Why do keyboards have caps lock at all? I suppose, there was some kind of need for caps lock when keyboards were invented, but I can't figure what that might have been

Usually the key just annoys me ...

Edit: Please don't say caps lock was made so that people in MMOs can shout and spam louder messages

Here you can check arbitrary LCGs for patterns:

http://www.vias.org/simulations/simusoft_lincong.html

In some games such pattern structures are annoying and you want a better RNG. In other games it doesn't really matter ...

One thing people could answer me is how hard it is to make a good custom pseudoRNG. Mostly out of curiosity, not that I'd have much of a use for it.

If you want a mathmatically sound generator, and do not have a good background in math, then do not try. (If you ever see books about RNGs in a library, you'll know why I say this )

Better just use one of the well known algorithms. The page that I linked in my first posting in this thread shows a simple but often "good enough" sort of RNG. If you need a better one, I'd just take a Mersenne Twister implementation, but not try to code my own.

A long time I was using my own LCG implementation, though, and it worked nicely.

You can think of the random seed as the first number in that series of numbers. If you always seed (initialize) your RNG with the same number, it will always generate the same series of numbers. This is actually useful for storing replays and similar stuff where you want the same "random" events to happen every time.

There are algorithms to create maps for games from a single seed, by using such RNGs. It's pretty cool to see a whole world grow from a single number

I once did that with solar systems for a space exploration game - the seed was just the position of the system in space, and so I didn't need to store millions of planetary configurations, but could create the whole game universe from just a few numbers.

It can be a mighty concept if applied right.

As Increpare pointed out, seeds are not portable between different RNGs. Thus if you want your program behave the same on all platforms, best is to use your own RNG implementation - or a free and portable one like the Mersenne Twister, which seems to be the most liked among developers anyways.

Internally any RNG keeps a state (a value or a group of values) to calculate the next "random" number upon the next call. The seed sets this internal state. In the most simple case, the seed just will be the internal sate for the next calculation.

Maybe this page also will help:
http://en.wikipedia.org/wiki/Linear_congruential_generator

In case of such an RNG, the seed is just the start value of the sequence of internal states. The first section of the Wikipedia page should explain it, and it's heavier on math than English ... if that is of any help

edit:

Trying to explain with a very simple RNG implementation, that shows the facts quite clearly:

Code:

new_seed = (seed * 16807) mod 2147483647

In this case the "random" numbers on each call are already the new seed for the next call (it's also the "random" result that you get on each call for this sort of RNG). More sophisticated RNGs have more complex formulas, but the idea of the seed being the base for each computation stays the same.

... but yeah, you could potentially use it for a shmup, with some creativity.  It's meant for static tiles, but you could use a tile (or a tilesheet of enemies) symbolize starting point for a baddie, though the editor won't really help you with editing behavior of enemies or anything.

That's what I have planned. One layer of the map will define "spawn points" for enemies, the number of the tile there will define the sort of enemy to spawn there. Movement patterns will be defined in some config file.

Will be fine to work together with someone Usually this also spawns a whole lot of new ideas, and it's fun.

Bad is though that work will keep me very busy till mid or even end of next week

But I'll be back

I'm so indy, I can't understand my own games.

That's one way to go about it. But with some left-field thinking you can still make it non-violent. For instance, the players are bugs in a house, shooting jars and causing mayhem (and no one dies, of course). For cooperative play, light-hearted settings are usually best, with silly details to laugh at.

Cool idea I love this forum for that, so many creative people here! Thank you for helping me with your suggestions.

[snip some more ideas]

Today my brain just isn't working. But I'll keep your suggestions in mind, an hopefully on a better day I can draw some inspiration from them.

So, with all this said --which I actually wrote before I read your last post--, I think that those graphics are okay to work with, but could use some spicing up.

On one hand I'm happy with them, since they trigger some memories of former games that I liked. But of course that is just me. And indeed there is room for improvement, some of them were made within a few minutes, just to get an idea how the elements will fit together if I go this route.

On the other hand, your ideas and suggestions are a whole lot more creative. Particularly the "things that come in pairs" theme looks intriguing to me. The idea of a split powerups, like a left and right boot, that both players need to collect, and which then will enable the "bad ass TIGer boot cannon" gives me a smile.

I'm not sure if I'm creative enough to get far on this route. But I like such fun and silly ideas. Once my head is easier again, I should also become more creative. Maybe I can make something like that

Thank you very much for the ideas and feedback!

Tiled is really good. I'm subscribed to the mailing list and spying on the development

But in this case it seemed easiest to reuse an older map editor of mine. (With that I can also reuse map data structures, IO code and similar).

Thanks for all the suggestions, though, they helped to get a good overview of the tools available

I'm currently making a small game (as a short break from ArcMagi) based on the same idea. Cheers there, and best of luck - we definitely need more cooperative shmups  !

Thank you! Arc Magi looks like a great, but also very work intensive project. Good luck with that, what I have seen so far is really outstanding!

Going with the "shapes" theme, I've been working on some bubbly scenery tiles. The structure loosely resembles a tunnel, and the player will navigate the vessel from left to right, scenery scrolling leftwards.

The "diamonds" and the "globes" are enemies, the "coins" will be bonuses to collect.

