steen Suurballe, here Steen are looking in an engine-engine match
under WinBoard by Tim Mann.

 

Steen Suurballe:

The art is to get both! High strength and in addition "entertaining" style. I believe 6.0 has both.

 

German, translation by Lisa Thul:

Die Kunst ist es Beides hinzubekommen. Hohe Spielstärke in der Kombination mit einem unterhaltsamen Stil. Ich denke, dass die Version 6.0 beides bieten kann.

 

start of page

 

07. Frank Quisinsky:

German: Und dann vergingen die Jahre und die vielen Deiner Fans taumelten desorientiert in den Foren umher und warteten regelrecht sehnsüchtig auf Gandalf News. Eine sehr lange Beta Phase zu Gandalf 6.0 lies Gandalf endgültig zu einer Legende werden. Es sickerten von einigen Testern immer mal wieder kaum glaubbare Ergebnisse durch und die Vorfreude auf Gandalf 6.0 stieg und stieg. Ein Steen Suurballe wird aber sicherlich keine kaufmännische Verkaufspointe im Sinn gehabt haben. Vermutlich warst Du mit Testergebnissen von Gandalf 6.0 Beta Versionen selbst nicht zufrieden. Hast Du Dir vielleicht zu hohe Ziele gesteckt? Warum warst Du gegebenenfalls nicht zufrieden?

 

English, translation by Prof. Dr. Djordje Vidanovic:
Then, what happened was that for two years many of your fans went through a period of frustration and real longing, expressed freely in different forums, expecting the latest on Gandalf. The long wait for Gandalf and the extended beta-testing contributed to its almost legendary status. All the time some results that were hard to believe were pouring in and the expectations soared, prompted by some testers.  This however was still not satisfying for you nor meant anything in a commercial sense.  You must have still been dissatisfied with the test results of the Gandalf 6.0 beta.  Or, possibly, you set a too ambitious a goal for yourself?  What wasn't to your liking yet?

 

Steen Suurballe:

Since 5.0/5.1 was received "badly" by the chess-community, I decided not to release 6.0 before I was 112% sure it was atleast +100 ELO better than the 5 series.

 

German, translation by Lisa Thul:

Auf Grund dessen, dass die Versionen 5.0/5.1 nicht ohne Vorbehalte von der Schachgemeinde angenommen wurden waren, entschied ich mich dafür, dass ich die 6.0 nicht veröffentliche bevor ich mir nicht 112% sicher bin, dass die 6.0 im Endeffekt +100 ELO besser sein wird als die 5er Serie.

 

start of page

 

08. Frank Quisinsky:

German: Zwischenzeitlich habe ich mich intensiver mit Gandalf 6.0 beschäftigt und bin wirklich restlos begeistert. Gandalf besinnt sich wieder auf bekannte Stärken. Strategische Zufolgen im Mittelspiel lassen Gandalf menschlich wirken. Bei Langzeitanalysen errechnet Gandalf im wahrsten Sinne des Wortes den vermeidlich besten Zug mit klaren uns aussagekräftigen Zugfolgen / Varianten. Vermutlich hast Du ganze Programmteile komplett neu programmiert? Bist Du mit neuen vielleicht noch unbekannten Verfahren am arbeiten und konntest diese nach einem Langzeittest von 2 Jahren optimieren?

 

English, translation by Prof. Dr. Djordje Vidanovic:
In the meantime I busied myself more intensively with Gandalf 6.0 and got completely inspired. Gandalf is playing again at its expected strength with strategic insight in the middle game, performing almost as a human. In long-term analysis Gandalf comes up with variations that are meaningful in a true sense. Steen, you have probably rewritten whole portions of the program? Have you implemented some completely new routines, unknown to us, and done some optimizations during this long two-year testing period?

 

Steen Suurballe:

Almost all parts of the evaluation has been rewritten + a new "careful" selective search has been made.

 

German, translation by Lisa Thul:

Nahezu alle Teile der Bewertungsfunktion sind umgeschrieben worden. Zusätzlich ist eine neue sorgfältige Selektivsuche erstellt worden.

