MD News

[RonHiler]

I have news on this front. Soon.

But first things first. I'm taking MD down from Source Forge and it will no longer be open source. The point of that was to hopefully get a little help whilst I did other things. That didn't happen So no reason for it to stay there. Just as well.

More coming when I get back from vacation next week.

[RonHiler]

Hi guys,

Well, I'm back from vacation in Florida. With me I brought roughly 73 mosquito bites. My five year old daughter had a blast with sparklers for the 4th of July, but we spent too much time outside during the dusk when the little bloodsuckers come out, and we both got eaten alive

While I was there I took a good hard look at where things were at. By things I mean Manifest Destiny, Sovereignty, Enzyme, and the DDO Character Planner. I gave a great deal of thought as to what ought to be the current priority.

The bottom line here is that I just have too many things going on at once, heh. The things I've listed are just the things you guys know about, it doesn't take into account the time I spend doing things that pay the bills, which, to be honest, comprise the majority of my time during the days (as is I'm sure true of everyone).

The easy one is the DDO character planner. It is done. There's a saying in programming circles that goes something like "No program is ever done, it is only done enough", which is SO true, heh. There are many things I still want to do with the planner, and I probably WILL do them. And of course, being based on an ever changing MMOG game, I will constantly be updating the planner until the MMOG goes away or I stop playing it (neither one of which I see happening any time soon). There will always be Planner updates. But the thing is, they are easy at this point. The basic functionality of the planner is complete, and making changes to keep up to date with the game, as well as adding the occassional upgrade is the work of a weekend here and there. The planner no longer takes up a significant portion of my time, and hasn't for a while.

Sovereignty has been back-burnered for a while now, simply because I don't have the set of tools I need to take it any further, which is why Enzyme was born in the first place

That leaves Enzyme and Manifest Destiny.

As we all know, I just finished a SERIOUSLY difficult portion of Enzyme, the triangulator. It was a major roadblock in the advancement of the project, and getting over it was quite the chore, but I managed it. It really looks like smooth sailing for Enzyme for a while until we hit another difficulty. Getting past the triangulator is good, but it also left me a bit burned out on it. I haven't moved on to the next step (which would be line segment and vertex additions) since completing the triangulator.

As for MD, nothing has been done on it since the end of 2004.

I took a really good look at MD, both it's initial incarnation and the most current one. The initial one had a lot of work done on it, to the point where we even "played" an alpha-test game. I think Pix was the game host, and I found a whole mess of bugs to fix, many of which were corrected as a direct result.

It's really a shame that version didn't continue. MOST of the difficult parts of MD had been worked out, excepting the A.I. Adding in a movement and combat A.I., and the game would have been in a playable state. And that's what I was attempting to do. I think I erred in trying to go with a neural net for the A.I. A much better choice would have been a state machine type A.I., which are not only much easier to code and customize, they fit the requirements quite a bit better. Ah well, I'm smarter now than I was back then.

With the second version, we faced a bad problem. I had let the license that I got that controlled the code to create our world expire, and I wasn't able to get ahold of the author to renew it. Thus we were faced with the task of doing it from scratch. That took quite a bit of work, and the end results were okay, but I wasn't exactly thrilled with some aspects of it (one, it took too long to create a world and two, the map tended to look a little blurry because of all the interpolation work going on).

Still, the problems with the second version could have been worked on a bit more, I think we were pretty close to having solutions I would have been happy with.

The basic problem with the second version is that I let things get out of control. There was some serious feature creep going on there. Feature creep is always a potential issue, especially when you don't work from a design doc (which I tend not to do). Really good ideas would come from me and from you guys, and we'd say "sure, let's add that in!". Seems harmless enough, but taken too far and the game continually gets ever farther from being done, heh. And that's not good for my morale

So, all of this is a long way of going about saying a couple of things.

1) Manifest Destiny is really a relatively simple game, at least in comparison to a MMOG or a game design suite. I'm thinking that if I cannot finish MD, I have no business writing the latter two. So, I'm going to back burner Sov and Enzyme for now, especially since I'm a bit burned on Enzyme anyway.

2) I'm a better coder now than I was during the initial version of MD, and even the last version. I'm going to restart MD from scratch. However, that doesn't mean the code in either of those versions is being tossed out. Rather, I'm going to use it as a base for the new version. In fact a lot of it I'm going to copy wholesale, maybe go through it once or twice to make sure it's up to my current coding standards. I'll also be pulling quite a bit of code from Sov and Enzyme. Because of that, a lot of the work should go pretty quick. There's no need to reinvent the wheel for much of the game

3) We're going to go with state machines for the enemy A.I., as well as for combat. I've worked with state machines before, and unlike Neural Nets, I'm quite confident we can get a decent A.I out of it relatively easily. It may need a lot of tweaking, but that would be true of any A.I., heh.

