Huh, interesting. Thanks for the info! Maybe I'll do it after all! If I ever get around to any of this stuff, I'll let the GOG community know.

All I can say is if you really want to do it then do it. The only thing holding you back is yourself.

I played the demo, and it was quite scary. Later I tried to find it on sale in my country, but I could not find it. I own it in GOG now, but I have yet to play it.

I used something equivalent to your method, but then someone pointed out to me that it did not guaranteed a solution in every cases. And they were correct.

Imagine a simple maze that is a square inside a square inside a square. Cut an opening on the inner squares, the smallest of which holds the exit. You begin on the periphery. I think your algorithm won't work because some walls are isolated.

You're supposed to be finding your way out of a maze, not prison. :P
Yeah, I think zeogold really has a bright future if he just unleashes the beast within. Never before have I seen a game review with such raw efficiency it makes developers cry.

It's the kind of feedback that people need in order to know if they're barking up the wrong tree or creating the next sleeper hit. (But then again, it takes some common sense for that as well.)

Go zeogold!

I'm with you on most of the suggestions. Get some prototype running first, and then flesh it out over a number of iterations.

However, starting with a good code base is a necessity, so that you don't struggle with it later. But you should obey the YAGNI rule.

My desires are a bit ambitious, and I really need a very good architecture and code discipline to pull it off (the elegant code — it does not need to be perfect, but it needs to be elegant). Fortunately, I don't have a time limit as you did. I just need to keep myself motivated.

Maybe in 2 or 3 years I'll be asking for beta-testers here. ;-)

Oh, and about the levels layout, what I mean is that corridors that lead to nowhere are a bit silly. Form should follow function.
Determine your space, set-up some rooms, then connect them with corridors in a logical manner. This is how architects work. In Wolfenstein 3D, they seemed to be like a maze because that was how it was common to do at the time, and I never knew where the hell I was. It was not that fun for me, and I only had the patience to map the first level.
Also, making the player test every wall for secret passages is silly.

Whoa! Thanks, man! You know what? You guys are so nice that you've inspired me. I think I WILL do it after all. I, as a student, just need to get some schoolwork out of the way, but once I do, prepare for the Puzzlemaster to make his debut!

You're most welcome... and yes, definitely finish school. That's the way you get your foot in the door. :)

I understand what you are saying. I suppose that, for me, playing with the code is the fun part. If I get some deliverable from it, great. If I don't, at least I had fun toying with it. I doubt I will find the time needed to put out something someone will enjoy playing, let alone something that could generate revenue.

At the moment I do wonder a bit what is the point of getting another game out. Why did you do it? There are dozens of games better in every aspect. Some of them are even free! So will anyone really take advantage of your work? Is the personal satisfaction you get worth the number of work hours you spent on it?

I suppose I am feeling a bit nihilist today. :-)

I am thinking more on this ambitious game because I want to compile my ideas for it first. I do have another, simpler game in mind that I consider developing first. It should allow me to try some ideas I have, and to play with some concepts I wanted to try out (genetic algorithms, domain-specific languages and Monte Carlo tree search come first to mind). Maybe I can finish its base form in a year.