Whee, answers and suggestions

@December:

My first testing code scrolled vertically, but currently I favor horizontal. Even with a 4:3 display, if one subtracts the status area at the bottom, the display is like widescreen, and actually the larger distance across makes it a tad easier to react to enemies and obstacles. I think I'm pretty much decided for horizontal scrolling.

What I had in mind was a mix of obstacles (parts of the scenery that will block the ships path), and formations/waves of enemies that move in interesting patterns. I don't know about bullet patterns, some of the enemies will shoot aimed at the ship, other will shoot in fixed directions, kind of curtain fire.

My first try was in 640x480, windowed mode. I think I'll stick with that, just because of not having a better idea. I'll try for a tile size of 32x32 to build the scenery and background. Must see how that works out. A test with a whopping two of background tiles looked alright.

Thank you for the pointer to "the2bears" that is looking very interesting. I'm sure I can draw a lot of inspiration from there

@agj

The idea was born from a discussion about the lack of games with cooperative modes, where two or more players play together. I don't really know if this "lack" is true or was just the impression of the people involved in the discussion. But it reminded me of an old shooter called "Blood Money" (Atari ST/Amiga, I think also on other systems), that had a nice two player mode.

The only real concept is, to let two players play together, cooperatively, and design the game at least in some parts so that they really need to play as team, not as two solo players. Besides that it is meant as easy fun. No big message to communicate there, I don't want to make political or religious games. In my past projects "peace" and "cooperation" have been messages, but "peace" obviously does not work in a shooter, so "collaboration" is left.

Considering this, I think I should make the game's artwork abstract, to escape or at least ease the moral impacts. Shooting things that resemble living creatures could trigger the wrong ideas, shooting inanimate things like robots is less problematic, ethics-wise, but still destructive. Shooting colored shapes might be the most innocent way to build a shooter.

Still this leaves the question of style. I have a few ideas, but must test what will work well in the game.

@Architekt:

I'll read through the thread, with some luck I can borrow some ideas from there. Thank you for the pointer

Sometimes it might be helpful to point people to the manual ...



Now I felt reminded of a smiley that I liked, but cannot find anymore. It's been a blueish smiley, ranting heftily and lengthily about something, until exhaustion stops him. I've tried to google, but cannot find it, and I don't remember which forum I have seen it on. Can anyone help?

In a moment of nostalgia, I decided to try my hands on a little shmup, in memoriam of a game that I once liked to play.

As a programmer I first focused on the technical aspects of the project, maps, ships, scrolling, player input, collisions and all that. Then I realized that I'll have a whole lot of options to present this to the player.

And now I feel a bit clueless. Starting with more technical questions like resolutions (320x200 like the old games, or use the capabilities of nowadays hardware?), the style (cartoony, realistic, surreal ...) the scenery (planet surface, space, submarine, tunnels of some sort) I feel puzzled, and uncertain which path to take.

I guess it is not that helpful to post such blurry questions here, but maybe someone has experiences with making shmups and can tell me what worked for you, or what didn't?

I can pixel to some extend, do some types of cartoony graphics, and if needed I can model and render realistic/surreal things. I'm just terribly undecided which path to go.

(Please forgive me the questions if they look stupid to you. I'm not a game designer, but mostly a programmer. That's why I ask for help with design, even if it's only graphics design and not game design.)

if you have a bigger group of testers, you can let some play the game as they want, and others get list of things to test - these lists will be made from bugs/problems found or suspected during development and earlier testing stages.

E.g. if you have a game about traffic and traffic lights have been a problem in the past, advice some testers to check the traffic lights for proper operation extra carefully.

Using gaming illiterate people as testers will also test the accessability/usability of your game, but you must expect less precise reports, at least expect that the terminology of such people does not match what you are used from gamers. E.g. gamers have a good idea of what "ping" means and "lag", but non-gamers will use different words for such effects.

Thank you, that is looking interesting, too

Next steps will be to decide what scenery to make, and what resolution to use. But I guess I'll post a new message over in the "Art and Design" section for that. Thanks all for the pointers and help so far

Nice

I'm positive that I can get the scrolling good enough. Next will be to decide on a style for the tiles, and then make a little tileset for testing the engine.

Which reminds me that since quite a while I wanted to play with a particle engine, maybe this project could be a good application for such. Time will tell, one step after the other

Slick looks good. Yesterday I checked LWGL which is used by slick to talk to OpenGL, but it didn't like my graphics card/driver ... but I had success with JOGL, which is easy to use, too.

Just stupid, that plain Java 2D API seemingly can not do screen/window updates at a steady pace, but has unpredictable jitter, which makes scrolling a bit jerky. I'm still trying to get along without OpenGL, because I don't want to bundle the fairly big libs with my tiny project - if it doesn't work out with plain Java, I'll use an OpenGL wrapper.

At the moment I think I'll use my old map editor, it was easy enough to adapt  for the new task. Now needing to make some tiles for more testing. Thank you all for the feedback

Thank you, I'll check it, and see if it's easier to write a file format converter, or expand my own editor.

So far I have a scrolling demo, woot! But Java is stupid if it's about smooth scrolling Must see if I can improve that or if I need one of the OpenGL wrappers.