4) I'm in the process of writing a pseudo-design document for MD. This is going to be our bible. I'll be posting it on the site so everyone can look it over. Those who have been here since MD won't be surprised by anything they see. But I would like you guys to look it over and see if I missed anything. Right now it's at about 20 pages covering a brief overview, how the game is played, and then a more detailed look at each interface screen (of which I've written most of them). It'll probably hit around 30 to 40 pages once I'm done with that. Then I'm going to add some information on world building (what the process is), unit templates, the unit items, and technologies. I wouldn't be surprised if we hit 100 pages or more. I do NOT expect to finish the last three with the first posted version of the document.

I will fill out the document further as time goes on and as coding continues. I'm not saying I will not change a game play aspect, but I would like to keep feature creep to a minimum. We have the design for a really good game, I don't think we ought to add to much more to it than what we have now.

One last small thing. Dungeons and Dragons Online. I'm quite active in the DDO game and community. That ain't gonna change. Monday nights I play with my Dedicated Group, Wednesday nights I play with a Perma-Death group, and various other nights I play with either my solo-only character and/or my normal character. I love DDO. It may be the best game I've ever played (IMO). It *certainly* is the best MMOG I've ever played. So you just have to deal with the fact that my spare time will be split between DDO and MD coding, and that's just the way it is

So there ya go. That is the news. I'm working on that document as a first priority, and I hope to be done with the first draft within the next week or two. I'll let y'all know when it goes up. I'd be really appreciative if some people would look it over (whether or not you are familiar with the design concepts of MD) and give critiques.

Let's finish MD and get it out there!

[Strike]

woot! Very exciting news!


1) Not having a design document is the #1 problem (IMO) in the project I'm currently working on, too.

2) Being able to reuse code you've already drawn up is always a good thing Especially when redoing a project from scratch, heh.

3) State machines, eh? Well, you should be able to define the game parameters well enough for that. Add some feedback routines to adjust the transition probabilities, and you should be able to get most of the advantages we wanted from NNs... I think.

4) I'll be happy to look over the design doc, assuming I can find time to read 20-30+ pages one of these weekends I keep feeding my time to things that should be low-priority. :/

[RonHiler]

3) State machines, eh? Well, you should be able to define the game parameters well enough for that. Add some feedback routines to adjust the transition probabilities, and you should be able to get most of the advantages we wanted from NNs... I think.

Yeah, exactly. I think the two primary things that will affect the transition parameters are the unit's moral and the stack's battle orders. Of course, dying will transition you straight into the "being dead" state But I'm not going to hardcode the transition probabilities, those will definitely vary depending on those factors, maybe some others I haven't thought of yet.

[RonHiler]

Hey Shane,

At one point you did an animated logo screen thing for R&J Cyberware. It was a landscape with a grid overlay on it. I thought I had it but I can't find it now. Do you still have it? And can we use it in the game as the logo screen?

[Shane Christopher]

I'll have to hunt it down on my computer at home. Won't be until tomorrow evening though. I'm sure I have it somewhere though and can probably improve it a fair bit now.

That's what it was designed for

[RonHiler]

Cool. Do you remember if the format is AVI? Or is it something else?

[PygoraMom]

Let's finish MD and get it out there!

HOORAY!

[RonHiler]

I'm going to post the design doc (first version) on the 21st or 22nd. I'm essentially done as far as it needs to go right now. It's 38 pages, but 5 or 6 pages are blank (just headings with nothing under them).

I want to take the rest of the week to read it over and see if I missed anything. To be sure, it's not complete (like I said, I fully expect it to hit 100 pages, especially once we start listing techs, templates, and unit items), but the overview is there and and full description of all interface screens is done. I think there is plenty enough there to get started on the coding.

[Pix]

I'm going to post the design doc (first version) on the 21st or 22nd. I'm essentially done as far as it needs to go right now. It's 38 pages, but 5 or 6 pages are blank (just headings with nothing under them).

I want to take the rest of the week to read it over and see if I missed anything. To be sure, it's not complete (like I said, I fully expect it to hit 100 pages, especially once we start listing techs, templates, and unit items), but the overview is there and and full description of all interface screens is done. I think there is plenty enough there to get started on the coding.

So once you've got the design doc done, and having the previous code available to you, then it should be pretty straightforward, meaning we can be done in six months, max!