Traveller Generation Engine

[robject]

Thinking about: an engine for generating the core "things" in Traveller: characters, star systems and worlds, world terrain, random equipment, random cargo.

I'm thinking of a data-driven framework, with a trade-off between flexibility and ease-of-use. Something that provides the user interface, organized based on the data presented to it, that's tuned towards generating Traveller things.

Something that will make it possible for developers to add needed functionality (via data) without needing a compiler. So we will have less of "I wish we had software that did XYZ" because we've got a platform that makes it easier to add a new function that does XYZ.

I'm not thinking of an interpreter, per se, though I'm wondering if it has one.

[robject]

Simplest engine is one which can return a random value between [0,1).

Next is one which also can return the nth element from a list.
Next is one which also can return the (m,n)th element from a matrix.
Next is one which can generate a random integer based on dice and mods.
Next is one which can support named aggregates of the above.
Next is one which has basic control flow...

Yes, this does contain an interpreter after all.

But... why bother? More effort goes to creating a framework that would be better spent just coding a solution, perhaps.

[Thunderbolt]

Intriguing idea, I wonder if you could build on top of something such as the Windows Workflow Foundation. I forget the terminology now but you could build custom shapes (actions?) and then assemble them into sequential processes.

Part of me wonders if the detailed "if this, if that, but not when this" logic that appears in design sequences (such as Book 6 star system generation) would produce a workflow diagram that nightmares are made of, which code would deal with more straightforwardly.

Such a framework would probably be harder to build (because you are not just solving one particular design sequence, but a number of them), but would be more accessible to others in the long run.

[robject]

And of course the real question is "who would bother to use it?" To which I think the answer is probably: no one.

[Thunderbolt]

Originally Posted by robject

And of course the real question is "who would bother to use it?" To which I think the answer is probably: no one.

The question that has killed many of my own thoughts of building Traveller software, along with the whole copyright thing of course...

[Keklas Rekobah]

I've often wondered about what the original 'Merchant Prince' software package actually did. I only know that it was once offered on a 5-1/2" diskette, and that it operated as a DOS-based executable utility (*.exe, not *.com).

Was it essentially a software version of the trade procedure featured in the 'Merchant Prince' LBB? If so, I could probably work it out for myself.

[ronaldo1]

ive been working on something in my spare time
its open source, and i am trying to develop it for pc, mac, linux

some of it is written in python, Mono.NET

http://battle.sourceforge.net/

[coliver988]

I've messed a bit with a trade program, it's been released with truncated data files over at the MGT Aids Yahoo group. Not exactly open source, but I am willing to share source code. The class documentation is mostly here (there are issues sometimes dumping lots of files to my ISP account; and this is simply the documentation on the class, which handles ships, worlds, and actual trade, breaking the game engine from the UI [the old n-tier programming model]): http://webpages.charter.net/coliver9...put/index.html

It can handle T5 which is pretty close to Merchant Prince, as well as classic & Mongoose. I finally decided to release a version w/o the tables detailed so that the user has to have the rules book to make the tables match the actual version in use. And it is really only a tool - pretty useless on it's own.

See the .Net thread for some details on where that left off.