Bay 12 Games Forum

Quote from: Ehndras on April 20, 2012, 03:59:20 pm

I'm confused. What is this? Multiplayer alien-enabled X-COM? o_O Or am I confused. Yep, I'm confused.

One person controls X-COM while the other handles aliens. It's got other features not in the base game (like high explosives breaching UFOs, sniper rifles, additional exit from the skyranger etc.). Sadly it's unfinished as the developers got a cease & desist from rights holders. You can still find it in the dark corners of the internet, though I don't know what's the latest version.

Unless someone can point me at an official source that says otherwise, I'm pretty sure that they never actually got a cease & desist. Judging from what I remember from their 'so long, farewell' post on their forums, they just saw the new games coming out the X-COM IP and were starting their own game studio up and decided that they'd better play it safe than sorry. Which is really unfortunate, because they were *so close*.

I found this particularly amusing:
Web of Trust Scorecard for YourWorldInc.com

So the option you're looking for already more or less exists, but you just don't want to you use it...?

You should check out the advanced world gen. It can actually be really interesting. Lots of things to play with. Although try just changing one or two values at a time. Everything has side effects (butterflies and all that) and that sounds like what you're hitting with your ice ball / deserts.

1) Create an online interface to let people share a Fortress, or create a turn-based game
2) Command line version of the game that could be played through SSH or on the terminal
3) 3D clients that would let you interact with your world as in WoW.

For the first two, try DFTerm, it'll allow you to host a DF session of Telnet/SSH. They had an experimental flash interface, but I think they took it out.

For the third, Stonesense is probably about as close as you're going to get. It's Isometric, but it's a bit better. So far as full 3D, it's never going to happen. Or rather never going to happen again. Remember how DF's full name is Slaves of Armok Chapter II: Dwarf Fortress? Go look up Armok 1. It was abandoned, precisely because of the full 3D thing.

Of course, building an API does take work--but just look at Twitter. They spent most of their early days investing in their API, and that's arguably why they got so big (because consuming, displaying, analyzing Tweet data is just so easy!). More devs -> more users -> $$$.

It would do a lot for the DF community. Relying on hacks to hook into DF is just messy (for developers and users). There are still huge parts of the game that are essentially black boxes to any third party application.

Look up discussions about open sourcing DF for discussions about why it's unlikely that DF is ever going to be open source / have an API. I know that there are direct quotes from Toady floating around somewhere, but I can't seem to find them at the moment.

An API has been discussed to death and (unfortunately) doesn't look like it's at all high on Toady's list of things that he wants to implement.

Basically, the idea is that an API would need to potentially be updated each version (resulting in extra busy work) and that any utilities that are built on the API have the risk of being too tightly integrated so that if the maintainer ever steps down, people will get upset. Of course, this already happens to some extent with DFHack and Dwarf Therapist, although both have good people maintaining them.

You should search for some older threads on the suggestions for an API. I'm sure you'll found some rather well thought arguments both for and against. (And a good number of flames. )

Which is where the option really belongs in my option. This isn't a game level thing, this is a world generation thing. Too many options in init.txt just start making it a bit unwieldy.

Java has a very long learning curve

They lied. At least IMO, the language is a fair bit simpler than C++ (if you're using the OO parts of C++). The libraries take some getting used to, but it's not really a long learning curve. If you want something a little more off the wall, try something like Lisp or Haskell.

With Toady's experience, I'd be willing to be he could learn how to use Java at a reasonable level in a day or two. To a level where he could have written DF in it? Weeks perhaps.

[Java] uses more processing power

Depends on what you're doing with it, as mentioned earlier in the thread and at the end of this post.

Overall, it sounds like your government friend was going off the standard anti-Java propaganda. Happens all the time.

---

Quote from: Proteus on February 21, 2012, 06:52:37 am

AFAIK Java is still in many regards a little bit slower than C/C++...

This.

Since Java (byte-)code is interpret on-the-go rather than fully precompiled, I doubt it will ever be as efficient as well-coded and -optimized C++. Also, the memory handling in Java is, to put it bluntly, incredibly sloppy.

Did you even read what czolus or myself posted earlier? Java has had a Just In Time compiler for about 15 years now. Sure, it doesn't compile ahead of time, but when the compiler detects a hot spot, it compiles it down to native code.

And while "well coded and optimized C++" well likely beat Java, the same could be said for well coded and optimized Java. If you know what you're doing, you can easily beat the average developer's C++ and give even well coded C++ at least a run for its money.

Disclaimer / on topic: Porting DF to Java really doesn't make that much sense, it's a decade old code base written by someone with at least a decade (probably more) of experience in C/C++ and AFAIK, little/no knowledge of Java. For a new DF-like project, maybe. For DF, not really.

That being said, I'm curious.

AFAIK Java is still in many regards a little bit slower than C/C++...

In general sure, although in some cases (including the intense numerical computations that C/C++ traditionally dominates at) the Java JIT can actually perform better. Although the RAM issue is a valid one (although you can do a lot of the bit twiddling optimizations that DF does in Java just as well as C/C++, it's just not the normal way of doing things).

I loathe Java.

Why? Java gets a lot of hate, but at least IMO doesn't usually deserve it. There are issues with boilerplate / verbose libraries, but those are issues with C# / big C++ libraries as well. And RE: performance, the point above.

(Since it's veering a bit away from DF suggestions, if you'd rather respond via PM or the like I wouldn't mind either way.)