Software Engineering for Maps

Well, I was just wondering: how do you guys control your projects? Do you plan it all in the beggining? If so, do you write it down with diagrams or something else? Or do you just "do it" and see what comes out? In any case, do you keep track of variables, spell names, dummy units, etc...? How? It's normal to see people showing percentage of the project, so I suppose they have it all planned, but what's the proccess you guys use to know when the project ends?

Well, just general questions. I'd love to know more if there's more to know

Most people who show percentages of work completed are pulling it out of their arse. It's not so exact of a science that percentages can be computed realistically; rather, people are reporting their intuitive feel for the progress of the map.

However, since a lot of the time and the hardest work comes in the last "ten percent" of the map, it's rather inaccurate.

Yes, I plan it out pretty carefully. I figure out what large tasks have to be accomplished and divide my planning into those tasks. When I start a task, I figure out what details have to be accomplished for the task to reach functional, and then finished, status.

As far as variable names, I name my variables following a very specific syntax so that, even when I forget what a particular variable may be, analyzing the syntax reveals its intended purpose.

I agree with Panto, the % stuff is mostly guesswork. I don't do it, and many other people don't either.

I have a very long development cycle for projects. It all starts with an idea, which may be inspired by another map, or something else completely. Then I put that idea in the fridge. I try not to act on impulse... if I did, I would end up with a bunch of half made maps which I'd start and never finish because I would loose interest. So as I work on my current project or two, I have half a dozen other ideas in my head, which I keep thinking about at times when I have nothing else to do, developing them, even thinking about how to do various details technically, how would the triggers need to work... I write some stuff down, although not too systematically, I don't do precise map specifications, just write down stuff like lists of heroes or items, balance tables where I can compare a bunch of numbers which I couldn't keep in my head without writing them down...

Then, one day, when I finish a project that I'm working on, I take a look at all the ideas I had in the meantime and try to determine which one has matured the most in that time. That's my next project. The process of work depends on what I feel like doing, I already know what the map will be like, so I can work on any part of it, I usualy do terrain first, but that's not an iron rule. Then come the triggered systems and lastly the content, heroes, items, all that stuff, but again, not in such strict order.

For me, I always draw out diagrams to help illustrate the idea for my team, I keep my variables and such in a .txt file kept with the mod, however soon I will be using locals so that wont be necessary. Anyway, in my opinion it is very important to make clear, written down, exactly what the goals of the project are. In my case I am the leader of a mod team, so I need to break up the project and delegate tasks to each member. All the more important to have a clear concise structure. Percentages are useless, the most you can do is give the map a status such as 'Concept Work' or 'Trigger Development' or 'Model Creation' etc. Even then however, at least for me, I have 5 people working on 5 different things at once.

For me, i do it piece by piece. since i'm lazy doing maps, and my schedule doesn't give me much time, yeah, slowly i do my work.

Like in my current AOS, i do 1 hero for every two days, then play it with my friends. I can say playtesting your map with friends help you not to lose interest, trust me.

Actually, I do the terrain first, then halfway down, I do heroes, items, etc. Then trigger works. I finish the terrain last, as a matter of fact, because i think it is the most boring one. But then I do the terrain best.

Now, after 4 months, I got over 65 different heroes, with triggers and everything, completed terrain, items, sounds, etc. Yeah, I can say i succeeded completing a COMPLETE map.

And I'm still working on it until now..

Lol ive never finished a map unfortunely , seems to spiral out of control theni start on a new one, but i did finish one and it turned out nice..

basically i go through this, lets say im bored at school or something, ill write down this:

Brief: what the map is about
Possible Features: what good features i may have
Terrain: What the sets gonna be
Custom models skins: am i gonna use any etc
triggers: i dont want to go overboard with triggers or itll end up in a mess

Then basically when i come to the map its very important, very very important, you do Terrainfirst, dont fall into the trap where you do little sections at a time, as this can result in much later problems, so once your happy with your terrain start your unit edditng, and triggers.

then good luck

Yeah, quite a perilous combination this one... School + Ideas for Maps... lost two days of algorithms and data structure so far hehe

Anyway, it's kinda hard to ask this in this community only, as far as I know. I was interested in knowing if people used complex processes to develop maps, but then I realized that as wc3 mapmaking is not something that will give you money, most people just won't go thru the hard working for it. And, refering to this forum again, just like the "professional" programming languages, most programmers like to program, and not write piles of documents, so we'd probably have a common sense here. But still, it's good to know

Hmmm...

Ok, I'll explain my system.

I constantly brainstorm new ideas but not new ideas for maps but new ideas for gameplay.

i.e. I thought of a backstab system, I attempted to implement it in wc3 with testing, figured out the pros and cons. And then decided whether or not to make a map out of it, and in that example I did, I made Ninja Clans.

Another example, I wondered if a space map was possible even though I can't model, I brainstormed a system of showing the game essentially using sprites and then set out to find a modeller for the base model. I did, it worked and I made Star-Factions.

Again, another example, I wanted a new spell system ( I constantly want a new spell system ); I came up with a system of interchanging items in various positions in an inventory that aligned with targetting spells, I tried it, found I could work it and made a basic map from the test map called Spellcaster. Unfortunately I never progressed the system.

So, that shows my idea creation process. How I imagine my maps, when actually implementing them I plan out unit attack tables, spells, heroes, maps, graphics, concept art even. My triggering consists of a similar system to Panto's from what I read of his, I use a consistent variable naming system although my trigger names can be rather confusing at times, I'll admit.

Creating terrain, I work from drawn maps which are carefully balanced. Hero design is usually about balance and consistent lore. Spell systems, I try to keep them engaging and more about skill than levelling.

Things like tech trees for races I design carefully but often revise my complicated systems. For example, Star-Factions originally had 5 races planned, I dropped 2 of them and added another, making 4. All their tech trees have been revised at least twice and their overall play styles have become more traditional overtime for balance.

It ends up that my plans are careful, but I often end up unsatisfied with how they mingle together. So I often edit them. Altogether my map making system is... expensive.