Up Close and Personal with Glenn Andreas



Some folks enjoy meeting the people behind the scenes of Ambrosia, some are curious about what goes into a new program, and some are just plain nosy. So each issue of The Ambrosia Times we interview a member of the Ambrosia family.

Ambrosia Times - Hello there! And who might you be?

Glenn Andreas - How about "Glenn Andreas" from "Fridley Minnesota"
(a suburb of Minneapolis)?

AT - Ah, welcome! Please tell us a little about yourself. How did you get into
programming?

GA - It paid better than a history degree would? (I've got a history minor, which
has proved invaluable for creating the Cythera scenario, BTW, so no knock against
history degrees) I guess it was one of those things that just came naturally to me - I enjoyed it, and I was good at it.

AT - What led up to programming games?

GA - Fortune and glory. And fun. I basically see myself as a storyteller with an unconventional way of telling the story. And as a job, it is sure more fun than things like printer drivers or internet firewalls (not that I didn't learn a lot from both those pursuits, but they weren't intrinsically fun).

AT - I really like the concept of Blobbo, where did you get the idea for it?

GA - Lot's of things. I've played a number of "push things" puzzles, from Sokobon to the museum puzzle in Zork II (or was it III?), and found that they offered a certain "absolute strategy" - there were a (relatively) small amount of rules (the behavior of each object separately) that could be combined with a finite state (the 30x18 board) that would produce near infinite variety. It wasn't the fact that pushing item A causes item A to move, but the fact that item B, which use to be on top of item A can now fall, which, brushing past item C causes item C to shoot right. Basically the complex interactions between simple things. I wanted a game that has as many possible interactions like this as possible, sort of a super-set of all existing puzzle games. And that was Blobbo.

AT - Do you have an affection for puzzles? If so, what kind of puzzle do you like the most?

GA - I find them interesting, usually from the conceptual point of view (what makes them work, what makes them a challenge), rather than the actual execution of the puzzle (what you are trying to do with it).

AT - Wasn't that Rubik's Cube the coolest? (I'm guessing you dig puzzles, I dig puzzles, please bear with me)

GA - The funny thing is, the most interesting part about it wasn't the whole "align all the colors", it was "how can you rotate all the parts and not have it fall apart". It was only once I got past that part that I began to figure out what the various transformations were that were needed to solve it. At that point, actually being able to solve the puzzle wasn't very interesting - once you knew the moves to switch three pieces, rotating one of them, applying those transformations was a mechanical process. Trying to come up with those transformations, that was where the real puzzle was for me.

AT - Do you play a lot of RPG's?

GA - Actually, I play almost no games for any great period of time. Usually, I take a game, play it for an hour or two, and then never touch it again. I'm more interested to see "how did they handle representing problem A" or "what unique approach to B" did the author come up with, and I usually never allow myself to become involved in the story line to any great detail (basically, I usually stop just as it is starting to get good).

AT - What do you enjoy doing in your spare time?

GA - Hm, "spare time". Those words, they sound familiar, but I can't quite place them.

AT - What do you do for a living? ;)

GA - Nothing. Seriously, writing Delver is my full time occupation right now.

AT - Let's talk a little about Delver - how do you feel about the challenge you have accepted? I imagine this project has been in the works for a while?

GA - I love the challenge. As for this project, the first line of code was written on the Labor Day holiday in 1995, so it's been a year and a half. I've been working on and off on various role playing games for years, but they never quite measure up to what I wanted, and so they never got finished - that's one of my biggest problems is that by the time I get something partially finished, there is a better way to do it, so I throw what I had out and start over. With Delver, I decided to accept certain limitations that I'd otherwise re-engineer a dozen times.

AT - From our last contact with you, we learned that there will be many possibilities that can be accomplished in Delver (water in bucket, wine in bucket for Cajun). What preparations are you making to compensate for all of the things that are possible? Are you trying to anticipate the user's actions, or are you just offering possibilities with one right answer to proceed?

GA - I try to cover as many bases as possible - obviously I can't cover everything, but I try to cover most of the interesting ones (or the more common, though dull, ones). In general, though, except for a handleful of key plot points, I try to avoid the "one right answer" approach. My rule of thumb is "at least two right answers". For example, if there is some important information, at least two people usually to know it. If you don't have the key for a door, you can break it down. Things like that. The "you crash on an alien world, and unless you have a screwdriver that you needed to pick up at the very beginning of the game you can't open the door to get out" type of puzzle is right out. Instead, if you had the screwdriver, fine, you can open the door, but if you don't, well you can take a crowbar to the window and break it open, or fire a disintegration ray at the door, etc... Again, take some simple behavior of simple things, and allow complex interaction to build "naturally" from them.

AT - Will you be pulling from your puzzle background in order to make the adventures in Delver that much more challenging?

GA - Not so much, because detailed puzzles don't usually fit within the context of an interesting story. There will probably be a few smaller puzzles (like how to bake bread), but, for example, you won't have to solve a Rubik's Cube to get a door open. I'm trying to focus on the story, and the interaction among the characters. The goal is more to make an interesting world with lots of stuff to get the player hooked. It is only at this point that they discover that they are caught up in a story that is full of twists and turns. The more they play, the more I want them to start to form theories about what is going on, and who is doing what. The puzzles will be more social, plot based, even emotional, rather than mechanical.

AT - The first adventure, Cythera, will only be the beginning, correct? Do you have plans to continue releasing adventures, or will you spend some time creating an editor for others to add to your engine?

GA - Actually, both. I've already got the basic "two sentence summary" for the Cythera sequel, and I've already put a couple of subtle clues in the first adventure, but I'd really like to rewrite the engine. One of the biggest limitations of Delver is that it is essentially a 2D engine - I want a real 3D engine. Something where you could stack crates up to form a makeshift stairway to reach something up high, for example.

On the other hand, I'd like to see others create scenarios using the existing engine. It is safe to say, however, that I probably won't be releasing the editor to the general public - not that the editor isn't ready for prime time (well, it isn't but that's more besides the point), or even that everything else isn't either - things like documentation on the editor, the run time system, system calls (I seem to add a couple every other week or so), and the like. The biggest is that this is just far too complex for the general public - you end up writing code (in a special scripting language) for just about everything (the only stuff that is hard-coded into the engine are things specifically related to rendering and window drawing & layout). Things like combat, magic systems, even examining a sign, these are all in the scripting language. The amount of support that would be needed for this would be overwhelming. The most likely way we'll handle this is to have a handful of "certified developers" who would be selected based on their ability to create this sort of thing.

AT - You have had some experience with shareware in the past, how do you feel about Ambrosia's shareware strategy?

GA - I really like Ambrosia's use of quality to separate themselves from the rest of the pack. I am so sick of seeing "gosh, I whipped this up in Hypercard over the weekend - if you like it, send me $10". I won't even begin to list the programs that I download that crash when I try them out. There are other bad examples - things like a word processor where you can't save anything, and it auto-quits after 15 minutes, or an appointment manager that only works for the first two weeks. You need to provide something of high quality and intrinsic value to get people to download and use something - and it needs to actually be usable in real-world situation, or what's the point.

Ed Note -- We are real happy to be working with Glenn, and are even more excited with his RPG engine. It's amazing what it can do. You can see a preview of Delver and Cythera elsewhere in this issue. We hope that Glenn finds some spare time soon, the nice wether is just around the corner for us northerners.