 

start of page

 

09. Frank Quisinsky:

German: Steen, mit diesen Äußerungen kommst Du mir nicht davon :-)

Verrate uns ein wenig mehr. Es kann doch nur die Suche in Kombination mit speziellen Schachwissen sein, oder? Vielleicht eine doppelte Suchfunktion die sich gegenseitig überprüft?  So eine Art Elf (gut) und Sauron (böse). Ich kann mir nicht helfen, vermute aber, dass Gandalf bei jeder Zugberechnung die Trilogie von Tolkien reproduziert und ein blitzsauberer Ring als Ergebnis herauskommt.

 

English, translation by Prof. Dr. Djordje Vidanovic:

Steen, could you please reveal a little more? What you said wasn't too clear J. Could the explanation be that you employed good search together with special chess knowledge, or something else, highly specific for Gandalf? Perhaps some kind of meta-search that checks on the search-tree? Some kind of Elves (the good moves) and Sauron  (the bad ones)? I cannot help myself thinking that Gandalf in the course of its computation reproduces actually an imaginary Tolkienian trilogy, coming up with a good move resembling a lightning ...

 

 

Steen Suurballe:

Again, both evaluation and the search is rewritten :-)

 

German, translation by Lisa Thul:

Noch einmal: Beide Bewertungsfunktionen und die Suche sind neu geschrieben worden. :-)

 

start of page

 

10. Frank Quisinsky:

German: In Deinem Test-Team scheinen viele Hobbits zu sein. Hobbits verlassen im Normalfall niemals Ihr zu Hause, sie sind treue Seelen. Du wirst ganz sicher immer wieder auf Deine fleißigen Tester zurückgreifen können. Gab es Phasen in der zweijährigen Entwicklungszeit zu Gandalf 6.0, die Dich dennoch zurückgeworfen haben und verzweifeln ließen? Wenn es so war, wer oder was konnte Dich wieder animieren?

 

English, translation by Prof. Dr. Djordje Vidanovic:

In the beta-testing stage of developing the new Gandalf you had some faithful and thorough hobbits who never left you in the lurch. You could always count on your beta-testing hobbits and their perseverance and industry. There must have been times in the two-year stretch when you felt you were not making progress which could have made you despair.  Who was that gave you back the impulse to go on with the project, and who filled you with new creative spirit?

 

 

Steen Suurballe:

Without my testteam no 6.0 would have been made! They are really a super support with special thanks to Dan Wulff and Christian Goralski.

 

German, translation by Lisa Thul:

Ohne mein Testteam würde es keine Version 6.0 geben! Sie sind wirklich eine super Unterstützung. Mit besonderem Dank an Dan Wulff und Christian Goralski.

 

start of page

 

11. Frank Quisinsky:

German: Wie oft hast Du Herr der Ringe gelesen, bzw. Dir angesehen?

 

English, translation by Prof. Dr. Djordje Vidanovic:

How many times have you read Lord of the Rings by Tolkien and do you find it an outstanding piece of fiction?

 

Steen Suurballe:

twice :-)

 

German, translation by Lisa Thul:

Zweimal. :-)

 

start of page

 

12. Frank Quisinsky:

German: Dein Distributor Lokasoft und natürlich Du selbst haben sich zu Recht viel Lob in Computerschachforen eingehandelt. Ein „fertiges“ Produkt nicht im üblichem, regelmäßigen Abstand von einem Jahr anzubieten um letztendlich Erwartungen von Usern gerecht zu werden betrachte ich als wirkliche Stärke. Mit Ed Schröder hast Du sicherlich auch einen weiteren sehr guten Berater im Team. Ed, dieser Fuchs und Haudegen … er hatte immer schon den richtigen Riecher und hat uns alle viele Jahre mit seinen Aktivitäten verwöhnt und in Testlaune gehalten.  Führst Du einen Teil des Erfolges von Gandalf 6.0 auch auf die Zusammenarbeit mit Lex Loep und Ed Schröder zurück? Die Beiden konnten Dir sicherlich aufgrund eigener Erfahrungswerte sehr gute Testergebnisse liefern?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Both your distributor Lokasoft and naturally you yourself  have rightfully gained praise in computer chess forums. For a "finished" product it is generally usual with regard to the expected playing strength offered to its users to appear about one year after the previous version. With Ed Schroeder you must have had yet another further very good advisor in the team. Ed, the "old" fox and computer chess warhorse... had always had the right ideas and kept all of us on our toes with his activities. Do you attribute a part of the success of Gandalf 6.0 also to your co-operation with Lex Loep and Ed Schroeder? These two could surely supply you with trustworthy empirical evidence and results?

