Starship Titanic: The Official Strategy Guide - Softcover

Adams, Douglas

Softcover

ISBN 10: 0609801473 ISBN 13: 9780609801475

Verlag: Random House Inc, 1998

Inhaltsangabe

A strategy guide to the CD-ROM game "Starship Titanic" with solutions and hints to solving the puzzle

Über diesen Titel

Aus dem Klappentext

DE HAS:<br><br>        SOLUTIONS so complete we despise you for needing to use them!<br><br>        HINTS so subtle you've got to be a bit of a smart-ass to understand them!<br><br>        DESCRIPTIONS of natural language parsing engines and object-oriented <br>          programming by people who ACTUALLY KNOW WHAT THEY'RE TALKING ABOUT!<br><br>        UNCENSORED PHOTOS of DOUGLAS ADAMS in the VERY ACT OF WRITING!<br><br>        ILLUSTRATIONS from people who've won REAL OSCARS!<br><br>        INSIGHTS into the SECRET LIVES of PROGRAMMERS!<br><br>        NO-HOLDS-BARRED back stories to all the CHARACTERS!

Auszug. © Genehmigter Nachdruck. Alle Rechte vorbehalten.

Introduction by Douglas Adams


I'm not allowed into the programmers room any more. We are about two days from final delivery of the game. They are chasing down the very last of the bugs, making sure that the cursors all point the right way, and that there are no hitherto unnoticed glitches in the control interface. So my habit of wandering in and saying, Hey, heres a great idea, why don't we put in a bit where...? is beginning to get to them. That and the long nights they've all spent sleeping on couches or floors around the office, living off coffee, cold pizza and crumbs of encouragement from people like me saying Do you think the parrot might look better in green? They're all brilliant heroes, I salute them, and I'm not at all surprised they don't want to see me this week.



A lot of people have been saying to me, with worried, searching looks, How come the CD-ROM is six months late? The answer is very simple. Its because its a bloody CD-ROM, that's why. All CD-ROMs are six months late. At least. Its an immutable law of the universe. The only surprising thing about ours being late is that we were surprised by it. We had no idea well, certainly I had no idea of the scale of the enterprise we were undertaking.
Probably the biggest task, certainly the biggest leap into the unknown, was handling the language interaction. Nothing on this scale had ever been attempted in a computer game before. In fact, everybody whose advice we sought said we couldn't do it and would be mad to even attempt it. A good challenge. Clearly the task was beyond the scope of any one person and we recruited a small team of dialogue writers to work with me, including Neil Richards and Michael Bywater. Michael is an old friend who had already worked with me sketching out the structure of the game. Neil knew little about computers and was therefore easily fooled into taking on the task of managing and editing the dialogue, a job that dominated his life for the next year. It was huge.



We started very simply, sketching out some of the things that we imagined a player might think of saying to the characters and what the characters should say in response. We brought in actors to record the lines, programmed them into the game and tested it to see how it worked. It was pathetic. For every input line we had thought of there were ninety-nine we hadn't. So we started on a second, much, much deeper and longer set of scripts to cover what we had overlooked, and got the actors in again, this time to record reams and reams of sentences, phrases, words, numbers, names, burps, sneezes and names of chicken recipes. This all got coded into the game, and again it wasn't remotely enough. The problem was increasing like a fractal, and was rendered almost impossibly more complicated by the fact that I had recklessly decided early on that each of the Bots would have a series of different moods and behaviors, each of which had to be covered for any given situation. So for instance, the BarBot might be in charming mood or a belligerent mood, might be telling the truth or lying, and might be more or less able to come to the point quickly. Or any combination of any of the above. This was a hideous, hideous problem. Especially for Neil who did most of the work on the BarBot and became a haggard, hunted creature as we went through this write, record, test, and rewrite iteration time after time over the course of a year.



We ended up with over sixteen hours of little snippets of dialogue. Over ten thousand lines. YOU WILL NOT GET BORED TALKING TO THESE ROBOTS. Some people who have tried the game have asked if this is an exercise in artificial intelligence. No. Not even remotely. Just as when a stage magician saws his assistant in half and then joins her together again it isn't an exercise in medical science. Its a trick, an illusion, albeit a highly complex and sophisticated one. When you're playing Starship Titanic the computer doesn't understand anything. It just assesses the players input according to a simple set of rules and then chooses an appropriate output. Watch any Presidential debate and you'll get the idea. Having gone this far though, it seems a pity that we didn't just go all out for trying to achieve artificial intelligence while we were about it. Hell, there's a good couple of days before we finally freeze the code, maybe Ill just pop along to the programmers room and see if they're up for it.

Details zur Ausgabe

• Verlag: Random House Inc
• Erscheinungsdatum: 1998
• Sprache: Englisch
• ISBN 10: 0609801473
• ISBN 13: 9780609801475
• Einband: Taschenbuch
• Anzahl der Seiten: 176
• Kontakt zum Hersteller: Nicht verfügbar
• Verantwortliche Person: Nicht verfügbar