maybe the answers to this question will be useful for me

[Oriebam]

Have you been working on a complex project? For how long? Did learn things as you went, successfully or did you have knowledge of things beforehand/stopped to learn? Did you give up(why?) ?

 

[Zed]

A complex project, yes. In IT. Developed a thingy for a hospital. Learned a lot. The best lesson I learned was "start with the essentials". If you got a userbase, what are the essential features. Start with those before anything else.

A complex game project, no.

 

[shihonage]

Have you been working on a complex project? For how long? Did learn things as you went, successfully or did you have knowledge of things beforehand/stopped to learn? Did you give up(why?) ?

I gave up on one project. It was a realtime stockmarket analyzer.

I spent 1 year on it. It lost $4000 on the stockmarket. In retrospect, I should've ignored a lot of literature on the subject and wrote algorithms intuitively. Kept it simple. By the end, it was too fundamentally wrong, there was too much to redo, and I dropped it. I went in with too many preconceived expectations, and also, too little experience with the stockmarket.

Monsterland took 1 year initially, it was my first serious game design/code project, iterative, and as such it evolved organically into something that felt "fun". It became a hard-hitting lesson on how writing spaghetti code wastes time.

Shelter started development in January 2000, but it became serious in 2007 when it went isometric and NMA put a video of it on their front page. It is, by far, the most complex thing I've ever worked on, and also a massive learning experience.

All these things people take for granted, like knowing what the mouse is currently over, have to be coded. Nothing happens by itself. Optimizations can be crucial even early on, otherwise you just waste too much time waiting for elementary things to happen hundreds of times.

If the previous steps are inelegant, future steps become impossible. If you have enough functionality to do a mockup in an hour, doing a real version of ANY sort of content (map, quests) will take FOREVER. The mockups are the illusion. You follow this illusion, under-develop your content generation tools, and sink in the quicksand the moment something real starts to happen.

This is why, after that first NMA video, I spent a shitload of time on dynamic sprite management, map editor that classifies tiles by materials, themes and facings, ShelterScript. And after the video with 3 quests, seeing how quickly it became unmanageable, I spent a year on dialogue/quest compiler.

 

[Forgotten Friend]

I've been making a game a really long time now. It's not so much an RPG as a tactical game using a 3D engine, which has all the best tactical and UI features of Jagged Alliance, Silent Storm, and to some extent the original X-Com. At least what I see as the best.

All those features are done for the most part and the game is playable and starting to seem fun, but even now I have a lot left to do. The biggest part that still sucks is the AI which is really crap so far but I recently change the orders system so that it works on a squad by squad system and so that squads patrol around, some squads guard locations, etc. It seems in almost all games the way it's handled is to have scripted AI that goes through the editor but I realized it's an immensely gay way to do things. In fact I'd say that most games don't have real AI at all in the traditional sense. It's not really solving a closed system but a bunch of halfassed scripts that hopefully work out worth a damn. I'm hoping my own AI will work out better and I'm encouraged but it's not something easy to see until it's done.

Game system is flexible and allows you to do virtually anything. You can even add in new stats, and they will work. It's an easy process because everything is built on stats behind the scene which are then added to the created stats as a bonus. So for example firing accuracy, health, fatigue, can all be affected by a stat. So you define something like perception and give it a bonus rate to affect firing accuracy or whatever else you want.

I have about 50 perks defined as well, and they work the same exact way as the stats. You can define how to get perks as well, ie how often per level.

For the game system itself the balance is really wacked out. Every time I get down to it I end up nerfing it because I still have stuff to code and therefore need to be able to go through combat very fast. I don't have facing implemented in any way. Since motion is continuous and not discreet, meaning there's no grid, it makes it hard to decide the best way to handle this, if at all.

As for learning stuff, I was already pretty uber as a programmer. For art it's another story since I've been doing all my own stuff. It doesn't look half bad if I say so myself, but there's a lot more to be done. Time it's taken doesn't surprise me, the only surprise is that I've stuck to it and that I have not had any major difficulties like I feared I would.