 

Steen Suurballe:

I have a very good relationship with both Ed and Lex, and they have helped me also in the test-phase, but then again without Dan Wulff + Christian Goralski, 6.0 would not have existed.

 

  Lokasoft - Home of ChessPartner

http://www.lokasoft.nl/gandalf6 (details to Gandalf 6.0)

 

German, translation by Lisa Thul:

Ich habe ein sehr gutes Verhältnis zu Ed Schröder und Lex Loep. Sie halfen mir auch in der Testphase, aber noch einmal ohne Dan Wulff und Christan Goralski würde 6.0 nicht existieren.

 

start of page

 

13. Frank Quisinsky:

German: Steen, zurück zu einer Frage die mich immer öfters beschäftigt. Auch aufgrund heutiger Möglichkeiten, die uns das Internet bietet, begrüßen wir mehr und mehr Schachprogramme auf unseren Festplatten. Die Programme von vielen talentierten und sehr jungen Programmierern erreichen heute Spielstärken die zwischen 2.400 – 2.600 ELO liegen. Du bist nun schon viele Jahre dabei und hattest trotz erbrachter Leistungen nicht das Medieninteresse auf Deiner Seite, dass andere Programmierer mit gar vermeidlich schwächeren Programmen genossen haben. Für einen jungen Programmierer ist es oftmals grausam von der Presse derart ignoriert zu werden. Was war Dein Patentrezept die Lust am Hobby niemals zu verlieren?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Steen, back to a question that has always intrigued me. Due to today's media possibilities, mostly those the Internet offers, we witness an increasing number of chess programs on our hard disks. Nowadays the programs of many talented and very young programmers reach the playing strengths between 2400 - 2600 ELO. Although you have been on the computer chess stage for a number of years you haven't enjoyed the media coverage that some weaker software got despite your program's achievements. For a young programmer as you were it could have been painful to be ignored by the press in such a manner. How did you manage not to lose the wish to go on with your work on Gandalf?

 

 

Steen Suurballe:

First of all the "love-of-the-game" and secondly believe that with years of hard work with a super test-team, anything is possible. So to the young talents: Get a team!
 

German, translation by Lisa Thul:

Als Erstes: die Liebe zu dem Spiel und als Zweites: der Glaube daran, daß mit Hilfe eines super Testteams  und der nötigen Zeit alles machbar ist. Also mein Rat an die jungen Talente: Findet euer Team!

 

start of page

 

14. Frank Quisinsky:

German: Den internen Kampf mit den Engine Protokollen hast Du erfolgreich, des Zauberers würdig, gemeistert. Du kennst beide freien Protokolle (WinBoard und UCI) wie wahrscheinlich nur wenige andere Programmierer. "Unseren" Gandalf kann eine sehr gut Kompatibilität zu den Protokollen sehr einfach nachgewiesen werden. Welches der beiden Protokolle bevorzugst Du? Möchtest Du den Protokoll Programmierern vielleicht noch einen Ergänzungswunsch oder Tipp mit auf dem Weg geben?

 

English, translation by Prof. Dr. Djordje Vidanovic:

You have won the internal fight with the engine protocol magicians with dignity. You have mastered both free protocols (WinBoard and UCI) like probably only few other programmers. "Our" Gandalf can handle the protocol compatibility quite well and it has been easily proven. Which of the two protocols do you prefer? Perhaps you have some additional comments and wishes?

