FortKnox's Journal: Muds, Graph Theory, and Open Source

OK, for those that read the "Developers! X3" JE, you know I was (poorly) asking some graph theory questions. The answers and responses were great (thanks helio!), but I have changed my design around to disclude expicit graphs.

I'm designing a mud. No, not like anything you used to play. Muds haven't been innovative since the 80s. The only mud to have any innovation is UO. I'm finding the middle ground (still text based, but having its own client and true innovation).

My original design had a graph whose nodes were a mini-2D coordinate system. This started getting really messy when I tried to find relationships between nodes. For example: a tavern would be one node, the street another, the house across the street another, etc etc etc. If I looked out of the house, would I see the tavern? What about on the second story? These were problems I was contemplating.
That's when a friend of mine came up with another idea. The first idea is very similar to the labrynths in the original Zelda. What about taking the modern approach (like UO) where you stay put, but the world moves. In other words, the entire town is one node on a graph and you have a 3D coordinate system within (3D array). That way, you do all your fancy relationships easily in the 3D array, and your graph doesn't need any fanciness attached to it.

So that's the solution. Now my next point: Open source. Most projects I mess with I open. But this mud is making me stingy. I put a lot of thought and hard work into designing this (and I'll put even more in coding it), so I don't want to immediately release it. This will sound cruel, but I don't want 16 year olds trying to install this, trying to steal my players, and trying to ask me questions like "D00D!!1 H0W D1D U J3T UR MUD 2 WERK?!?!!11" I want to make a name for myself in the cult-like mud community, THEN release my source once everyone knows where it came from and who I was.

Is this terribly egotistical? Anyone make something that they poured their sweat in and can give me a second opinion on why I should open source from the start?

I would assume

[sielwolf]

That if you decided to get complex you might need to do some ray tracing. It wouldn't be so difficult since you aren't working with floating point values (i.e. trying to hit a bounding box to test if you "see" it or not). Of course that would only be necessary if you were doing something complex. If there is only ONE thing that is to the "north" of your position, it isn't a big deal. But if you want to say "You see a street. Beyond that you see a two story house. Beyond that you see the spires of a ca

Egotistical? I think not.

[dthable]

Personally, I've done the same thing with two libraries of code - a J2EE DAO framework and my Perl Tk::Widgets. I started developing both while unemployed to keep my skills up, and when I got my current job, used them in projects at work without opening them up to the world. Some day soon, I'll look at making an open branch of the codebase, but for now, I think the company that got me off the street should have a little kickback for hiring me.

You're not egotistical or stingy.

[Dthoma]

There's nothing wrong with wanting to keep your hard work under your control. That's fine. You wrote the software, you do what you like with it.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:OSS and CSS...

[FortKnox]

Sure. I intend on making an engine so diverse, you should be able to create a full mud (including skills/attributes/etc...) through the engine itself. Everything, including constants, exist in a DB (get loaded into a singleton).
Also, I want to make it into a compiled class-jar, so you can build onto it without me having to open source it.

Also, I intend on making the design of the game open from the start. If you'd like I'll point you to the site I open up (when I finish) with my plans and design docs.

Re:OSS and CSS...

[(H)elix1]

I want to make it into a compiled class-jar, so you can build onto it without me having to open source it.

Hate to say it, but if it is java it is very easy to create reasonable java code from a class file. Snag a copy of jad [tripod.com] and run it on a class. It was solid enough that many of us on the consulting side would jad the code base rather than go through the hassles of getting source control access. It really is that good (or bad, depending on your perspective). Won't stop the determined regardless, but

Graph Theory?

[gmhowell]

Graph Theory? I'm totally lost dude. But I will gladly help alpha test on either Linux or WinXP or Win2k. (or NT4 from work;) Perhaps some documentation work.

Don't feel compelled to OSS from the beginning. I can see only two reasons to open source early/right away. First is if you find yourself either unable to code something or code it quickly. You may want and/or need the help of others. Of course, this doesn't mandate OSS. Perhaps someone in your JE group can help you on that specific bit of code. The s