I've been feeling my will to work evaporate the last 6-9 months just because I've been working on it so fucking long but I still hope to release a demo by the end of 2012.

In b4 cries of troll or Prosper. Maybe I will call my company Prosperous Games.

 

[Destroid]

You could post some screenshots.

 

[Forgotten Friend]

I cropped out the interface, this is just a teaser to show off the way the path gets marked. Just like in Jagged Alliance. Unfortunately the decals can only be one color for some reason I can't fathom. The black footsteps mean it will take multiple turns. The red area means you won't be able to attack once you get to that point, using the currently selected attack mode.

I don't really want to reveal the whole project yet, but I am sure that glorious epic butthurt will ensue for reasons that will be obvious later on.

 

[20 Eyes]

I'm working on a 2D strategy/RPG game using XNA. I've made some (very) simple games with it, and now I'm trying to tackle something that is probably way out of my league. I'm comfortable making basic stuff like side-scolling shoot-em-ups, but I have bigger goals. I'm learning as I go, through reading tutorials, books on C#/XNA, trial and error, ect. The Microsoft RPG sample is very helpful.

The last iteration had a basic state manager (for switching screens), basic collision, working input through keyboard and mouse, a controllable character, a test map with a tileset I threw together in 5 minutes, several menu screens that went in and out of the gameplay screen, but not much else. I'm in the middle of rewriting a lot of it, though. My goal is to have the game look about as good as an average Super Nintendo game, just with better (PC oriented) GUI and a higher resolution. I'm not far enough to comment on the gameplay, but XCom, JA2, Final Fantasy Tactics are big influences. It's 99.999% likely to be vaporware, but it's a fun process. Right now I'm still early into the basic foundation of the game and looking forward to getting into the less tedious stuff.

 

[janjetina]

I'm working on a semi-complex production optimization project right now and in addition to that I'm developing a complex diagnostic program (data gathering takes up most of the time in that one). I've been working on the first program since the early spring (I have a fully functional working version as of today, I just have to add "bells and whistles"), and on the second one for two years (with a few smaller modules finished).

I've done some semi-complex stuff in the past as well.
For all projects prior knowledge is essential, but there are also specific things that I am learning as I go along.

No game programming for me, there is no money in it. Props to those who are into it.

 

[DakaSha]

I have worked on shitloads of bigger projects. All before i was ready and all were abandoned. Right now im just fucking around with conceptual shit. I think I'm ready to work on something a bit bigger (such as phone or flash games) but I need an artist so it's not happening anytime soon

 

[Cassidy]

Years ago I coded in BASH(glorious) an AWESOME button that downloaded, installed and configured automatically the latest nVdia driver for Linux without any user input other than executing the script in a terminal. Unfortunately it was very specialized in the specific distro and version it was made for due to its automation cutting off compatibility potential.

That was my greatest achievement in any kind of project related to computers and the only thing I ever finished, and I guess that says much in my level of interest and skills in coding. Even today the only "programming" language I sort of know how to use* is shell script and I gave up on seriously considering programming as either profession or hobby years ago.

As for modding, I dabbled with UnrealED, Neverwinter Nights toolset(which I remember hating), Thief's DromED, JRPGMaker, Deus Ex' UnrealED among others and usually all I managed before getting bored and giving up was to get a couple of base geometric shapes(in the case of the 3d editors) and basest NPC dialogues up. I guess I don't have what it takes or just always felt that the opportunity cost of not spending the same time with something more fun instead was too high.

Besides, nowadays doing Let's Plays is the lazy alternative for people who appreciate the idea of entertaining others sharing similar tastes without putting too much effort so why should I even bother trying to mod something in a over 10 years old engine that may end being crap?



* printf "Hello world" doesn't count

 

[Destroid]

I have worked on shitloads of bigger projects. All before i was ready and all were abandoned. Right now im just fucking around with conceptual shit. I think I'm ready to work on something a bit bigger (such as phone or flash games) but I need an artist so it's not happening anytime soon

Make game, acquire artist.