 

 

Steen Suurballe:

Well when an interface is working its working. The easiest thing to implement was the UCI. Very "clean" interface, thanks to Stefan.

 

German, translation by Lisa Thul:

Also wenn die Schnittstelle einmal funktioniert, dann funktioniert sie. Das am einfachsten zu implementierende Protokoll ist UCI. Eine sehr 'saubere' Schnittstelle, danke an Stefan Mayer-Kahlen.

 

start of page

 

15. Frank Quisinsky:

German: Auch an Dir wird das Thema Chess960 nicht gedankenlos vorbeigegangen sein. Immer mehr Schachspieler erfreuen sich an Chess960 (oder auch FRC = Fischer Random Chess). Gerade aufgrund der Mittelspielstärken bei Gandalf vermute ich, dass Gandalf beim Chess960 kaum von einem GM zu bezwingen sein würde. Hast Du Dich mit Chess960 schon beschäftigt? Möchtest Du nicht erneut mit gutem Beispiel voran gehen und als TOP-Programmierer für eine Kompatibilität sorgen (wie auch schon bei Einführung vom UCI Protokoll oder beim WB Protokoll)?

 

English, translation by Prof. Dr. Djordje Vidanovic:

I think that we shouldn't skip the topic of Chess960.  An increasing number of players enjoy Chess960 (also FRC, Fisher Random Chess). I assume that due to Gandalf's strong middle game play even Grandmasters could hardly defeat it. Have you already dealt with Chess960? Wouldn't you like to go set a good example and be the vanguard of Chess960 as a TOP programmer, searching compatibility (as you did with Winboard and UCI)?

 

 

Steen Suurballe:

I really only like "real" chess :-)

 

German, translation by Lisa Thul:

Wirklich, ich liebe nur das 'wahre' Schach. :-)

 

start of page

 

16. Frank Quisinsky:

German: Ist es aufgrund der Programmstruktur von Gandalf überhaupt möglich einen Deep Gandalf zu programmieren oder würde Dich das bei Deinen sicherlich vorhandenen Verbesserungsideen zu weit zurückwerfen bzw. würde es zu viel Zeit in Anspruch nehmen?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Is the code structure of Gandalf easy to adapt to SMP (symmetric multiprocessing), so that we could have Deep Gandalf or would it be too demanding and perhaps even take you back?  Would it take too much time that you might otherwise devote to improving the program?

 

Steen Suurballe:

Well, "maybe" a deep-7 will come but I believe its very hard work and the time spent on "deep" will be lost in making chess-knowledge so I really don't know yet.

 

German, translation by Lisa Thul:

Nun gut, 'vielleicht' wird ein deep-7 kommen, doch ich glaube, es ist wirklich harte Arbeit und die Zeit die dafür geopfert werden muss, wird mir beim Erweitern des Schachwissens fehlen, von daher kann ich es zur Zeit noch nicht beantworten.

 

start of page

 

17. Frank Quisinsky:

German: Eine kurze Zeit nach dem Erscheinen von Gandalf 6.0 steht ein erstes Update zur Verfügung. Kannst Du uns kurz beschreiben was Du geändert hast? Werden weitere Updates für Deine Kunden folgen bzw. sind weitere Updates vorgesehen. Vielleicht hast Du noch eine Trumpfkarte im Ärmel und möchtest diese im Wege eines Updates zur Verfügung stellen? Interessant wäre auch mal eine gewagte Testversion als Update anzubieten, natürlich mit einem entsprechenden Hinweis ... wir schreien alle nach BETA Versionen :-)

 

English, translation by Prof. Dr. Djordje Vidanovic:

Shortly after the first appearance of Gandalf 6.0 the first update was available. Can you describe briefly what you changed? Will there be further updates for your customers and are they planned at this point? Perhaps you have still another trump card in the sleeve and would like to include it in the eventually upcoming update, just like a magician? It might be interesting to offer a daring version in an update for us who long for new beta-versions? :-)

 

 

Steen Suurballe:

