next »

[Bishop]

Hey,

I'm currently working with a friend on an RTS game with destructible environments. You play as a band of space pirates raiding enemy ships, all the combat takes place on board the enemy space ship you're raiding. You can do things like disable the enemies shield generator to then be able to ship to ship missiles or storm the security control room to remove fog of war (cus you can see out the ships CCTV). We're hoping for highly emergent gameplay by trying to simulate the workings of a ship as well as offering dynamically destructible terrain.

Onto the issue we're facing. We're creating the destructible walls using pixel perfect collision in a worms style geomodding. We've got the path finding implemented but processing diagonal walls is effecting the frame rate. I feel we should limit walls to vertical and horizontal only and move to a more grid based system but my friend feels thats too large a sacrifice.

Do you feel it's right to change the design to cut time out of development (or increase time for other areas)? Or should we never compromise the original vision?

PS. I don't want to get too into the nitty gritty details on the technical side. Lets assume we can get it working fine, but it'll take 2 months of upkeep, or we simplify the game layout to save those 2 months. Thoughts?

[Udderdude]

Tile based terrain worked for Starcraft .. their idea of destructable terrain is large objects you get all your units to smack around for 7-8 seconds.  I think you'd be fine scaling it down a bit.  The overall effect on gameplay should be about the same, anyway.

[rivon]

I'd do what feels the best/most fun. So, if you can get the simple tile-based version working in a few days/week, then try it. If it feels bad/not fun, do the original version. It's little possible sacrifice which would save you a much bigger amount of time if it proved fun.

You can also experiment with tile sizes. The destructible ones don't have to be as big as the original tiles.

[Wireless Mouse]

My advice is this: don't confuse niggling details like diagonal walls with the overall vision of your design. I can't think of any games I ever played where I thought to myself, "This would be the best thing ever IF ONLY it had diagonal walls!" Nope. That's (probably) never going to happen.

[Moczan]

And you don't even have to sacrifice all diagonals by going to tiles, creating a 3-4 variants of diagonal tiles will give you much more flexibility when creating levels and still have the speed benefits of tiles.

[Archibald]

If this is your first game, then it is no brainer, do everything you can to cut down development time. 99.999% of games are never finished or drag too long to finish. You can always release a sequel with all the features that didn't make it into the first one.

[C.D Buckmaster]

Well, how do you think changing it will affect the player's experience?  Would the lack of true diagonal walls hurt their enjoyment, are they even necessary in the first place?  Think about how a player would interact with a diagonal wall, and if it would be fun or intuitive.

[stevesan]

I really don't see any particular reason why pixel-perfect collision would make an RTS more fun. My favorite game with destruction was X-Com, and that was completely tile-based. So unless you have a compelling game play reason to keep it, it's probably not worth the technical hassle.

Game development is about change and adaptation. It is absolutely OK and often necessary to change from your original vision - if you're doing something innovative, you can't possibly predict all the problems that will come up, so your original vision is probably mostly wrong anyway.

It would be a shame if you tried to do it, and then you got discouraged cuz it was introducing too many technical issues. Or you release it, but it's so finicky and buggy due to that part that no one enjoys playing it.

Simplicity is a virtue.

[stevesan]

I'd do what feels the best/most fun. So, if you can get the simple tile-based version working in a few days/week, then try it. If it feels bad/not fun, do the original version. It's little possible sacrifice which would save you a much bigger amount of time if it proved fun.

You can also experiment with tile sizes. The destructible ones don't have to be as big as the original tiles.

This. Right now, it may feel painful to let go of your original vision. But I find that once you implement the simpler thing and see how awesome and workable it actually is, that original vision doesn't seem to precious any more and you'll toss it aside without a thought.

[C.A. Sinner]

I really don't see any particular reason why pixel-perfect collision would make an RTS more fun. My favorite game with destruction was X-Com, and that was completely tile-based. So unless you have a compelling game play reason to keep it, it's probably not worth the technical hassle.

yeah. adding detail does not necessarily equal adding quality

[Fandrey]

It depends on that "2 Months" vs "Actual time budget" of the game. If it's 2 months of a 2-year-time-budget, it's still worthy to check it out. If that 2-months is, say, 30+% of actual time budget, i'd say you better simplify it.

I do think over-detailing things in an RTS isn't necessary though, since most likely players will be too busy to notice that (while you still sacrificing your development time and game performance).

Often times, ideals just don't click with reality.

===

I'm sorry if i'm not explaining it well. My english just sucks.

[SHilbert]

IMHO, I doubt people would consciously notice that it's missing diagonal walls, especially in man-made structures like space ships. So I wouldn't worry too much about removing diagonals if they don't serve a gameplay purpose. They can definitely make things more attractive visually but they don't always add to gameplay.

I know of at least one game where the doorknobs on doors are always on the right, regardless of whether you are looking at the inside or outside of the door. You would never notice unless you were told. I think it's kind of the same thing.

I can sympathize with unexpected pain from diagonals; I am tempted to remove them entirely from my game as well. I already had to remove 45° angles. (Don't ask.)

[Ashaman73]

Do you feel it's right to change the design to cut time out of development (or increase time for other areas)? Or should we never compromise the original vision?

I'm doing it all the time, to be honest. Adding some features, making them far to complex and cutting it down at the end to make it easier and better. Here's a nice quote

Perfection is archieved not when there is nothing more to add, but when there is nothing left to take away.

[Bishop]

Yea, this is generally how I felt about it. I showed my buddy this thread, trying to convince him to cut the corner, but he's still insisting it gets done. Now that he's going to do all of that code though, I have to admit it's alot easier when it's not your own 2 months, to just agree. Also I didn't make it fully clear, the diagonal will enable curved and odd shapes as well, although I doubt they'll be many of those.

I'll see if I can persuade him to get it working aligned first and then see if he can be bothered. That a really idea.

[Graham.]

Work on the work you believe in; let him do the same. If he's happy, you should be happy.

It either one of you ever feels conflicted about your own work, like ever, then cut something. _Always_ cut something. The only time you don't cut is when you have an externally-motivated deadline. Why would you work on something you don't believe in? That's what human intuition is for. Conflict means something.

A lot of the time the process of cutting forces you to re-evaluate what's "really important" in your game. If you can't choose your priorities, it often means you don't really understand what your game's about anyway. These kinds of conflicts are really opportunities to hone your design. They mean that you haven't been clear with yourself about what you're trying to achieve. If you become clear, your design will improve. Sacrificing is always a healthy move.

The Intel "M" series (the pre-cursor to the duo-core line) was a re-design to save on power consumption for laptops. They cut half their shit out of there, and the processor became twice as fast for the same cost (_and_ lower power consumption). Then they slid into duo-core. Nice, right? When your shit doesn't serve your goals, you redesign, and you end up with something even better than you originally imagined.