Ease of Implimentation...Major Consideration?

[Krice]

It's the programmer's responsibility to interpret the design and implement it in a progressively functional manner.

The design or plan should always be complete before it can work even in theory. One of the major problems is that the design is not complete (not detailed enough) or it has a flaw and it doesn't work when it's implemented. I think design/planning is the least known and mastered aspect of game programming, maybe because it's the worst part of it: Create a complete design and stick to it? No thanks, I'll just come up with something and see what happens.

[kipar]

There are plenty of things that are easier to design than to implement.
There are plenty of features that even harder to design properly, so they will be fun, not just complex and boring.

About the question - I often cut potentially funny features if their implementation seems too difficult. The problem is that I often don't look at their ideas. I'm sure that creating interesting multiplayer platformer with procedurally generated levels (such as described JumperRL) is very difficult, so such idea won't come in my head.
    I'm sure that creating interesting MMORPG (roguelike or not) is just impossible, so i won't even try.
    I won't try to create roguelike with detailed battle system, complex body parts and material system, just because I'm sure it won't be fun without some "spark" in design.
   Well, even creating something like ADOM or Angband, creating pure roguelike without "twist" seems impossible to me, because such roguelike will be boring. So in my projects I always use an additional ideas, e.g. unusual setting or battle mechanic - I'm not sure that i can create something interesting without them.

[tuturto]

I come up with new ideas and discard most of them in regular basis. Sometimes they don't fit to the theme of the game, sometimes they're hard to design well and sometimes they're too hard or laborous to implement. I'm deliberately not writing huge and detailed design document before writing code, because I believe that well planned is never implemented. Rather I like to design and implement one thing at a time, discarding and editing things as the design grows. The goal is to keep the game design and codebase as good as possible at all times. If I were to design everythign before starting writing code, I would be basing lot of design on assumptions.

[Krice]

If I were to design everythign before starting writing code, I would be basing lot of design on assumptions.

This is not related to programming but game design itself. In other words you need to test some feature to see if it works in gameplay's point of view. But it could in theory work already in the design. I think designing has bigger role when you get more experienced in game development. I guess it's something you don't learn quick.

[tuturto]

If I were to design everythign before starting writing code, I would be basing lot of design on assumptions.

This is not related to programming but game design itself. In other words you need to test some feature to see if it works in gameplay's point of view. But it could in theory work already in the design. I think designing has bigger role when you get more experienced in game development. I guess it's something you don't learn quick.

Yes and no. If I'm giving character class special attack to compensate slower healing rate, I'm assuming that slower healing rate needs compensation and that special attack is good for that. If I'm first playing around with character that has slower healing rate, I probably get idea if it needs compensation.

Granted, the more experienced you are, the bigger steps you can take in designing a game. I'm not Sirlin, so I have to take smaller steps.

[Ancient]

I find design much harder than implementation for the most part. Potential features get thrown away usually because of not fitting to game theme, being silly shout outs or not contributing to gameplay. Very rarely something pops up that is discarded because one would have to redo half of underlying architecture to include it.

Some features that are found difficult to implement or requiring significant extension to the engine are delayed. When enough good proposals that require container items (or similar solution) to be put into game you will eventually find them in PRIME.

When you have a great idea that fits into theme and would be very fun you have designed absolutely nothing yet. Design means you know by exactly what mechanics interaction will be handled.

I recommend taking five minutes with a pencil and sheet of paper whenever something potentially excellent pops up to your mind. With smaller ideas you can do all the thinking in your head. I have a hard time explaining thing without bad examples so here it goes.

For PRIME 1.9, we (me and Psiweapon) wanted more psykers. There was proposal of adding Bene Gesserit who would make use of The Voice. Sounds neat, eh? So decide how exactly it works! Thats design. Psiweapon handed me phrases for reverend mothers to speak and told what actions on part of the player would be blocked in effect.


 - "You will not show belligerence!" / unable to make direct attacks
 - "You will not read the Nam-Shub!" / unable to use computers
 - "You will not indulge your mutant nature!" / unable to use psi/mutant powers


Later I had a look at other timed effects and determined best durations for this to work so that getting commanded with The Voice is an obstacle but never something that would ruin your day. Unless you make several mistakes that is. Then one needs to compare the effects against other powers in PRIME. For example not showing belligerence is similar to fear in many respects. It means one has to consider whether it is still interesting and worth keeping this effect.

Next, implementation phase. In case of the Voice that was not difficult. Engine has supported timed conditions already so all I had to do was to handle user interface commands and modify some of them depending on current conditions. Finally, playtest. Nothing can substitute that.

To sum up, I sometimes discard ideas because of implementation difficulty but often just delay them. Ideas get far more often dropped because of design considerations.

[st33d]

I've just resigned myself to the notion that RPGs take a lot of time to make.

[requerent]

I've just resigned myself to the notion that RPGs take a lot of time to make.

Seven days?

[guest509]

  Wow. Learning a lot about people's process.

[Pueo]

One of the major problems is that the design is not complete (not detailed enough) or it has a flaw and it doesn't work when it's implemented. I think design/planning is the least known and mastered aspect of game programming, maybe because it's the worst part of it: Create a complete design and stick to it? No thanks, I'll just come up with something and see what happens.

Huh, I think I find myself saying (about my project) "Have a complete design but need to implement it? No thanks, I'll just keep designing."  I guess that's just me, always thinking, never doing. I think the major problem for me is I'm not a guru programmer, and then it's like, "Oh, magic should be able to teleport you through walls! How do I do that?!"

I also like Krice's comment about how the best way to build a RL is to start with a robust engine and slowly add features.