Lokasoft collected various small bug-reports and when the amount has stabilized we realised 6.01 as hopefully good support. Only my test-team will see the "7" beta's :-)
 

German, translation by Lisa Thul:

Lokasoft sammelte verschiedenste kleine Fehlermeldungen und als das Ergebnis sich stabilisiert hatte, realisierten wir die 6.01 mit einem hoffentlich guten Support. Lediglich mein Testteam wird die '7' Betas zu Gesicht bekommen. :-)

 

start of page

 

18. Frank Quisinsky:

German: Sofern ich mich aufgrund unserer vielen Kontakte aus Zeiten der Gandalf 4.32f-h Entwicklung erinnere, schaust Du selbst gerne bei engine-engine Partien zu. Welche Programme setzt Du gegen Deinen Gandalf am Liebsten ein. Welche Verfahren zur Spielstärkeverbesserung setzt Du sonst ein? Vielleicht eine Sammlung von speziellen Teststellungen? Möchtest Du uns Deine beiden Lieblings-Test-Postionen verraten?

 

English, translation by Prof. Dr. Djordje Vidanovic:

If my memory serves me right from the time of our frequent contacts from during the Gandalf 4.32f-h development, you like to watch engine vs. engine games. Which program do you usually use to play against Gandalf for this particular purpose? Which methods to improve the playing strength do you prefer?  Perhaps you analyse a set of special test positions?  Could you perhaps let us know which are the two favourite sets of test positions that you use?

 

 

Steen Suurballe:

Gandalf 6.0 has one problem as i see it, and thats is Shredder. It is just such a good engine with a superp search. I almost don't use chess-positions anymore, just engine matches.
 

German, translation by Lisa Thul:

So wie ich das sehe, hat Gandalf nur ein einziges Problem und das ist Shredder. Es ist eine wirklich gute Engine mit einer fantastischen Suche. Ich nutze kaum noch Schachpositionen, sondern Engine-Wettkämpfe zum testen.

 

start of page

 

19. Frank Quisinsky:

German: An unseren persönlichen Mails habe ich immer an Dir geschätzt, dass Du Sachverhalte realistisch betrachtest. Die Situation im Computerschach werte ich aufgrund der vielen frei verfügbaren Programmentwicklungen samt Sourcen für den kommerziellen Markt als zumindest bedenklich. Wie schätzt Du die Situation ein?

 

English, translation by Prof. Dr. Djordje Vidanovic:

From our personal emails I have regarded you as a person who views circumstances realistically. How do you see the situation in computer chess in the light of the many freely available program developments including sources and what could this mean for the commercial market, could the current moment be seen as precarious for the commercials?

 

 

Steen Suurballe:

Well i really don't see myself as "super" professional. I am just happy to reach as many end-users as possible and that is alot easier when I have a "pro" like Lokasoft to help me promoting gandalf.

 

German, translation by Lisa Thul:

Nun, ich sehe mich nicht wirklich als einen 'super' Profi. Ich bin glücklich, so viele Anwender wie möglich zu erreichen. Das ist wesentlich einfacher, wenn ich ein Plus wie Lokasoft habe, die mir helfen, Gandalf zu vertreiben.

 

start of page

 

20. Frank Quisinsky:

German: Wie auch immer die Situation auf dem kommerziellen Markt ist, Gandalf ist eine wirkliche Bereicherung, glänzt durch Kompatibilitäten und Fans sämtlicher GUIs können Dein Programm nutzen und kommen voll auf Ihre Kosten. Du bist userorientiert und freust Dich selbst über erzielte Verbesserungen. Gandalf 6.0 hat sich gegenüber der Vorgängerversion um satte 100 ELO verbessert und ein solcher ELO Sprung bei einer kommerziellen Software und einer schon vorhandenen hohen Spielstärke ist äußerst selten. Wie wird es mit Gandalf weiter gehen? Vielleicht sind noch andere Pläne rund um Gandalf in Deiner Schublade, vielleicht einen "Pocket Hobbit" (Pocket PC oder ein Programm für einen Palm)?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Although Gandalf has always been in the commercial market, and an actual money-making project, its compatibilities are exceptional and various fans of all GUIs can use your program and  be fully served by it. You have a user-oriented program and make constant improvement. Gandalf 6.0 has improved over the predecessor version around 100 ELO and such an ELO jump in the commercial software is extremely rare. Where is Gandalf heading? Perhaps some other plans with Gandalf are already in your drawer, possibly a "Pocket Hobbit" version (Pocket PC or a program for a Palm)?

