Overall Project Issues -- A Proposal for how to Proceed

[Lord God Jinnai]

Levels i don't mind. Its things like having a technology tree and a basic technology setup that i don't want to see.

In fact as the tech model lead i am insisting you not even bother coding up a tech tree model as it will eventually become too unwieldy or not make sence like civ2. Coding up a tech tree as far as i can see produces no benifits. Anything short of that you can try.
[This message has been edited by Lord God Jinnai (edited September 28, 2000).]

[Mark_Everson]

LGJ:

So long as F_Smith (If he turns out to be the tech coder) programs the standard model as it exists, he's free to do any options he wants. That is part of the deal. Even coding up his own options he's making much faster progress than anyone else we have! We will have to address what options make it into the final release when that time comes.

Beör:

The no-buildings econ allows for much more management by the player in meaningful ways that anything in any 'buildings' game I know of. Can you give examples in the main econ thread of what you think is limiting about the system?

Manu:

Thanks for all your constructive comments. I think we are heading firmly in the direction you are pushing for, even if we don't have all the details you would like.

Take Care,

Mark

[Richard Bruns]

LGJ, don't fight the options! I already made that mistake. F_Smith has already agreed that the basic game will probably use our system. The other thing is just there for scenario editing and testing purposes.

You can't stop F_Smith from coding the functionality for the tree model, because a good OO code will have that as an option anyway. I don't think he ever planned on creating the tech tree; I think he just wants to make the editor that allows mod makers to make the tree.

We will decide the shape of the final game after playtesting. But for now, we need functionality and flexibility!

[Lord God Jinnai]

I agree with functionality and flexability.

However, even for scerio options i don't think we should have it there. We need 'standards' in the final product that all scerio creators can use, but expand on, not have one do his thing and another do another it would cause chaos for people downloading these scenerios (remember you haveto consider that into the factor as well, not just what the scenerio creator himself will want).

Besides remember how unwieldy you said a tech tree would be Rich? Even with coding it in OO it will still be just as unweilding to use, maybe not to code.

[Beör]

Mark

I never said that the no-buildings approach is limiting. On the contrary, it is without limits, and that may be the reason that I am just a little concerned. I said I was afraid that it - along with the other meticulous models - would reduce the 'cause-effect' that makes Clash a game not a SimCiv. I would like to gÃve the player a feeling that when he acts, the game re-acts. And preferably not by changing some distant parameters, that might have longterm-effects on gameplay, but little effect on what is currently going on. I know this has been discussed in the Game vs Sim thread, and that there are differences of opininon as to what the group wants.

So, its not that I oppose the no-buildings approach in particular (and therefor I wont reply in the thread you suggested). Actually I think it's a rather innovative, and might be fun to play. I guess it was just an example of a general feeling I have.

(Anyway: I suspect that what is going to happen is that F_Smith is going to code it both ways . Then it will be up to the gametesting which option shoud be the default and which should be optional.)

In general I'm all for options. The OO approach renders itself very nicely to this, provided we have good OOD. This is why I think the discussions we're having in the OO Mapsquare thread are so valuable. By concentrating on basic design issues, it forces us to look much more on gameplay. What options do we really want the player to have? Where can a unit go etc?

I wish we could have the same discussion on interfaces. Personally, I am not good with interfaces - everything I do is cluttered and messy. But the general discussion of what the interface will be like, will give a very good picture of exactly what options the player has. If it's in the interface, he either can influence the parameter, or it is vital to his understanding of the state of the realm - so to speak.

So maybe someone with greater understanding of interfaces than I should start/revive an OO interface thread.

[Richard Bruns]

Look at the Demo 5 Interface thread. I recently posted an interface plan for the technology system that got lost in the shuffle and didn't get critiqued.
[This message has been edited by Richard Bruns (edited September 29, 2000).]

[puree]

ok im new here, and havnt yet been through all the pages and forums, but ill put in my twopenneth anyway.

if you want some coding done then give me some interfaces and ill implement them. i havnt had time to go through or form an opinion on how good or bad the models are, so i dont care whose approach they follow.

there is a lot of talk about what is going to be in in version 20 or 30 or 40, you need a version 1 before you can do them.

manurein is essentially correct, work out your architecture and stick with it till it breaks- if it does ( dont fix what aint broke), and get those interfaces out.


if this is going to be java/oo then how about someone putting up a class diagram ( or component diagram)for newcomers like me ( sorry if there is one that i havnt seen yet) - a picture says a thousand words.

What people say they will implement may not be what they will implement, (dont take this personally F-smith, you are putting a lot of effort into this, more than anyone else by what mark is saying) e.g. f-smith wants m-v-c , and i would agree on that, but from his posts on the mapsquare design that is not what was going to be produced ( through misunderstanding ), there should be more focused discussion between people in the same role to ensure the coders or the desingers or the architects (...) are on the same wavelength in terms of thinking and expectations understanding etc.

mark is allowing coders to add in their bits as long as they get in the expected model. i would disagree, use the time(always a valuable commodity) on what is expected, besides what is the point of implementing something that wont get used for no other reason than others dont know what candy you have added in. ( or am i totally off it here, and all the coding is being done by 1 person, who obviously knows all the extras he's put in ).

ok moaning over, what component do you want me to code. i dont care just give me the design spec and the INTERFACE.


ps a story

before i joined my current company, they had come up with an idea for a cool product. the 3-4 people working on it were all bright blokes and came up with wonderful object models ! and really got into implementing all the really good bits, eventually (about 1-1/2 years from what i gather) they thought it was just about ready - but lo and behold the whole lot fell over in a heap, they had spent so long on the little things deep down in the design , that they had neglected to get a simple running version out, and could not now fit all their bits together. The new java programmers in the company spent about 5-6 weeks starting from scratch getting and out a full working version vastly better than the original, why ? start simple, come to the extras when you know you have something to add them to. - hence my refernce near the top to version 20+.

[F_Smith]

puree:

Hey, no offense taken, I assure you.

You're completely correct, I will make mistakes. I need help/oversight. Especially in the area of graphix and 'view' code. That is not my 'bag', by any means.

There is an object design for a government testing tool I was building. I'll bump that up. Otherwise, the basic architecture is under development right now. And you've already been a big contributor. There were 4 previous demos, but it seemed a good idea to redo the basic demo architecture. So we're basically starting an 'alpha' prototype from scratch.

I also completely agree with a focus on prototyping a piece of software into production, then adding/improving the components along the way, as needed. I promise that will absolutely be the way I approach things.