 

[Excommunicator]

Have you been working on a complex project?

Yes.

For how long?

The longest, around 6 years (so far), including learning processes.

Did learn things as you went, successfully or did you have knowledge of things beforehand/stopped to learn?

If you are pushing yourself to achieve something real, then it has to be both.

Did you give up(why?) ?

I'm not familiar with that term.

 

[The_scorpion]

Dunno if (larger scale) mods count as complex projects. In my mind, they were complex, given that some of the mods required to coordinate lots of people around the globe and required years of experience on the subject to even start working on them, then took years to be finished or publicly released in some form. Or canceled. For me, i learned most of the skills "on the job" but that may have to do with the fact that almost no information could be found beforehand, so that might be irrelevant for the O.P's purposes. The more easily accesible information you have beforehand, the better, but i guess there simply is a limit, depending on the topic, on how much theory you can learn before trying to apply it into practice. On the other hand, if you learn vital skills only later during the project, you might be sorry that you can't properly apply the new skills later on as the development has already progressed to a point beyond where you could still use the new skills to good effect.

 

[DakaSha]

I have worked on shitloads of bigger projects. All before i was ready and all were abandoned. Right now im just fucking around with conceptual shit. I think I'm ready to work on something a bit bigger (such as phone or flash games) but I need an artist so it's not happening anytime soon

Make game, acquire artist.

That is of course the most obvious answer and its not 'wrong' per se but ive just come to notice that it can be incredibly important to code WHILE having the artist on board for multiple reasons. I'm not saying i would never start writing without one. Just i would much prefer to have one. Also it is much more likely I would finish something when somebody elses hard work is on the line

 

[Forgotten Friend]

I have worked on shitloads of bigger projects. All before i was ready and all were abandoned. Right now im just fucking around with conceptual shit. I think I'm ready to work on something a bit bigger (such as phone or flash games) but I need an artist so it's not happening anytime soon

Make game, acquire artist.

That is of course the most obvious answer and its not 'wrong' per se but ive just come to notice that it can be incredibly important to code WHILE having the artist on board for multiple reasons. I'm not saying i would never start writing without one. Just i would much prefer to have one. Also it is much more likely I would finish something when somebody elses hard work is on the line

Most of the answers like how many polygons can characters be, how big can levels be how big textures, all need to be answered by the programmer. At first I had wanted to avoid the art side completely but it's just not realistic. Otherwise you are going to wind up realizing you have a bunch of completely useless shit at some point as happened to me.

 

[DakaSha]

There is also just the psychological aspect (and its not a small thing). Working on something this long without seeing results can be pretty hard to deal with

 

[shihonage]

I have worked on shitloads of bigger projects. All before i was ready and all were abandoned. Right now im just fucking around with conceptual shit. I think I'm ready to work on something a bit bigger (such as phone or flash games) but I need an artist so it's not happening anytime soon

Make game, acquire artist.

That is of course the most obvious answer and its not 'wrong' per se but ive just come to notice that it can be incredibly important to code WHILE having the artist on board for multiple reasons. I'm not saying i would never start writing without one. Just i would much prefer to have one. Also it is much more likely I would finish something when somebody elses hard work is on the line

Most of the answers like how many polygons can characters be, how big can levels be how big textures, all need to be answered by the programmer. At first I had wanted to avoid the art side completely but it's just not realistic. Otherwise you are going to wind up realizing you have a bunch of completely useless shit at some point as happened to me.

Yeah, or you get isometric sprites rendered in 6 directions while you need them in 8 directions, and no, that doesn't mean rendering just extra 2 directions, it means redoing everything.

 

[Forgotten Friend]

There is also just the psychological aspect (and its not a small thing). Working on something this long without seeing results can be pretty hard to deal with

You don't have to go the whole way. I plan to eventually get some kind of help, but you need to do a lot of groundwork before you should do even a little coding or artwork.