 

Steen Suurballe:

Well it seems "7" needs another 100 to beat Shredder, so....:-)
 

German, translation by Lisa Thul:

Nun, es scheint, dass '7' noch einmal 100 ELO bedarf, um Shredder schlagen zu können, also ... :-)

 

start of page

 

21. Frank Quisinsky:

German: Version 6.0 fällt bei kürzeren Bedenkzeiten nicht mehr zu deutlich ab als die Vorgängerversionen. Welchen Stellenwert haben Blitzpartien für Dich?

 

English, translation by Prof. Dr. Djordje Vidanovic:

Gandalf 6.0 does not lose its strength at shorter time controls the way previous incarnations did. How do you estimate the importance of short time control games?

 

Steen Suurballe:

We never play blitz when testing. We normally have the fastest time control of 40 moves in 20 minutes which seems to work the best.
 

German, translation by Lisa Thul:

Wir bedienen uns nie der Blitzfunktion beim Testen. Normalerweise nutzen wir die schnellste Zeitkontrolle von 40 Zügen in 20 Minuten, welche scheinbar am Besten funktioniert.

 

start of page

 

22. Frank Quisinsky:

German: Bei Engine Turnieren scheint Gandalf mit dem eigenen Buch generell stärker abzuschneiden als bei Turnieren mit Stellungsvorgabe. Wie hoch schätzt Du den Vorteil bedingt durch das eigene Buch ein? Verschiedene Programmierer vermuten, dass eine Engine durch ein gutes Buch um bis zu 200 ELO an Spielstärke zulegen könnte. Teilst Du diese Meinung. Sind es nicht eher so maximal 100 ELO?

 

English, translation by Prof. Dr. Djordje Vidanovic:

In engine vs. engine tournaments Gandalf seems to get more handicapped when starting its play from preset positions than from its normal book. How highly do you assess the advantage of the own book? Different programmers suspect that a good book could add up to 200 ELO to the playing strength of a program. Do you share this opinion? Is it even possible to talk about maximum 100 ELO?

 

Steen Suurballe:

Dan Wulff's is rightly super-optimized for gandalf. have much its worth in ELO i don't know. Im just happy to have Dan making me a great book!

 

German, translation by Lisa Thul:

Tatsächlich ist Dan Wulff's Buch super-optimiert für Gandalf, doch ich kann den Wert des Buches nicht in ELO einschätzen. Ich bin einfach nur glücklich, dass mir Dan Wulff ein so großartiges Buch kreiert hat.

 

start of page

 

23. Frank Quisinsky:

German: Abschließend eine Frage zum Gandalf Buch. Dan Wulf bezieht sicherlich mindestens drei Eröffnungsbuch-Renten aufgrund seiner "jahrzehntelangen" Arbeit an dem Gandalf Buch. Wenn wir mit der Zeitmaschine unterwegs wären würde uns Dan Wulff sicherlich immer wieder mal mit ein paar Buchvarianten begegnen. Das neue Gandalf Buch ist wesentlich größer geworden. Kann sich Gandalf wirklich deutlich mehr verschiedene Varianten leisten? Dan Wulf muss sehr überzeugt von Deinen Verbesserungen sein, denn sonst hätte er eine solch immense Vergrößerung des Buches wohl kaum gewagt? Wie lange arbeitet Dan genau an dem Gandalf Buch? Hast Du eine Schätzung in Stunden parat? Tauscht Ihr regelmäßig Informationen über gute oder schlechte Buchvarianten aus oder überlässt Du Dan komplett den Part des Eröffnungsbuches?

