Galactic Sovereign Development Log

This is possibly the final version of the emitter program, aside from adding in actual alpha blending options, which I keep forgetting to do.  It’s certainly not perfect, but it’s not like it’s really part of the game, so as long as it’s complete enough to kick out some good effects with minimum hassle, that’s good enough.
New features:
Each complete emitter system can have many individual emitters, each one individually tunable.  This allows for creating much more sophisticated effects, and is the way the Unreal Development Kit does it.  Each complete emitter corresponds to a single file.
Emitters can now be saved and loaded from files.  The interface for doing so is on the top of the screen.
Fixed that flickering/blinking screen problem.  Then created another slightly different one.  Then fixed that one too.
You can now define 100 separate data points instead of just 10 on all the “graphable” variables.  So you can now have essentially instantaneous changes, or much smoother curves- I found the 10 point system was visibly jerky with some transitions.
There is now a “displacement” mode.  Instead of just saying how fast stuff is being spit from the center, you can just graph exactly how far it is over time, if you want.  So you can have stuff that goes back and forth, stays in the same places, goes towards the origin, or starts somewhere other than at the origin.
There is now a “rotation” value.  This spins all the particles around the origin.

Also included is a couple of fairly silly-looking tech-demo emitters and a png with instructions and controls on it.

So, next up: Back to actually working on the game, instead of emitters.  Next thing to add: Support in the game for actually “attaching” emitters to objects, then giving the player’s ship a rocket trail emitter, which I feel is really going to be a significant improvement.  Then a wormhole emitter.  Then wormholes, and multiple galaxies.

emitterV3

Well, the emitter code is in, and you can now play with actually spraying particles around.  I’ve included a sample of what you can do by messing with the color values.  It’s a little inefficient, though, if you spawn a whole lot of particles.  On my computer, it drops to 60 FPS at 10,000 particles being processed per frame.  I’m just not really sure if that’s good enough or not.
Some notes on use:
First of all, type in a number and hit enter to change whatever value is red.  Click on the names on the left to change which value is selected.  Click twice to pick the value in the second column.
The first column (VAL) is the overall value, the second column (VAR) is how much randomness there is for that value.  Values are evenly distributed throughout that range, though I’m considering changing that.
For the values where a graph shows up on the bottom, draw on it with the mouse.  The X axis is the lifetime of the particle.  It doesn’t update the emitter immediately when you change the graph like it does when you type in new numbers, though- to force it to update, hit “R”, for “refresh”.  Also hit “R” if it is flickering or displaying giant colored boxes- it does that sometimes, I’m pretty sure I know why, but I haven’t fixed it yet.

Valid values for “texture” are 0 through 16.  I particularly recommend 8, and 15 and 16 are good once you’ve set up some color changing stuff with the RGB graphs.  Those two are the only two new textures, the rest of them are just textures that were already being loaded for other stuff.
The top value is now how many milliseconds it waits per spawn.  Values between zero and one are valid.
“Acceleration” doesn’t do anything right now.
The VAR value for “ejection dir” will change how wide a spray it is.  It’s in degrees.  The VAL will change what the main direction is.
The lifetime of the particle is actually measured in ms, but displays in seconds for simplicity.  So for 1 second, you actually have to type in 1000.
Hit “E” to export your values to emitter.txt, or “X” to export the “expanded” values- for every ms in the lifetime of the particle, the varying values are precomputed and saved in the “expanded” version.
So, up next: enable viewing multiple emitters at once for building more complex effects, and write the loading function, so you can load emitters instead of just saving them.

emitterV1

First a little background: If you aren’t familiar with what an emitter is, it’s something that spits out a bunch of semi-transparent images (individually called particles).  You set a bunch of rules about how the particles behave, maybe set up some basic physics, give them some randomization, then start spitting them out, and you can get some pretty nice looking stuff which is actually fairly simple in terms of code.  So there are a whole lot of examples up on youtube.  This is approximately the complexity level I’m aiming for. (that link has music.)
The game really needs some engine trails, if for nothing else than to let you tell which corner of the ship is the front at low levels of zoom, so it’s emitter time, and I decided not to go down the dead-end street of typing in hard-coded values, or tuning them via a file like I do with most of the data in the game.  No mere mortal can just type values in and know what it will look like, so you’ve really got to have some realtime feedback.  So I’m working on a full-on emitter generating program.
Right now, I have the UI stuff down, and a big empty space ready for the actual emitter to go into.  Now I just (just…) need to write the actual emitter code that will take all this data and turn it into what it’s supposed to be.
So, a totally new program is included.
Most of the names will be self-explanatory, except “val” and “var”, which are respectively “value” and “variance”.
q to quit like always
d toggles between display of the raw data and displaying the emitter (which doesn’t do anything yet)
pick the thing you want to edit on the left by clicking on it
Type in a number and hit enter, and you’ll give a new value to whatever number is highlighted in red.  Decimal points allowed.  Note that if you enter something above a few decimal places, it will start giving scientific notation with an exponent and the number won’t fit in the column any more, but numbers that big aren’t really useful in emitters, so I’m not worried about it.  Also, it will take full float precision, but only display the first three decimal places.
Where it gets interesting is if you pick one of the values in the middle, where it makes sense to change the value over the lifetime of every object emitted, and you get this box at the bottom where you can draw a line to represent a multiplier value to be applied over time.  That part I think is actually pretty neat.
And finally, hit e and it will export the entire dataset into emitters.txt, where it can (eventually) be read in by the main game.

emitterUI