If I had to do again I'd learn the basics of art workflow before even starting. Then choose engine, then do a basic mockup which doesn't take long. Then have something to show while looking for people. It's very hard to get good people, takes a lot of time to communicate with them. You don't need to do it all alone but if you already have a basic design people can see and know the import process and the technical specifications you can easily communicate the project to people and make sure you only recruite people who are on the same page.

Otherwise you can waste a lot of time on people and wind up with no real help from them.

 

[Excommunicator]

It sounds like DakaSha wants the artist on side most of all as source of energy and motivation. Having someone who can visualise your ideas immediately is a powerful thing, as is knowing that they run on a wholly different source of energy than your own body.

So what are these things that you currently have in the conceptual stage, DakaSha?

 

[DakaSha]

It sounds like DakaSha wants the artist on side most of all as source of energy and motivation. Having someone who can visualise your ideas immediately is a powerful thing, as is knowing that they run on a wholly different source of energy than your own body.

So what are these things that you currently have in the conceptual stage, DakaSha?

Yep you're right about the artist

I work mostly on proceduraly generated content. right now my main thing is rewriting my (kickass ) random dungeon generator in a cleaner way in Haxe. Other then that i just do fun things. I taught myself haxe/as3 and started up my blog/website thing as a way to keep a journal. this way I'm actually finishing (smaller) projects and can look at them later. Its a way of forcing me to finish things (and it works so far ) and a huge source of motivation

Kinda helped that my one and only subscriber is literally the one guy that inspired me to start a blog xD
This guy: http://simblob.blogspot.com/

Game wise i just have a bunch of TBS and roguelike ideas floating around. *shrug*

 

[Forgotten Friend]

Just because I'm bored and masochistic, another screenshot with interface cropped out. Brought to you by Tom of Finland.

 

[tiagocc0]

It sounds like DakaSha wants the artist on side most of all as source of energy and motivation. Having someone who can visualise your ideas immediately is a powerful thing, as is knowing that they run on a wholly different source of energy than your own body.

So what are these things that you currently have in the conceptual stage, DakaSha?

Yep you're right about the artist

I work mostly on proceduraly generated content. right now my main thing is rewriting my (kickass ) random dungeon generator in a cleaner way in Haxe. Other then that i just do fun things. I taught myself haxe/as3 and started up my blog/website thing as a way to keep a journal. this way I'm actually finishing (smaller) projects and can look at them later. Its a way of forcing me to finish things (and it works so far ) and a huge source of motivation

Kinda helped that my one and only subscriber is literally the one guy that inspired me to start a blog xD
This guy: http://simblob.blogspot.com/

Game wise i just have a bunch of TBS and roguelike ideas floating around. *shrug*

Interesting, I will need to proceduraly generated content for my MoO2 like game, want to be partners in crime?

EDIT: I've been using Amit's site for inspiration too.

 

[DakaSha]

Interesting, I will need to proceduraly generated content for my MoO2 like game, want to be partners in crime?

what are you planning on having procedurally generated? I'm guessing at least the galaxies ^^
also: Are you planning on selling your game or will it be freeware?

edit: oh and what language/libraries are you using?

 

[tiagocc0]

Interesting, I will need to proceduraly generated content for my MoO2 like game, want to be partners in crime?

what are you planning on having procedurally generated? I'm guessing at least the galaxies ^^
also: Are you planning on selling your game or will it be freeware?

Yep, galaxies and opponents too. There will be a wide range of traits for the player to choose from, and there will be some ready to use races that the player can use or play against. But I want the option for the game to generate a random opponent too.
I'm planning on selling it, without DRM and the code will be released too to help people mod the game or create similar games on their own.

 

[DakaSha]

tiagocc0: I recommend you open a thread on the proceduarlly generated/random elements of your game i your sub forum when you have some time. Then we can talk there with the input of others and if it turns out I want/could help you in some way then we can sort out the details later.

It would just help me get an idea of what you even have planned