 

English, translation by Prof. Dr. Djordje Vidanovic:

In conclusion one question on the Gandalf book. Dan Wulff could obtain at least three opening book retirement funds due to his "decades" of work on the Gandalf book. If we possessed a time machine we might meet Dan Wulff again and again with a couple of book variants... The new Gandalf book has grown rather large. Can Gandalf really clearly afford more different variants?  Is the immense enlargement of the book beneficial to Gandalf and can the Magician successfully handle well the numerous extended lines? How long exactly has Dan worked on the Gandalf book? Do you contribute to the book or is it mainly Dan's share of the mutual project?

 

Dan Wulff, the opening book author from Gandalf.
Here Dan is thinking about a new killer opening line vs. AnMon!
An (ANgel) Mon (deMON) is a chess program by Christian BARRETEAU (France)

 

Steen Suurballe:

I believe Dan has worked on the book now for 11 years!!

 

German, translation by Lisa Thul:

Ich glaube Dan Wulff arbeitet jetzt seit 11 Jahren an dem Buch.

 

Dan Wulff:

Gandalf has become a lot better. The size of the opening book has nothing to do with that. The main point is to know variations, without actually having to play them actively. Of course there are a lot of new variations that are active in the book. The number of active systems has not increased that much. I've been working on the book since 1993. I try to find improvements almost every day. Difficult to figure how long I've been working on the book, several thousand hours, at least, I think. I really cannot say excactly. We (Steen, the Beta Team, and myself) often discuss variations, improvements, changes, propabilities, and so on. I have also gotten a lot of support from the Beta team, in finding really good variations for Gandalf.

 

German, Dan Wulff
Gandalf ist deutlich besser geworden. Die Größe des Eröffnungsbuches hat allerdings nichts damit zu tun. Hauptsächlich geht es darum, Varianten zu kennen aber nicht aktiv auszuspielen zu lassen. Natürlich sind auch eine Menge neue, aktive Varianten im Buch. Die Anzahl der verwendeten Eröffnungssysteme sind nicht viel größer geworden. An dem Buch von Gandalf arbeite ich seit 1993. Fast jeden Tag versuche ich Verbesserungen zu erzielen. Schwierig zu sagen wie lange ich an dem Buch gearbeitet habe, wahrscheinlich viele tausende Stunden. Genau abschätzen kann ich das nicht mehr. Wir (Steen, die Beta-Tester, und ich) sprechen oft über Varianten, Verbesserungen, Änderungen, Ausspiel-Häufigkeiten usw.. Sehr schön ist der Umstand, dass ich auch im Beta-Team Unterstützung erhalte um gute Varianten für Gandalf zu finden.

 

start of page

 

 

Frank Quisinsky:
German: Im Namen der restlichen Fans von Gandalf (und das sind nicht wenige) wünschen wir Dir und natürlich Deinem Distributor Lokasoft weiterhin viel Erfolg. Für Deine Bereitschaft zum Interview möchte ich mich bei Dir bedanken, lieben Gruß an Dan und den hart arbeitenden Hobbits :-)

 

English, translation by Prof. Dr. Djordje Vidanovic:

On behalf of the numerous remaining fans of Gandalf who aren't here with us I wish you and your distributor, Lokasoft, lots of success in the future. I would also like to thank you for your willingness to give me this interview. Best greetings to Dan and your hard-working hobbits :-)

 

http://www.lokasoft.nl
http://www.lokasoft.nl/gandalf6 (details to Gandalf 6.0)
http://www.lokasoft.nl/order.htm#321 (order information)

 

Schweich / Germany, February 08th, 2005
Frank Quisinsky

German: Webmaster können die Seite auch direkt starten unter:
http://www.playwitharena.com/directory/interviews/ssuurballe_iv.htm
Diese Seite darf in anderen Webseiten eingebunden werden.

English: Webmaster can start directly this page under:
http://www.playwitharena.com/directory/interviews/ssuurballe_iv.htm
This page can be included in other websites.

 

start of page